Volume 13 Issue 2

Never Trust Always Verify

Authors- Mr. Prakash Hongal, Mr Shrikant Jogar, Ms. Aishwarya Vernekar

Abstract-The Zero Trust Security Model (ZTSM) has proven to be a modern alternative to traditional perimeter-based network security, achieving a significant driving movement in recent years. Traditional perimeter security architectures are plagued by the risk of inherent vulnerabilities, particularly the single point of failure, making organizations susceptible to cyber threats. In response, businesses will not accept implicit trust and shift to Zero Trust, a security paradigm that treats all hosts as if they were exposed to the open internet. This model enforces continuous authentication, the cheapest access to privileges, micro segmentation, and rigorous identity testing to prevent unauthorized access and lateral movement of threats. This article examines the Zero Trust architecture framework, its implementation strategies, benefits, and comparative benefits compared to legacy security models. To ensure robust protection against modern cyber threats, we will set up a multi-layered zero trust architecture that includes three cascade layers, four enablers and five nuclear safety attributes.

Hospital Management System

Authors- U.Harshitha, Y. Sudharshini, S. Shiva Kumar, Vineeth Kumar, S.Anusha

Abstract-This paper presents a web-based hospital management system that allows patients, doctors, and administrators to interact with the hospital’s information system through a web interface. The system is built using HTML5/CSS3, JavaScript, Bootstrap, XAMPP, PHP, MySQL, and TCPDF technologies. Web-Based Hospital Management System (HMS) enables various hospital and medical processes to be performed online. It consists of registration, login of patients, and booking their appointments with doctors storing their details in the system. It provides a login page for patients, doctors, and admins each have their username and password. It consists of three modules. Those are the patient, doctor and admin. This Web Application maintains authentication to access the information. Administrators can see patient and doctor information and appointment schedules and add new doctors as part of administrative tasks. A database was created one for the patient and the other for the doctors so that admin can access it. The Patient module includes booking appointments and checking prescriptions. A patient can pay a doctor’s Fee online. The doctor module allows doctors to view appointments, give prescriptions and search for patients. Web-based technology provides a wide range of online services in practically every industry. The majority of jobs may be completed online, which helps to minimize the workload, expense, and effort. The paper discusses the concept of a web-based platform that would enable various hospital and medical processes to be performed online utilizing Web networking technologies, which could be crucial for implementing the functionality of online medical administration. This will aid in the administration of patients, the management of doctor schedules, and the maintenance of patient data that are accessible throughout the hospital online patient data storage, management, communication, analysis, and updating. Therefore, by implementing this web-based application many tasks that would be time consuming and inconvenient can be accomplished.

Placement Management System

Authors- Annam Vaishnavi, Ravula Manasa, Kanukuntla Nikhil, S.Anusha

Abstract-From a student’s perspective, placements can bring a wide range of benefits and opportunities. Training and management of placement is a crucial part of an educational institution in which most of the work is done manually. Manual system in the colleges requires a lot of manpower and time. With this project we aim to develop a web portal to solve this issue. The project is aimed at developing an application for the placement department of the college. The system is an application which will be accessed and effectively used throughout the organization with proper login enabled. It can also be used as an application for the Placement Officers in the college to manage the student information about placement thus reducing the manual work and consumes less paperwork. The system also provides the facility of viewing the personal and academic information of the student. The system gets the requested list of candidates for the companies who would like to recruit the people according to their eligibility criteria.

A Review of Thermal Analysis Techniques for Engine Cylinder Fins with Varying Geometries

Authors- Research Scholar Rajat Yadav, Assistant Professor S.N. Dubey, Professor Dr. Mohammed Ali, Assistant Professor Dr. Palash Soni

Abstract-This review examines the thermal analysis of engine cylinder fins with varying geometries, focusing on their critical role in heat dissipation for internal combustion engines. The study evaluates four distinct fin designs—angular, curved, rectangular, and cylindrical—using Computational Fluid Dynamics (CFD) analysis performed in Ansys Workbench. The research demonstrates that fin geometry significantly impacts cooling efficiency by influencing temperature distribution, airflow patterns, velocity profiles, and pressure distribution across the engine cylinder surface. The analysis reveals that curved fins exhibited superior thermal performance compared to traditional designs, achieving higher heat transfer coefficients and more effective cooling. This enhanced performance can be attributed to the curved geometry’s ability to create optimal turbulence patterns that maximize contact time between cooling air and fin surfaces. The findings highlight the importance of optimizing fin design beyond conventional rectangular geometries to improve thermal efficiency in air-cooled engines. This review synthesizes these results to provide insights for engineers and designers seeking to enhance engine cooling systems through geometric optimization rather than material substitution alone.

DOI: /10.61463/ijset.vol.13.issue2.188

AI Mock Interview Chatbot Using Gen AI

Authors- Mr.Madanachitran R, Austin A, Balaji K, Rajappan M

Abstract-Job interviews are a critical component of the hiring process, but several interviewees struggle to articulate their skills. In this paper, we propose developing a Generative AI-based mock interview chatbot that offers real-time, interactive, and adaptive interview practice. Unlike traditional mock interview methods that are costly and non-scalable, this chatbot leverages Gemini AI, React, and Next.js API to generate industry-domain interview questions programmatically and deliver instant feedback. The platform relies on Neon PostgreSQL for database and Clerk for role-based authorization to offer users an enriched experience. The chatbot dynamically adjusts question complexity levels to the user skill level (beginner, intermediate, expert) and offers technical, behavioral, and situational answer insights. By integrating AI-based feedback processes and performance metrics, the platform encourages user activity, self-esteem, and actual interview readiness. In comparison with other available techniques like pre-recorded video questionnaires and template-based chatbots, the proposed system offers end-to-end data-driven interactive mock interviewing. Deployed on Vercel for hassle-free scalability, this AI-based career solution aims to bridge skill gaps, improve candidate performance, and maximize job market fit.

DOI: /10.61463/ijset.vol.13.issue2.189

Generative Chat Model for GitHub Repository Using LLMs

Authors- Dharmaraj J, Graceshon J S, Murugan S, Dr. J. Yogapriya

Abstract-GitHub repositories serve as vital resources for software development, containing source code, documentation, and configuration files that define a project’s architecture. However, navigating and understanding complex repositories can be challenging, particularly for developers unfamiliar with a given project. This paper proposes an interactive platform that leverages Natural Language Processing (NLP) and a Large Language Model (LLM) to enable users to query GitHub repositories efficiently. By inputting a repository URL, users can ask specific questions and receive detailed, context-aware responses generated by the LLM. The system processes various file types, including README.md, source code, and configuration files, to extract relevant insights. The frontend is developed using HTML, CSS, and JavaScript, while the backend utilizes Flask for data processing and routing. This model surpasses traditional keyword-based searches by offering human-like, semantically rich responses, significantly improving repository accessibility and developer productivity. The proposed system provides an innovative approach to enhancing software repository comprehension using AI-driven conversational agents.

DOI: /10.61463/ijset.vol.13.issue2.190

Co-Training Transformer for Remote Sensing Image Classification, Segmentation and Detection

Authors- Mrs.E.Nivaditha, S.Dhinesh, M.Gokulanand, S.Manojkumar

Abstract-The rapid advancement of remote sensing technologies has significantly enhanced the ability to monitor and analyze the Earth’s surface, facilitating applications in environmental monitoring, urban planning, and disaster management. Remote sensing image classification, segmentation, and object detection are fundamental tasks in this domain, enabling automated interpretation of satellite and aerial imagery. Traditional machine learning models, including Convolutional Neural Networks (CNNs), have shown promise but often struggle to handle the diverse and complex nature of remote sensing data, which may include variations in scale, resolution, and lighting conditions. In this work, we propose a novel approach for remote sensing image analysis by cotraining a Transformer-based architecture with a state-of-the-art detection framework, specifically Detectron2. The Transformer model excels at capturing long-range dependencies and global context in image data, while Detectron2, a high-performance object detection framework, is capable of fine- grained instance segmentation and detection tasks. By co-training these two models, we aim to leverage the strengths of both architectures to improve classification accuracy, segmentation precision, and detection performance for remote sensing images. The co-training strategy involves joint learning where the Transformer model is used to capture semantic features from the remote sensing images, while Detectron2 refines spatial boundaries and detects objects of interest. We demonstrate the effectiveness of this approach through a series of experiments on benchmark remote sensing datasets, showing significant improvements in classification accuracy, segmentation masks, and detection performance compared to traditional methods, solutions for real-world applications such as land cover classification, urban monitoring, and disaster response.

DOI: /10.61463/ijset.vol.13.issue2.191

Intelligent Search and Rescue System Using IOT

Authors- Mrs.P.Saranya, S.Gowtham, K.P.Hariprasanth, P.Manivel

Abstract-Recently personal security has become a sensitive issue. Small kids, ladies, as well as aged people need to have their secure against kidnapping, rape, chain snatching respectively. There are different areas & scopes of security. Recent social incidents gave us motivation to develop personal security system. Kids, aged people & ladies mostly not able to fight to criminal for self security. Today’s world is full of rush and most of the women work independently to support their family. They have to work till late night. For such women, safety is the most important requirement. The security issue for such women comes forward because cases of harassment and rapes on those women are increasing. Best suitable system for those women will be a portable system which the women will be able to carry will her and easy to use. Portable system will generate a shock which will make to attacking person to get back. After generation of shock the message will be sent with the help of Global System for Mobile Communication (GSM) on the particular number stored and the location of those women is traced with the help of Global Positioning System (GPS). If the message is not checked by the particular number mentioned, the system will continuously give the call until the message is checked by the particular number mention.

DOI: /10.61463/ijset.vol.13.issue2.192

A Study on Physical and Mechanical Properties of Reactive Concrete

Authors- Jegatheesh Kumar. S, Assistant Professor Poonkuzhali. A, Neelasundaravalli. M

Abstract-The term reactive concrete has been utilized to depict a fiber-supported, superplasticizer, silica rage concrete combination with exceptionally low water-concrete proportion, described by the presence of extremely fine totals all things considered of the common total. It has having compressive strength of about 130 MPa. Filaments are consolidated in Reactive concrete to upgrade the break properties of the composite material. Reactive concrete is perceived as a progressive material that gives a blend of super-high-strength and brilliant solidness. Nonetheless, the creation of responsive powder concrete isn’t yet accessible with the restricted examination in this region. This paper researches the mechanical and new concrete properties of reactive concrete.

DOI: /10.61463/ijset.vol.13.issue2.193

Designing a 3D Quantum Circuit Structure Using Interconnected Hexagonal Flowers for Enhanced Redundancy and Decoherence Reduction

Authors- Himadri Maity

Abstract-Quantum computing has made significant advancements, yet challenges such as decoherence, error propagation, and scalability continue to limit its efficiency. This paper presents a novel 3D quantum circuit architecture that leverages an interconnected hexagonal structure to enhance fault tolerance and minimize decoherence. The proposed design consists of 314 qubits, including 2 control qubits and 312 target qubits, distributed across 20 hexagons arranged in a hollow cylindrical structure. Each hexagon contains 19 qubits, positioned at both its vertices and internal intersections formed by edge connections. A key feature of this architecture is optimized gate efficiency, where each qubit is connected to at least four others using only two active CNOT gates per qubit, significantly reducing gate-induced errors. Additionally, redundancy is implemented across the structure, ensuring that if decoherence occurs, the lost quantum information can be recovered through alternative pathways. Hadamard gates are applied to all qubits to maintain superposition, while Quantum Non-Demolition (QND) measurement and weak measurement techniques are utilized to enhance measurement precision and minimize quantum state disturbance. This research introduces an innovative, fault-tolerant, and highly scalable quantum circuit, paving the way for more robust quantum computing architectures. By addressing fundamental limitations in decoherence and quantum error correction, this design has the potential to surpass existing superconducting quantum computers in both reliability and computational power.

DOI: /10.61463/ijset.vol.13.issue2.194

Karshaka Mithra: Transforming Indian Agriculture with Technology-Driven Solutions

Authors- Bysani Saroja, Myla Visanth Kumar, Sasanala Ramesh, Chinthala Seetha

Abstract-Indian agriculture faces significant challenges, including unpredictable weather, fluctuating market prices, limited access to timely information, and regional communication barriers. To address these issues, ‘Karshaka Mithra’ was developed as an innovative, technology-driven platform aimed at empowering farmers with real-time insights, advanced tools, and collaborative opportunities. This comprehensive solution integrates multiple features such as accurate weather forecasting, multilingual voice assistance, market insights, and government support resources, all accessible through a user-friendly interface. Real-world incidents, such as unseasonal rains damaging crops in Odisha (2020), price crashes distressing onion farmers in Maharashtra (2019), and pest infestations affecting wheat yields in Punjab (2021), underscore the urgent need for such a platform.Karshaka Mithra offers a five-day weather forecast, real-time updates on humidity, rainfall, and storm probabilities, and a multilingual “Speak Out” feature to support illiterate farmers. Additionally, it provides a community interaction platform for collaboration, daily expert-led sessions, detailed market insights for informed decision-making, and a resource hub featuring content on modern farming techniques and drone technology. To enhance farm security, the platform includes animal and fire detection systems with instant alerts in multiple languages. Furthermore, it offers hands-on simulation training for agricultural drone applications to improve productivity. By bridging the gaps in knowledge, communication, and technology, Karshaka Mithra aims to revolutionize Indian agriculture, fostering sustainable growth and enhancing the livelihoods of farmers nationwide. The involvement of both government and non-government organizations in promoting awareness and usage of this platform can significantly elevate agricultural standards and drive a technology-first approach to farming.

DOI: /10.61463/ijset.vol.13.issue2.195

Intellicheck – Plagiarism Detector Using HMM, BERT and Abstract Syntax Tree

Authors- SVS Satish, S D Anirudh, P Chandu, K Suhaas Varma, Assistant Professor P.Jyothi

Abstract-In the modern era of widespread access to information and digital resources, plagiarism detection has become increasingly essential for preserving academic integrity and protecting intellectual property. This research focuses on the design and development of a dual-functional plagiarism detection system capable of analyzing both text and source code submissions. The text-based detection module employs a Hidden Markov Model (HMM) to evaluate textual similarity and utilizes a BERT model to understand and compare the semantic meaning of the content. The source code module leverages Abstract Syntax Trees (AST) to identify structural similarities in programming code, offering precise detection of plagiarized logic and patterns. The proposed system bridges the gap between surface-level content analysis and deeper contextual and structural understanding, making it a robust and comprehensive tool for a wide array of applications. This paper also addresses implementation challenges, evaluates system performance metrics such as precision and recall, and suggests potential future enhancements, including the incorporation of multilingual support and the expansion to more programming languages. By integrating advanced technologies, this system provides a reliable solution for detecting plagiarism in academic, research, and software development environments, ensuring fairness and originality.

DOI: /10.61463/ijset.vol.13.issue2.196

IOT – Based Fall Detection System

Authors- Associate Professor Y V Nagesh Meesala, Perumalla Harini, Buddala Sravani, Jagupilla Ganesh, Muchu Madhan Mohan

Abstract-This paper presents an IoT-enabled system for fall detection and epilepsy monitoring, integrating accelerometers, Arduino Uno, and NodeMCU to ensure real-time health tracking. The system detects falls and seizures through movement pattern analysis, triggering instant alerts via the Blynk IoT platform. It provides an LCD display for status updates and a buzzer for immediate user alerts. By combining fall detection and seizure monitoring in a single system, this approach enhances safety, enables timely intervention, and simplifies care management for individuals at risk, demonstrating the transformative potential of IoT in healthcare monitoring.

DOI: /10.61463/ijset.vol.13.issue2.197

An Experimental Investigation on Application of Biopolymer in Concrete

Authors- M P Iniya, R Mahalakshmi, S Swathi

Abstract-Biopolymers are being investigated as an environmentally friendly substitute in the manufacturing of concrete due to the growing need for sustainable building materials. This study investigates the use of xanthan gum and cellulose derived from bagasse ash as biopolymer additions in concrete. A byproduct of sugarcane production, bagasse ash is high in cellulose and provides an affordable, eco-friendly substitute for conventional concrete additives. Amicrobial polysaccharide called xanthan gum is known to improve cementitious mixes’ viscosity and workability. These ingredients include cellulose, xanthan gum, fine and coarse aggregate, and cement. The cube specimen measures 150 x 150 x 150 mm, while the cylinder specimen measures 150 mm in diameter and 300 mm in height. Experiments are conducted to examine differences in properties including compressive strength, split tensilestrength and slump cone test. This study highlights the potential of cellulose and xanthan gum as eco-friendly alternatives to conventional chemical additives in concrete, providing a pathway for more sustainable construction practices.

DOI: /10.61463/ijset.vol.13.issue2.198

Students’ Final Year Projects Record Management System (Sfyprms)

Authors- Johnson Tunde Fakoya, Mobolaji Olusola Olarinde

Abstract-Efficient management of students’ final year projects is critical to ensuring the seamless operation of academic programs in higher education. Traditionally, manual processes have been used for recording and managing these projects, leading to inefficiencies such as delays, errors in data handling, and challenges in information retrieval. This study explores the design and implementation of a Students’ Final Year Projects Record Management System (SFYPRMS) to address these challenges. The proposed system leverages modern database technologies and user-friendly interfaces to automate key processes, including project registration, supervisor allocation, progress tracking, and archival. By digitizing these workflows, the system enhances data accuracy, ensures secure storage, and improves accessibility for stakeholders such as students, supervisors, and administrators. Key features include real-time project status updates, automated notifications, and a robust search function for retrieving archived projects. A prototype of the system was developed and tested in a university setting, demonstrating significant improvements in efficiency, transparency, and user satisfaction compared to traditional manual methods. The findings suggest that implementing a digital record management system for final year projects can streamline administrative tasks, support compliance with institutional policies, and improve overall academic performance. This paper highlights the potential of integrating technology into project management processes in higher education, providing a scalable solution that addresses the growing complexity and volume of academic records. Future research may focus on extending the system’s capabilities, such as incorporating AI for predictive analytics and enhancing system interoperability with existing institutional platforms.

DOI: /10.61463/ijset.vol.13.issue2.199

Foot Check: Advance Detection of Diabetic Ulcer Using AI-Driven Image and Text Processing

Authors- Preethiga D, Sathya D, Sathasri M, Mr.S.Baskar

Abstract-Diabetic ulcers are a serious complication of diabetes that can lead to infections and amputations if not detected and treated early. Current diagnostic methods rely heavily on manual evaluation, which is subjective and error-prone. This project proposes an AI-based system for the early diagnosis and classification of diabetic ulcers using deep learning for image analysis and Natural Language Processing (NLP) for clinical text evaluation. Convolutional Neural Networks (CNNs) will classify ulcer images by severity, while Transformer-based NLP models will analyze clinical notes and patient histories for contextual insights. By integrating these modalities, the system aims to enhance diagnostic accuracy and provide healthcare professionals with a reliable tool for timely intervention. This AI system not only automates ulcer detection but also supports decision-making, reducing the burden on medical staff and improving patient outcomes. It offers a comprehensive understanding of a patient’s condition, facilitating personalized treatment and remote monitoring, especially for those in underserved areas. Ultimately, this innovative approach seeks to transform diabetic ulcer care, improving healthcare efficiency and accessibility.

DOI: /10.61463/ijset.vol.13.issue2.200

Micro-Architect on RISC-VISA

Authors- Dhanush S, Deepan R, Chandru M, Geethalakshmi M

Abstract-The major problem with the ISA architecture is its power consumption ratio for the ALU operation .It has a MUX Combinational circuit for their memory access purpose . Comparing with other logic gates mux have additional logic gates which is needed for the purpose of path selection leading to consume more power . As an alternative we use more multipliers but the X factor want to consider was the execution time must be very quick and efficient to my signed numbers . Our approach uses Baugh – Wooley algorithm which over comes all the drawbacks of other multiplier and has nearly of 24.32% power reduction .This can be measured in two ways as the time consumption can also be calculated by the RTL design and the power efficiency can be calculated by the synthesis method where can generate power reports.

DOI: /10.61463/ijset.vol.13.issue2.201

Cryptographic Encryption for Secure Data

Authors- Preethi K, Preethiga K, Shanmugapriya S, V.Krishnakumar

Abstract-In order to safeguard sensitive data, this project will integrate steganography with sophisticated cryptographic encryption to improve the security of digital communications. Ensuring data transmission security and integrity has become essential in an era of growing cyber threats. By allowing users to encrypt and embed messages into a variety of multimedia assets, such as text documents, audio, video, and photos, this technology ensures both confidentiality and covert communication. Strong security is achieved by implementing steganographic techniques like LSB and F5 to ensure that concealed messages stay undetected and high-level encryption algorithms like AES-256 and RSA to protect data integrity. Instead of alerting prospective attackers, this method enables safe data transport, unlike traditional encryption, which might cause suspicion. The system features a user-friendly interface designed for seamless and efficient secure communication. It also incorporates multi-factor authentication and role-based access control to prevent unauthorized access. Furthermore, advanced anti-steganalysis techniques are employed to counteract modern detection methods, ensuring that hidden data remains secure against forensic analysis. This comprehensive solution addresses evolving cybersecurity challenges, making it suitable for personal, corporate, and government applications. By combining encryption with steganography, this project ensures secure, covert, and reliable data transmission and storage, providing an innovative approach to digital privacy and protection against cyber threats.

DOI: /10.61463/ijset.vol.13.issue2.202

Enhancing Aviation Safety through Secure Controller-Pilot Data Link Communication Protocols

Authors- Subalakshmi K, Suvedha S, Vanitha D, K. Yazhini

Abstract-Millions of passengers depend on safe and effective air travel every day, making aviation safety a top priority in the global transportation sector. By reducing reliance on voice communication, which is prone to mistakes and misunderstandings, Controller-Pilot Data Link Communication (CPDLC) protocols have become a game-changing technology to improve safety and operational efficiency. Pilots and air traffic controllers can communicate directly digitally thanks to CPDLC, which makes it possible to send precise, succinct, and unambiguous instructions. This paper explores the critical role of secure CPDLC protocols in mitigating risks associated with traditional voice-based systems, such as frequency congestion and language barriers. By integrating advanced encryption techniques and robust authentication mechanisms, secure CPDLC ensures data integrity and confidentiality, safeguarding against cyber threats and unauthorized access. The study also looks at how CPDLC affects situational awareness, how much less work controllers and pilots have to do, and how it helps keep air traffic operations in complicated airspaces running smoothly. This study emphasizes innovations and best practices in creating and implementing secure communication systems through case studies and analysis. This study highlights the significance of ongoing adaptation and development, as well as the potential of CPDLC to transform aviation efficiency and safety in a world growing more interconnected by the day.

DOI: /10.61463/ijset.vol.13.issue2.203

Hybrid Infrastructure for Enhanced Link Defense Using Quantum Security and Dynamic Channel Hopping

Authors- Sowmiya G, Sophia R, Sindhuja R, T.Dinesh kumar

Abstract-Wireless blockchain networks face increasingly sophisticated jamming attacks that exploit RF vulnerabilities, manifesting as targeted interference, signal manipulation, and consensus disruption attacks. Traditional defense mechanisms, primarily relying on frequency hopping spread spectrum and basic cryptographic protocols, prove insufficient against advanced cognitive jammers and quantum-capable adversaries The framework implements a hybrid consensus protocol that synergizes Proof of Stake with practical Byzantine Fault Tolerance .The system employs a deep learning-based anomaly detection model utilizing a Long Short-Term Memory network architecture .The network activates quantum-aided dynamic channel hopping, leveraging quantum key distribution for secure channel state information exchange and pseudo-random sequence generation .For enhanced scalability, implementation of a novel blockchain sharding mechanism that dynamically partitions the network .The integration of quantum communication principles, specifically BB84 protocol for key distribution and quantum Byzantine agreement, provides information- theoretic security with a key. Signal integrity is maintained through adaptive Rayleigh fading compensation using Maximum Likelihood Estimation and Low-Density Parity Check codes with minimum Hamming distance , ensuring robust error correction in compromised channels. The system exhibits a packet delivery ratio under severe jamming conditions, with quantum bit error rate .Network throughput remains stable of optimal capacity during multi-channel jamming attacks, while the sharding mechanism demonstrates linear scalability .The comprehensive solution establishes new benchmarks for secure wireless blockchain deployments, effectively addressing both current and emerging quantum-era security challenges.

DOI: /10.61463/ijset.vol.13.issue2.204

A Matlab Simulink Modeling of Solar Based EV System with Control of its Utility Parameters

Authors- Ajay Yadav, Assistant Professor Abhay Awasthi

Abstract-The transition towards sustainable transportation solutions has led to an increased demand for electric vehicles (EVs). One of the critical challenges in EV adoption is the availability of efficient and fast-charging infrastructure. This paper presents the design and simulation of a solar-based fast charging station for electric vehicles using MATLAB. The proposed system integrates solar photovoltaic (PV) panels, power electronics, energy storage, and charging management techniques to provide a reliable and sustainable solution. The design process involves selecting appropriate PV panel configurations and sizing them to generate the required power for fast charging. To ensure continuous power availability, an energy storage system, such as batteries, is integrated into the system. The power electronics components, including DC-DC converters and inverters, are designed to efficiently manage power flow between the PV panels, energy storage, and the EV charging units. Advanced control strategies are implemented to regulate the charging process, considering factors like battery state of charge, EV battery specifications, and grid interactions.

A Survey on Digital Consumer Behavior based Product Recommendation Techniques

Authors- Yogini Sarathe, Professor Rahul Patidar, Professor Jayshree Boaddh

Abstract-As the number of internet users are increasing day by day. This tends for most of the market to move towards online shopping. Some of website provide user rating for different product but they do not recommend any user to purchase. So bridging the gap between user social behavior and product requirement is done. In this paper a brief survey of product prediction algorithm is done. Paper has discuss different challenges that need to be address for researchers. Various research were detailed for the different recommendation systems. Different models were list in the paper for recommendation of product and services. In order to compare models normal evaluation parameters were also elaborated.

Enhancing Concrete Durability through Partial Replacement of Fine Aggregate with Ceramic and Granite Waste

Authors- Dhavashankaran D, Dinesh V, Nishanth K, Varun S

Abstract-Waste from development and demolition is growing daily, and natural resources are running out. Many nations governments and researchers are attempting to determine the best way to handle this predicament. Ceramic and Granite debris from the building sector needs to be used efficiently. Numerous researchers have discovered that it may be used to make concrete by replacing some or all of the finse particles. Finding appropriate substitutes that can partially or significantly replace sand is urgently needed due to the steadily rising demand for river sand and the dwindling supply. The experimental investigation of the mechanical strength characteristics of M25 grade concrete using granite and ceramic waste in place of some of the sand is the focus of this project. The samples were cast with 10%, 20%, 30%, 40%, and 100% replacement of sand using ceramic and granite waste in order to study the mechanical qualities like compressive and split tensile strength. They were then tested for varying curing times, such as seven, fourteen, and twenty-eight days. Byproducts of the manufacturing and construction sectors, ceramic and granite waste are frequently dumped in landfills, contributing to environmental contamination. These materials are good substitutes for conventional fine aggregates in concrete because of their superior mechanical qualities, which include high hardness, endurance, and resistance to chemical attacks.

DOI: /10.61463/ijset.vol.13.issue2.205

Experimental Investigation on Pervious Concrete

Authors- B. Sasivarman, S. Ajithkumar, T. Kamalesh, R. Loganathan

Abstract-Pervious concrete made from uniform graded material consisting of Portland Pozzolana Cement aggregate, Admixtures and Water. Because pervious concrete contains no fine aggregates such as sand, it is sometimes referred to as “no fines” concrete. It is a special type of concrete having a high void content of about 30%, and is becoming popular nowadays due to its potential of reduce water runoff of the drainage systems which can provide a water flow around 0.34cm/second. In this project, detailed various literature has been under taken to fully understand the properties and applications of pervious concrete. An alternate has been made to develop a mix of pervious concrete and compare with that of the conventional concrete. The strength properties of hardened concrete include compressive strength of cube, split tensile strength of cylinder, flexural strength of prism, load deflection behaviour of beam for various mix proportions have been studied. Based on the results of the experimental study, some important conclusions have been drawn.

DOI: /10.61463/ijset.vol.13.issue2.206

Practical Methods to Control Conducted and Radiated Emissions in DC/DC Converter Based Design

Authors- Akshay Khot, Sanket Bhor, Anand Kendre

Abstract-Electromagnetic interference (EMI) is one of the biggest challenges faced during the production of any electronic device. If the EMI profile of the system does not meet the accepted standards, then it becomes necessary to take measures to reduce the influence of these unwanted interferences so that the equipment can be used in the real world. Practical methods to control conducted and radiated emissions in DC/DC converters involve a combination of design techniques, component selection, and filtering strategies. One key approach is to identify and mitigate noise sources within the converter, such as fast voltage and current transitions, and to optimize the converter’s layout and grounding to minimize radiation. Filter design and optimization involve selecting appropriate components, such as inductors, capacitors, and ferrite beads, and configuring them to achieve the desired attenuation and impedance characteristics. Parasitic components, such as series inductance and series resistance, can have a significant impact on EMI and should be considered in simulation and design. By combining these practical methods, this paper gives designers insight on how we can effectively, control conducted and radiated emissions in DC/DC converter based designs and ensure compliance with relevant EMI standards.

DOI: /10.61463/ijset.vol.13.issue2.207

Predictive Telemedicine Model for Early Heart Attack Detection in Dialysis Patients Using Multimodal Data and Machine Learning

Authors- Assistant Professor Mr J.P Pramod, Boddupally Pravalika, Gyajangi Nandini

Abstract-The dialysis patients are prone to developing heart attacks, as fluctuations in blood pressure and electrolytes occur very often. Such an early diagnosis is extremely important to ensure timely interventions. The current work presents a predictive telemedicine model for continuous monitoring of dialysis patients, thus providing them with early warning for a heart attack. Patients on dialysis, more often are liable for heart attacks due to frequent fluctuation of blood pressure, electrolyte balance, and intrinsic medical illness. Early detection of this event is the prerequisite treatment for patients. In light of this need, there is a proposed predictive model aimed at predicting heart attacks or chances of a heart attack in patients who have come for dialysis; remote monitoring and early intervention over telemedicine channels will surely be beneficial. In applying these various machine learning algorithms, logistic regression, random forest, SVM, KNN, Gradient Boosting, and finally XGBoost classifiers have been used. Using relevant clinical features like vital signs, lab findings, and other demographic data in a given database, the models will be trained to identify certain patterns that are correlated to the risk of having an attack. XGBoost had the highest score at 92.6% with excellent precision, recall, and F1-score, which shows that the model was pretty robust in predicting the events.

DOI: /10.61463/ijset.vol.13.issue2.208

Enhancing Concrete Strength and Durability with Bacterial Self-healing Mechanisms

Authors- P. Prabhu, R. Gokulraj, D. Prakash, K. Santhosh kumar

Abstract-Self-healing, also known as bacterial concrete concrete, presents a promising solution to enhance the durability and sustainability of infrastructure. This project aims to explore the mechanisms behind bacterial concrete’s self-healing properties, evaluate its performance under various environmental conditions, and assess its potential for widespread implementation in construction projects. When cracks appear, the bacteria Bacillus Megaterium come into contact with water and nutrients. Concrete is the most widely used construction material, but its susceptibility to cracks significantly reduces its durability and structural performance. To address this challenge, bacterial self-healing mechanisms have emerged as an innovative and sustainable solution for enhancing concrete’s strength and longevity. This method involves incorporating specific types of bacteria, such as Bacillus species, into the concrete mix along with a suitable nutrient source like calcium lactate. When cracks form and moisture infiltrates the concrete, dormant bacterial spores germinate and activate, precipitating calcium carbonate (CaCO₃) crystals that seal the cracks. They then undergo metabolic processes that result in the precipitation of calcium carbonate, a key component of cement. The concrete cubes was tested using 150mm x 150mm x 150mm cubes for compressive strength.

DOI: /10.61463/ijset.vol.13.issue2.209

LipNet: Deep Learning for Visual Speech Recognitions

Authors- Atharva Mahesh Khandagale, Atharva Sunil Bagave, Aditya Santosh Pande, Rishabh Hitesh Jain, Professor Kalaavathi B

Abstract-LipNet is a deep learning system that completely reimagines the approach towards visual speech recognition. It makes predictions over whole sequences of words by forcing a watch of lip movements in videos to a single process. Unlike previous techniques that rely on hand-crafted feature extraction and processing in separate stages, LipNet models this in one end- to-end process. It uses spatiotemporal Convolutional Neural Networks to capture visual features and RNNs with LSTM units for handling sequences. A salient feature of LipNet is the use of Connectionist Temporal Classification loss, which will enable it to learn directly from unsegmented data. Tested on various challenging datasets like GRID, LipNet has set a new standard in automated lip-reading accuracy. Its streamlined design and impressive performance open up exciting possibilities in areas like accessibility, silent communication, and security, making it a major step forward in this field.

DOI: /10.61463/ijset.vol.13.issue2.210

Experimental Investigation on Floating Concrete Using Light Weight Materials

Authors- A.karthick, K.Gurumoorthi, S.Harivignesh, M.Ragul

Abstract-This project mainly deals the concrete to float in water by using the combination of light weight aggregate and Air entraining agent of Aluminium powder. The main idea of our project is that the light weight aggregate in the concrete lowers the concrete’s self-weight, so that concrete’s density also reduces. Hence its density is less than 1000kg/m3, the concrete floats in water. For making this type of light weight concrete, we don’t know the proportion to be mixed in concrete, So that we have adopted the trial and error method to solve this problem. The density of concrete comes under the limit of 500 to 900 kg/m3 . The proper mix design of the floating concrete is not arrived, so we have taken the mix ratio from the help of journals. We have used various materials like Fly-ash, gypsum, Lime powder, Pumice stone, Aluminium powder, Polypropylene, GGBS, Vermiculate, Sand, Cement etc. we have successfully achieved the floating property of the concrete from the combination of below ingredients Cement, Lime powder, Gypsum, Fly-ash, Aluminium powder, Polypropylene, Sand . We have partially replacing the cement by Fly-ash (48%), Lime(17%), Gypsum (6%), and then fine aggregate of the sand is replaced by 50% of polypropylene. We have added air entraining agent of Aluminium powder by 2-10%. Finally the floating property and the compressive strength of the concrete are tested.

DOI: /10.61463/ijset.vol.13.issue2.211

Experimental Investigation of Concrete with Recycled Aggregate, GGBS and Steel Fibers

Authors- Gowtham S, Madhan raj S, Ramki R, Vishnu K

Abstract-As construction is increasing day by day in last decade, it is concerned for processing of waste material with the help of recent technology. Ground Granulated Blast Furnace Slag (GGBS), and steel fibers, aiming to evaluate their combined effects on mechanical properties, durability, and microstructural characteristics. The current study presents the experimental results of Recycled aggregate (RA) with different percentage 0, 20, 40, 60, 80, 100% in concrete replaced by natural coarse aggregate. Casting of cube, beam, cylinder mould has been done. In addition to this replacement of cement partially with MK also has been introduced. The flexural strength, compressive tensile strength and split tensile strength of M-35 and M-25 have been determined and also there design mix has been prepared. The results show good effect up to 60% aggregate replacement but on further increases the replacement ratio of aggregate, strength decreases. Therefore sustainable concrete may be produced to make environment safe and eco-friendly.

DOI: /10.61463/ijset.vol.13.issue2.212

The Journey of Employees in an Organization: An Exploration on Employee Promotion & Organization Elevation

Authors- Dr. Sukanta Mishra

Abstract-In most third world countries of the world, people always consider not from their brains but stomach, because of the higher rate of unemployment and dwindling nature of the economy, coupled with the over reliance on government for employment, the major source of job satisfaction in the third world countries is promotion. Today, employees are considered as an asset for the organization and the success of an organization largely depends on the people, as it is not possible for survival without them. The growth of an employee to higher ranks is Employee Promotion. This is a powerful strategy for motivating employees. Employee promotion serves as a crucial milestone, symbolizing growth, recognition, and new opportunities. It is a moment when hard work and dedication are acknowledged, and individuals are entrusted with increased responsibilities. Promotion and Competence have direct impact on Job Satisfaction and Performance. More successful job promotion and a high degree of competence improve workers’ satisfaction, which affect the improved output of employees. Promotion is not just a mere change in job title; it carries significant implications for both employees and organizations alike. It involves an increase in salary, position, responsibilities, status, and benefits. This aspect of the job drives employees the most—the ultimate reward for dedication and loyalty towards an organization.

DOI: /10.61463/ijset.vol.13.issue2.213

Research on the Influence of Flight Altitude on the Aerodynamic Characteristics of Su-27 Fighter Aircraft

Authors- Son Nguyen, Truong Thanh Nguyen, Khuat Van Huy

Abstract-The article focuses on studying the influence of flight altitude on the aerodynamic characteristics of the Su-27 fighter. The research was conducted on the basis of applying SolidWorks software to build a 3D model of the aircraft, combined with the Fluent module of Ansys software to simulate and calculate the changes in aircraft characteristics such as drag coefficient, lift coefficient and longitudinal moment coefficient. From there, some recommendations are given to users to improve flight performance and enhance flight safety when the aircraft changes flight altitude.

DOI: /10.61463/ijset.vol.13.issue2.214

Automatic Number Plate Recognition System

Authors- Dipali Ambadkar, Sanika Ghyar, Isha Dhande, Akansha Sawale, Dr. K.N Kasat

Abstract-With increasing urbanization, the demand for efficient parking management systems has risen significantly. Traditional methods relying on manual intervention are prone to human errors, security risks, and inefficiencies. This paper proposes an automated paid parking management system utilizing number plate recognition (NPR) technology to streamline vehicle identification, entry-exit monitoring, and digital payment processing. The system integrates machine learning, image processing, and web-based technologies to enhance security and optimize space utilization. Experimental results indicate a 97% accuracy rate in standard lighting conditions, with real-time transaction processing under three seconds.

DOI: /10.61463/ijset.vol.13.issue2.215

Tripura’s Natural Beauty and Biodiversity: A Mini Amazon of the Northeast

Authors- Research Scholar, Mr. Nimai Sarkar

Abstract-Tripura, a picturesque state in north-eastern India, is renowned for its lush green landscapes, diverse wildlife, and rich biodiversity, earning it the title of a “Mini Amazon” of the region. Encompassing dense forests, vibrant river ecosystems, and unique flora and fauna, Tripura plays a crucial role in maintaining ecological balance. This paper explores the state’s natural beauty, highlighting its protected areas, endemic species, and the symbiotic relationship between its indigenous communities and nature. Additionally, it examines the challenges posed by deforestation, climate change, and urbanization, while discussing conservation efforts and sustainable development initiatives. By emphasizing Tripura’s ecological significance, the study underscores the need for effective environmental policies and community-driven conservation strategies to safeguard its rich natural heritage for future generations.

DOI: /10.61463/ijset.vol.13.issue2.216

Applications of Inorganic Catalysts in Lignocellulosic Biomass Valorization

Authors- Olufunso O. Abosede, Kelechi Peace Kanu

Abstract-The valorization of lignocellulosic biomass is crucial for advancing renewable energy and sustainable chemical production. This review explores the role of inorganic chemistry in overcoming biomass recalcitrance through the applications of inorganic compounds which enhance pretreatment, depolymerization, and catalytic conversion. Recent advances in transition metal complexes (Fe, Co, Cu) and metal-organic frameworks (MOFs) are analyzed for their efficiency in lignin breakdown and biofuel synthesis. Beyond biomass, inorganic compounds also contribute to solar energy conversion, hydrogen production, and energy storage. The review highlights emerging trends, challenges, and future directions, emphasizing the need for cost-effective catalysts and integrated bio refinery strategies to improve sustainability.

DOI: /10.61463/ijset.vol.13.issue2.217

Artificial Intelligence Can Help Improve Education

Authors- Dhruvi Kheni, Shreya Sonani, Professor Bhoomika B. Chauhan

Abstract-Artificial Intelligence (AI) is considered important in education research and practice in many ways. With AI-powered tools, we can improve data analysis, provide personalized learning, facilitate the creation of content, and offer assistance to teachers. Along with this, the ethical controversies of the AI implementations in education and the likely futures of how this might affect education systems across the globe are further discussed in the paper.

DOI: /10.61463/ijset.vol.13.issue2.218

Effective Usage of Silica Fume Ash as the Replacement of Cement in Concrete

Authors- K.Pradeep, R.Dhivakar, K.Tharwin, A.Balakumaran

Abstract-The construction industry continuously seeks sustainable alternatives to traditional materials to reduce environmental impact and enhance performance. One such alternative is silica fume ash, a byproduct of silicon and ferrosilicon alloy production, which has gained attention as a partial replacement for cement in concrete. This study explores the effective utilization of silica fume ash in concrete mixtures, focusing on its impact on strength, durability, and workability. The high pozzolanic activity of silica fume ash enhances the compressive strength of concrete by refining the microstructure and reducing porosity. Additionally, its incorporation improves durability properties, such as resistance to sulfate attack, chloride penetration, and alkali-silica reaction. However, proper mix design and curing methods are crucial to achieving optimal performance, as excessive replacement levels may adversely affect workability. This research evaluates various replacement ratios of silica fume ash in cement and presents experimental findings on the mechanical and durability properties of modified concrete. The results highlight its potential in producing eco-friendly, high-performance concrete, contributing to sustainable construction practices. The increasing demand for sustainable and high-performance concrete has led to the exploration of supplementary cementitious materials. Silica fume, a byproduct of silicon and ferrosilicon alloy production, has gained attention as a viable replacement for cement due to its pozzolanic properties. This study investigates the effective utilization of silica fume ash as a partial substitute for cement in concrete, aiming to enhance mechanical strength, durability, and environmental sustainability.

DOI: /10.61463/ijset.vol.13.issue2.219

A Comprehensive Review on Optimization and Process Parameters in Viscosity of Nanofluid by Using RSM Method

Authors- S.Baskar, S.Ramasubramanian, L.Karikalan, M.Ruban, Madhan Kumar G

Abstract-During the past decade, nanotechnology with its rapid development has grabbed the attention of scientists, scholars, and engineers. Nanofluids are one of the surprising outcomes of this technology that could increase the efficiency of thermal systems remarkably. Nanofluids containing solid nanoparticles have a higher viscosity than common working fluids; hence, measuring the viscosity is necessary for designing thermal systems and estimating the required pumping power. In the current review study, an attempt has been made to cover the latest experimental studies performed on the viscosity of nanofluids. An experimental investigation is very vital for the analysis since the theoretical models usually underestimate the nanofluid viscosity. Through experiments, the real effects of volume fraction, temperature, particle size, and shape on the viscosity of nanofluids will be determined.

Fabrication of IoT-Based Solar Water Pumping System

Authors- S. Ajith Arul Daniel, Vijay Ananth Suyamburajan, S. Sivaganesan, C. Gnanavel, T. Gopalakrishnan

Abstract-This study presents the design and fabrication of an IoT-based solar water pumping system, aimed at addressing the challenges of traditional water pumping methods in rural and agricultural settings. Traditional AC motor-driven pumps and fuel-operated pumps are often inefficient, costly, and environmentally harmful. The proposed system utilizes solar energy as a renewable power source, integrated with IoT technology for remote monitoring and control. The system comprises a solar panel, battery, DC pump, and IoT module, enabling users to monitor water levels and control the pump via a web interface. The system is cost-effective, easy to maintain, and environmentally friendly, making it suitable for applications in agriculture, home appliances, and commercial purposes. The total cost of the system is estimated at Rs. 12,000, with a detailed breakdown of material and labor costs provided.

The Impact of Smart Cricket Bat Grips on Batting Efficiency and Control

Authors- Dr.B.Prabakar

Abstract-The evolution of cricket equipment has significantly influenced player performance, with advancements in bat technology playing a crucial role. One such innovation is the smart cricket bat grip, designed to enhance batting efficiency and control. This study examines the impact of smart grips that incorporate adaptive material technology, pressure sensors, and sweat-resistant properties to optimize grip and reduce slippage. The research evaluates improvements in shot precision, hand fatigue reduction, and overall batting stability. The findings indicate that smart grips contribute to better control, improved shot execution, and enhanced player comfort, making them a potential game-changer in modern cricket. A statistical analysis was conducted to assess the impact of smart grips compared to conventional grips, with results demonstrating a significant improvement in performance metrics.

DOI: /10.61463/ijset.vol.13.issue2.220

Experimental Investigation on Performance and Emission Characteristics of CI Engine Fuelled with Canola Oil Biodiesel Blends

Authors- Vijay Ananth Suyamburajan, S. Sivaganesan, C. Gnanavel, T. Gopalakrishnan, S. Ajith Arul Daniel

Abstract-The increasing demand for energy and the environmental concerns associated with conventional fossil fuels have led to the exploration of alternative fuels. Biodiesel, derived from vegetable oils, has emerged as a promising alternative due to its renewable nature and lower emissions. This study investigates the performance and emission characteristics of a compression ignition (CI) engine fuelled with canola oil biodiesel blends. The biodiesel was produced through the transesterification process, and blends of B10, B20, and B30 were tested in a single-cylinder diesel engine. The results indicate that canola oil biodiesel blends exhibit comparable performance to conventional diesel, with a slight increase in brake thermal efficiency and a reduction in specific fuel consumption. Emission analysis revealed a significant reduction in carbon monoxide (CO), carbon dioxide (CO₂), and hydrocarbon (HC) emissions, although nitrogen oxide (NOx) emissions were slightly higher. The study concludes that canola oil biodiesel, particularly the B30 blend, is a viable alternative fuel for diesel engines, offering both environmental and performance benefits.

Experimental Study on a DI Diesel Engine Fueled with Mahua Oil Methyl Ester and Diesel Blends

Authors- S. Sivaganesan, C. Gnanavel, T. Gopalakrishnan, S. Ajith Arul Daniel, Vijay Ananth Suyamburajan

Abstract-With the rising number of internal combustion engine-based vehicles, energy demand is also increasing. The growing consumption of petroleum products and environmental pollution remains a significant global concern. Since fossil fuel availability is highly limited, the need for alternative fuels is becoming increasingly critical to ensuring energy security and environmental protection. In recent years, biodiesel has been produced from both edible and non-edible vegetable oils due to its renewable nature and lower emissions. In this study, various blend ratios of B20, B40, B60, B80, and B100 of MEOM (methyl ester of mahua) with diesel are analyzed in a single-cylinder DI diesel engine at 1500 rpm. The findings indicate that the B20 MEOM blend exhibits lower emissions and improved performance characteristics.

A Review on the Comparative Modal Analysis of Structural Beams Using CATIA

Authors- Research Scholar Nakul Bharti, Assistant Professor Neeti Soni, Assistant Professor Dr. Rajesh Rathore, Assistant Professor Praveen Patidar

Abstract-The study of beam dynamics is fundamental to structural engineering, influencing the design and application of beams in buildings, bridges, and mechanical components. This review examines the methodologies and applications discussed in the thesis “Comparative Modal Analysis of Beams Using CATIA,” focusing on modal analysis principles, material optimization, and computational approaches. The document serves as a guide to understanding the dynamic behaviors of different materials and geometrical configurations in beams, leveraging advanced simulation techniques.

CIFAKE: Precision Image Classification and Explainable AI for Detecting Synthetic Generations

Authors- D. Chakra Satya Tulasi, Singuluri Annapurna, Pasumarthi Samyuktha, Busi Srikanth, Golla Sneha Ratna, Gampala Lakshmi Shiva Teja

Abstract-The rapid advancements in synthetic data generation have resulted in AI-generated images that are nearly indistinguishable from real photographs, posing challenges in data authenticity and reliability. This study aims to enhance the detection of AI-generated images using computer vision techniques. A synthetic dataset resembling the ten classes of CIFAR-10 is created using latent diffusion, providing a direct comparison between real and AI-generated images. The classification task is framed as a binary problem, distinguishing real images from synthetic ones. To achieve this, a Convolutional Neural Network (CNN) is trained to classify the images with optimal performance. After fine-tuning hyperparameters and evaluating 36 distinct network architectures, the best-performing model achieves an accuracy of 92.98%. Additionally, explainable AI techniques, such as Gradient Class Activation Mapping, are applied to interpret the model’s decision-making process. Interestingly, the analysis reveals that rather than focusing on primary subjects, the model relies on subtle background inconsistencies to differentiate real images from synthetic ones. To support further research in this domain, the newly generated dataset, CIFAKE, is made publicly available for future studies.

DOI: /10.61463/ijset.vol.13.issue2.221

NeuroShield: AI-Powered Brain Stroke Detection Using Advanced Machine Learning Algorithms

Authors- A. Daiva Krupa Nirmala, Pedireddy Harika, Singarapu Aneela Deepthi, Bulusu V S L N Bhaskara Teja, Bonam Ajay Saatvik, Pavan Puppala

Abstract-Stroke remains the second leading cause of death worldwide, underscoring the necessity for timely and precise predictive models to support early intervention. This research investigates advanced machine learning techniques to enhance stroke prediction accuracy. Initially, traditional classifiers such as Decision Tree (DT), Random Forest (RF), and Support Vector Machine (SVM) were employed. To further strengthen predictive performance, Extreme Gradient Boosting (XGBoost) and Light Gradient Boosting Machine (LGBM) were later introduced. Model performance was assessed using various evaluation metrics, including accuracy, sensitivity, error rates, and log loss. Results indicate that XGBoost achieved an outstanding accuracy of 98%, while LGBM also played a crucial role in boosting overall predictive accuracy. These findings highlight the significant impact of sophisticated machine learning models in stroke prediction. By integrating state-of-the-art predictive analytics into clinical settings, this study aims to facilitate faster and more accurate diagnoses, ultimately improving patient care and advancing stroke detection methodologies.

DOI: /10.61463/ijset.vol.13.issue2.222

SkyFare AI: A Generative AI-Powered Web Platform for Real-Time Flight Price Prediction

Authors- P. V Komali, Pentapati Prasanthi, Saride Swetha Rama Veni, Ayinala Bhargavi Sowmya, Asritha Goppu, Bhula Vivek

Abstract-The dynamic nature of the aviation industry presents significant challenges in accurately predicting flight fares, as factors like fluctuating demand, fuel price shifts, and route complexities contribute to pricing unpredictability. This research introduces a novel approach that harnesses generative artificial intelligence (GAI) to improve real-time airfare forecasting. By incorporating generative models, deep learning techniques, and historical fare trends, the system enhances prediction accuracy. A GAI-driven web-based framework enables the model to analyse intricate patterns and interdependencies within historical airline data, allowing it to detect complex relationships and adjust to evolving market conditions. Using deep neural networks, the model processes multiple influencing factors, extracting crucial insights for a deeper understanding of airfare variations. Furthermore, the proposed approach emphasizes real-time forecasting precision, facilitating swift responses to market fluctuations and serving as an effective tool for optimizing airfare pricing strategies.

DOI: /10.61463/ijset.vol.13.issue2.223

Securing Android Apps: Deep Learning-Powered Threat Detection from Tweet Analysis

Authors- Ch. Veera Gayathri, Addala Bhargavi, Korada Jayanthkumar, Pinnam John Suresh, Oleti Srinivas Reddy, Telu Dheera Guru Chakravarthi

Abstract-Smartphones have become an essential part of modern life, making security and privacy critical concerns, particularly for the Android operating system, which dominates the smartphone market. However, its widespread use also makes it a prime target for malware attacks, posing significant risks, especially to Internet of Things (IoT) devices that rely on Android applications. To mitigate these threats, this paper proposes a multi-layered malware detection approach that integrates deep learning techniques with real-time threat intelligence from Twitter. By updating a malware hash database every 48 hours using Twitter data, our system remains up to date with emerging threats. Additionally, we employ a Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM) units to analyse application permissions, achieving a 94% detection accuracy. This comprehensive approach strengthens Android security by delivering a proactive and adaptive malware detection system, effectively countering evolving cyber threats.

DOI: /10.61463/ijset.vol.13.issue2.224

AI-Powered Global Solar Radiation Forecasting: Fusing Machine Learning with Satellite Vision

Authors- K. Harika, Telugunta Sahana, Surampudi Yamini Satya Sai Priya, Jafari Kulsum, Thadisetti Siva Venkata Sai Kumar, Shaik Sameeruddin

Abstract-Accurately predicting Daily Global Solar Radiation (DGSR) is crucial for applications in renewable energy, agriculture, and climate research. This study explores the potential of Machine Learning (ML) algorithms combined with satellite imagery to enhance DGSR forecasting. Traditional ML-based approaches rely on meteorological parameters such as temperature, wind speed, atmospheric pressure, and sunshine duration, along with radiometric factors like aerosol optical thickness and water vapor. In this work, we investigate the integration of normalized reflectance from satellite images across multiple spectral channels to refine solar radiation predictions. Two ML models, Artificial Neural Networks (ANN) and Support Vector Machines (SVM), were employed as regression techniques. The findings highlight that the selection and quantity of input parameters significantly impact the accuracy of DGSR predictions. Furthermore, ANN demonstrated superior performance over SVM, achieving an RMSE of 212.21 W h/m², NRMSE of 3.46%, MAPE of 2.85%, MBE of -7.26 W h/m², and an R-value of 0.99. In contrast, the SVM model produced an RMSE of 441.95 W h/m², NRMSE of 6.6%, MAPE of 5.62%, MBE of 69.46 W h/m², and an R-value of 0.98. These results underscore the effectiveness of leveraging satellite imagery alongside ML techniques to improve the accuracy of DGSR forecasting.

DOI: /10.61463/ijset.vol.13.issue2.225

Real-Time Disaster Forecasting: Harnessing Social Media and NLP for Early Crisis Detection

Authors- K. Srikanth, Yarra Lalitha Dabi Varshini, Rayudu Satish, Yarra Koushik, Kondisetti Phaneendra Saketh, Pulidindi Harsha Vardhan

Abstract-Natural disasters such as wildfires, earthquakes, floods, cyclones, and heatwaves have significantly impacted social media users, who actively express their sentiments about these crises online. Analysing location-specific public emotions during such events is essential for policymakers and emergency response teams to make informed decisions. To address this need, we introduce a fully automated artificial intelligence (AI) and natural language processing (NLP) framework designed to extract and analyse sentiment trends related to disasters at specific locations. Our proposed system utilizes AI-driven sentiment analysis, named entity recognition (NER), anomaly detection, regression modelling, and the Getis-Ord Gi* algorithm to process multilingual social media content. Supporting 110 languages, our approach facilitates comprehensive sentiment monitoring and disaster intelligence extraction from social media data. The system was implemented and tested on live Twitter data collected from September 28 to October 6, 2021, processing 67,515 entities in 39 languages and identifying 9,727 location-based entities with over 70% confidence. These insights enabled real-time mapping of disaster-affected locations along with sentiment-based disaster intelligence. Performance evaluation demonstrated the system’s effectiveness, achieving an average precision of 0.93, recall of 0.88, and an F1-score of 0.90, with an overall accuracy of 97%, highlighting the reliability of our AI-driven disaster monitoring framework.

DOI: /10.61463/ijset.vol.13.issue2.226

AI-Driven Drug Discovery and Precision Medicine: Revolutionizing Healthcare with Machine Learning

Authors- N.V.S. Sowjanya, Reddy Hema, Dara Twinkle Roy, Gutti Devi Veera Prasanna, Chikkala William Naveen, Bejawada Sai Pavan Rama Krishna

Abstract-Machine Learning (ML) has emerged as a transformative force in drug discovery and personalized medicine, significantly improving efficiency, accuracy, and cost-effectiveness. This study explores the latest advancements in ML applications within these domains, focusing on breakthroughs and challenges. It examines key ML algorithms instrumental in identifying drug candidates, predicting interactions, and understanding disease pathways. Additionally, we propose an innovative framework that integrates advanced ML techniques with large-scale biomedical datasets to enhance drug efficacy predictions and patient-specific treatment responses. A comprehensive literature review outlines major milestones and benchmarks, providing a foundation for our research. Our methodology employs cutting-edge ML models, including deep learning and reinforcement learning, to analyse complex biological data. The effectiveness of our approach is validated through rigorous experimentation on real-world datasets, demonstrating its potential to optimize drug development and enable personalized treatment strategies. The results section provides an in-depth analysis of model performance, supported by statistical tables and graphical representations. Ultimately, this paper highlights the revolutionary impact of ML in reshaping drug discovery and precision medicine while identifying future research opportunities to address existing challenges.

DOI: /10.61463/ijset.vol.13.issue2.227

AI-Driven Business Intelligence for Adaptive Pricing Strategies in E-Commerce

Authors- M. V. Rajesh, Nelluri Sai Jyothi, Pynda Satya Harshitha, Palla Rakesh, V V S R K Kamalesh Janapa, Kola Uday Surya Prakash

Abstract-In the rapidly evolving landscape of e-commerce, businesses must adopt dynamic pricing strategies to maximize revenue and maintain a competitive edge. This study explores the integration of machine learning (ML) and business intelligence (BI) in optimizing dynamic pricing, addressing the limitations of traditional pricing models that struggle to adapt to shifting digital market conditions. Despite the proven benefits of ML in various business applications, its full potential in online pricing remains underexplored, particularly when combined with BI. Existing research lacks a comprehensive understanding of how ML and BI can work synergistically to enhance pricing decisions. To bridge this gap, this study employs the Support Vector Machine (SVM) algorithm, chosen for its effectiveness in handling complex, nonlinear interactions within large datasets. By leveraging BI technologies to gather, process, and analyze critical data, this study establishes a robust framework for real-time pricing decisions. The findings indicate that integrating ML-driven BI systems enhances pricing accuracy and enables businesses to respond swiftly to market fluctuations. The adaptability of the SVM model allows for precise, context-aware pricing decisions, ultimately strengthening a company’s ability to navigate the dynamic e-commerce environment.

DOI: /10.61463/ijset.vol.13.issue2.228

Advance Alternate Fuels for IC Engines

Authors- Associate Professor Dr. Bijivemula Narayana Reddy, Assistant Professor Mr. K. Reddeppa, Mr. S. Assistant Professor Firoz Basha, Assistant Professor Mr. Y. Chandrasekhar Yadav

Abstract-The increasing concerns over environmental pollution, energy security, and the depletion of fossil fuel reserves have driven significant research and development in advanced alternate fuels for internal combustion (IC) engines. These fuels provide a sustainable and cleaner alternative to conventional gasoline and diesel while maintaining compatibility with existing technologies. Advanced fuels such as hydrogen, biodiesel, ethanol, methanol, compressed natural gas (CNG), liquefied petroleum gas (LPG), biogas, dimethyl ether (DME), ammonia, and synthetic e-fuels offer a range of environmental and performance benefits, including reduced greenhouse gas emissions, improved engine efficiency, and lower particulate matter. However, their adoption faces challenges like production costs, fuel storage and distribution, infrastructure limitations, and the need for engine modifications. Innovations such as dual-fuel engines, advanced combustion technologies like HCCI (Homogeneous Charge Compression Ignition) and PCCI (Premixed Charge Compression Ignition), and fuel additives are critical in addressing these barriers. This paper explores the characteristics, advantages, challenges, and applications of advanced alternate fuels for IC engines, emphasizing their potential to decarbonize transportation and ensure a sustainable energy future. By integrating these fuels with evolving engine technologies, the automotive industry can significantly contribute to reducing environmental impact and achieving global sustainability targets.

DOI: /10.61463/ijset.vol.13.issue2.229

Blockchain-Based Platform for Agricultural Supply Chain

Authors- Ms. Jeya Dharshini, Assistant Professor S, Mr.Yuvaraj. V

Abstract-The traditional agricultural marketing system, characterized by multiple intermediaries, often reduces farmers’ profits and inflates consumer prices. These intermediaries, including wholesalers, commission agents, and retailers, not only add costs but also diminish transparency, limiting farmers’ access to real-time market information and fair pricing.. This direct interaction allows farmers to negotiate better prices, schedule transactions efficiently, and monitor market trends in real time. Additionally, the platform can facilitate digital transactions, reducing reliance on cash and making the process more transparent and secure. Farmers benefit from increased profit margins due to the elimination of intermediaries, while consumers enjoy lower prices, as the supply chain is streamlined. The system promotes fairness, ensuring that farmers have better control over their produce, from setting prices to deciding where and when to sell. Moreover, market personnel can access a wider pool of suppliers, increasing competition and ensuring the availability of fresh produce at competitive prices. This innovative solution not only improves the livelihoods of farmers but also enhances market efficiency, making agricultural products more accessible to consumers at fair prices.

DOI: /10.61463/ijset.vol.13.issue2.230

Personalized Mood Melody through Facial Emotion Analysis Research

Authors- Professor Dr. Y. Subba Reddy, Chilamakuru Vara Mounika, Dalal Nafisa, Avula Mamatha, Gittala Sunil Kumar

Abstract-In the modern era of human-computer interaction, leveraging artificial intelligence to enhance personal well-being has gained significant attention. This research explores the innovative concept of generating personalized music based on facial emotion analysis, aiming to improve mental health and emotional regulation1. Music has long been recognized for its profound psychological effects, capable of influencing mood, stress levels, and cognitive functions. By integrating advanced machine learning algorithms with facial recognition technology, this study proposes a system that dynamically adapts music playlists to match and potentially uplift an individual’s emotional state. The methodology involves real-time facial emotion detection using convolutional neural networks (CNNs) that classify emotions such as happiness, sadness, anger, surprise, fear, and neutrality. High-resolution facial images or video streams are analyzed to extract facial landmarks and micro expressions, which are then processed to determine the prevailing emotional state.

DOI: /10.61463/ijset.vol.13.issue2.231

Cloud Based Blood Bank Management System

Authors- Manish Hedau, Aditya Mhasaye, Shreyash Deshmukh, Ayush Rangari, Professor Ashish. G. Kokate

Abstract-Current blood seat systems are manual and fragmented, resulting in data inaccuracies, Inefficient tracking of donors and lack of centralized management. These limitations Make it difficult to access donor stories, monitor blood inventory and answer effectively for emergencies. This report proposes a cloud -based blood bank Management System to optimize operations through a centralized database. THE The system connects blood banks, hospitals and donors, allowing real -time updates, Efficient donor management and improved coordination. It supports automated Registration, safe data handling and quick access to blood availability, minimizing scarcity and ensure better resource management. By integrating cloud technology, GPS and web platforms, the system enhances operational efficiency and Responsibility, creating a network of reliable and cohesive blood donations.

DOI: /10.61463/ijset.vol.13.issue2.232

Best Practices for Design Rules in Vehicle-Level EMI/EMC Compliance

Authors- Sanket Bhor, Anil Joshi

Abstract-Electromagnetic Interference (EMI) and Electromagnetic Compatibility (EMC) are critical considerations in the automotive industry, directly affecting vehicle performance and regulatory compliance. This paper presents a comprehensive analysis of EMI/EMC challenges encountered across all vehicle fuel Platforms. Through detailed case studies, the paper identifies common sources of emission noise, explores rectification methods, and documents key lessons learned to enhance design rules and compliance. The study covers issues such as improper harness routing, insufficient grounding, and earlier release electronic components, which lead to emission noise across different frequency ranges. The findings underscore the importance of meticulous design practices, including proper routing, shielding, grounding, and the use of up-to-date hardware and software, in achieving vehicle-level EMI/EMC compliance. This work aims to provide practical guidance for automotive engineers and designers, contributing to the development of vehicles with improved electromagnetic compatibility.

DOI: /10.61463/ijset.vol.13.issue2.233

Farm Defender: Intelligent Perimeter Security

Authors- Sakshi Gadewar, Pragati Dubey, Prajakta Gorade, Sayli Ingole, Assistant Professor Mr. A. V. Saywan

Abstract-Agriculture is a crucial sector that supports economies and livelihoods, but it is constantly threatened by wild animals that invade farms, damage crops, and pose risks to livestock. Traditional methods of farm protection, such as manual surveillance and fencing, are often ineffective, labor-intensive, and costly. To address these challenges, this project is designed as an advanced, automated solution that ensures continuous monitoring and protection of farmland. This system provides real-time wild animal detection, immediately alerting the farmer to potential threats while ensuring quick and efficient response mechanisms. One of the key features of the Farm Defender is its directional LED indication system, which helps farmers identify the location of detected animals within the farm. In addition, the system is equipped with an alarm such as fences, scarecrows, and manual patrolling, mechanism, which produces a loud warning sound when an animal is detected. This not only alerts the farmer are either ineffective or require continuous human but also acts as a deterrent, scaring away animals before they can cause harm. Sustainability and renewable intervention. With the growing adoption of energy utilization are central to this project. The entire system is powered by solar energy, making it self-sufficient and ideal for remote agricultural areas where electricity supply may be unreliable. To enhance its efficiency and ensure continuous operation, a windmill is also integrated into the system, generating additional power, which is stored in a battery for uninterrupted performance. This project represents a significant step forward in protecting agricultural resources while reducing dependency on manual intervention, ultimately contributing to more resilient and productive farming practices.

Multi-dimensional Classification of Social Media Discourse Patterns and Electoral Outcomes in Madhya Pradesh: A Novel Machine Learning Framework Incorporating Regional Linguistic Features

Authors- Research Scholar Pavan Kumar Goyal, Dr. Prashant Sen, Dr.Anil Pimplapure

Abstract-This study proposes a novel machine learning framework to classify and analyze social media discourse patterns and their correlation with electoral outcomes in Madhya Pradesh, India. By incorporating regional linguistic features specific to the Hindi language variants spoken in the region, we develop a multi-dimensional classification system that captures nuanced political discourse across multiple platforms. Our approach combines natural language processing techniques with supervised learning algorithms to identify patterns in social media content that correlate with voting behavior. Results demonstrate that regionally-specific linguistic markers serve as significant predictors of electoral outcomes, outperforming standard sentiment analysis metrics. The findings contribute to understanding the evolving role of social media in shaping political landscapes in linguistically diverse regions and offer methodological innovations for studying digital political communication in non-Western contexts.

Temporal Dynamics of Social Media Engagement and Voter Mobilization in Rural Versus Urban Madhya Pradesh: A Hybrid Deep Learning Approach with Geospatial Correlation Analysis

Authors- Research Scholar Pavan Kumar Goyal, Dr. Prashant Sen, Dr.Anil Pimplapure

Abstract-This paper examines the relationship between social media engagement patterns and voter mobilisation across rural and urban areas of Madhya Pradesh, India. Utilising a hybrid deep learning approach combined with geospatial correlation analysis, we investigate temporal dynamics of digital political communication and its differential impacts on electoral participation. Our methodology incorporates recurrent neural networks with attention mechanisms to analyse time-series social media data, complemented by geospatial mapping of voter turnout statistics. Results indicate significant divergence in social media influence between rural and urban constituencies, with urban areas demonstrating more immediate response to digital campaigns while rural regions exhibit delayed but potentially more sustained engagement patterns. The study contributes to understanding the evolving role of digital platforms in Indian democratic processes and offers insights for targeted voter outreach strategies that account for regional digital divides.

Magnesium Titanate (MgTiO3) Properties, Characteristics and its Applications

Authors- Neeharika Bakhla

Abstract-In recent years the dielectric ceramic materials with high dielectric constant, large quality factor and a near zero temperature coefficient of resonant frequency are being thoroughly studied for fabricating miniaturized circuit. Magnesium titan ate (MgTiO3) is one of the leading and a near zero temperature coefficient of resonant frequency are being thoroughly studied dielectric material for microwave applications such as dielectrics in resonators, filters and antennas for wireless communication, radar and global positioning systems operating at microwave frequencies. Some recent reports [1] show that with the partial replacement of Mg by Co the ceramic shows better dielectric properties we have synthesized Mg_0.94 Co_0.06 TiO_3 ceramic in our lab and characterized its structural and dielectric properties of this material. Using this material, the revolutionary miniaturization of integrated circuitry has resulted in demand for microwave dielectric thin films. Thin films may have the advantages of lower crystallization temperatures and smaller devices than a bulk ceramics and can be integrated in microelectronic devices.

The Effectiveness of Financial Inclusion Schemes Awareness and Access in India

Authors- Dr.S.Saranya, Associate Professor Dr.K.Chandrasekar

Abstract-Financial inclusion refers to the process of ensuring that vulnerable groups, such as low-income individuals and communities, have timely access to financial services and appropriate credit at a fair cost. The primary goal of financial inclusion schemes is to provide secure financial solutions to India’s underserved sectors without discrimination or bias. The primary purpose of the study is to examine the efficacy of financial inclusion knowledge and access using empirical evidence from Tamil Nadu. This study is descriptive in nature, with a target audience in Tamil Nadu. The structured questionnaire collects data using Likert five-point rating scale questions. According to statistics, the degree of financial inclusion in various districts in Tamil Nadu has improved over time, but the bulk of rural districts remain in the medium inclusion group. According to this research study, financial inclusion helps to economic success and prosperity, which in turn stimulates activity and elevates the standard of life in all segments of society. This result illustrates that financial inclusion policies have the potential and capability of transforming the face of Tamil Nadu, and they are unquestionably working to improve all citizens. Because this study is confined to financial inclusion, future research might look into other aspects of financial inclusion like perception, involvement, and citizen empowerment, as well as a comparative analysis of different financial inclusion plans in different countries.

DOI: /10.61463/ijset.vol.13.issue2.234

Night Vision Rover

Authors- Krishnakant Vasule, Sarvesh Datir, Sanika Waidneshkar, Jhanvi Thakre, Asistant Professor Ashish. G. Kokate

Abstract-Robots are employed in specialized applications such as managing hazardous situations and performing tasks that demand high precision and speed. Often, dangerous incidents arise due to human negligence. Unauthorized individuals may inadvertently cross national borders, and it is impractical for soldiers to constantly monitor these areas. Border security requires continuous surveillance to detect any unusual activities. Traditionally, this task has been performed by human personnel; however, human limitations, such as fatigue and exhaustion, can lead to errors. The objective of this project is to enhance border security by providing a dependable and efficient solution. To achieve real-time surveillance and monitoring of border regions, an intelligent remote monitoring system has been developed. The proposed system focuses on the design and development of a “Night Vision Rover” equipped with wireless cameras to detect human presence and obstacles in remote areas, transmitting the collected information to a central location. The “Night Vision Rover” represents an advanced approach to strengthening border security measures. This system is built on a Node MCU platform integrated with the Blynk application, enabling seamless communication. The rover is powered by a motor driver that controls its four-wheel mobility, ensuring swift and adaptive movement. The integrated sensor suite includes a PIR sensor for human detection, an ESP32 Camera for clear real-time visual surveillance, and a sound sensor to capture and analyze audio disturbances near the border. These sensors operate collectively to monitor border regions, relaying instant notifications to the Blynk application. This functionality enables prompt responses to potential threats, enhancing border security and ensuring the safety of surrounding settlements.

DOI: /10.61463/ijset.vol.13.issue2.235

Referencing and Acquisition of Conceptual Tools in the Teaching of Architecture

Authors- Sabrina Ghattas, Kaouthar Zair

Abstract-The studio project in the first year of the undergraduate architecture program at the National School of Architecture and Urbanism in Tunis provides an introduction to architectural design. It offers the learner the opportunity to understand architectural space by perceiving, analyzing, designing, and finally transposing it through drawing and various modes and tools of architectural representation and expression. In this article, we present the methodological approach we have implemented to achieve one of the important didactic objectives: the acquisition of architectural references and interdisciplinary knowledge related to the field of architecture.

DOI: /10.61463/ijset.vol.13.issue2.236

Impact of Virtual Reality in Education

Authors- Assistant Professor Mr. Rakesh Jaiswal, Aditya Krishna, Lucky Singh Rajput, Divyansh Rathore, Kishore Bole

Abstract-Virtual Reality (VR) is rapidly gaining recognition as a powerful educational tool with the potential to revolutionize traditional teaching and learning methods. By immersing learners in interactive, simulated environments, VR bridges the gap between theoretical knowledge and real-world application. Unlike conventional classroom instruction, VR allows students to actively engage with learning content in three-dimensional, experiential contexts—enhancing comprehension, fostering critical thinking, and improving knowledge retention. This paper explores the transformative role of VR in education across various disciplines including science, engineering, medicine, history, and language learning. Through an in-depth review of existing literature and analysis of case studies, the study investigates how VR enhances student motivation, facilitates personalized and inclusive learning experiences, and promotes the development of practical skills in a risk-free environment. Furthermore, the research highlights how VR helps educators create more learner-centric environments, encouraging self-paced exploration and active participation. The findings also shed light on the technical, financial, and infrastructural challenges associated with VR adoption, such as equipment costs, content availability, and the need for teacher training. Despite these challenges, evidence suggests that VR-integrated instruction leads to significantly improved academic outcomes and learner satisfaction. The paper concludes that VR is not merely an auxiliary learning tool but a disruptive technology capable of shaping the future of global education systems. With careful implementation, policy support, and ongoing innovation, Virtual Reality can foster deeper engagement, broaden access to quality education, and redefine the very nature of teaching and learning in the 21st century.

DOI: /10.61463/ijset.vol.13.issue2.237

Role of Plumbing in Mitigating Water Scarcity Through Sustainable Industrial Practices

Authors- Ebrahim A. Omar

Abstract-Water scarcity is a growing concern in the Philippines, particularly in industrial sectors where high water consumption exacerbates resource depletion. This study examines the role of plumbing in mitigating water scarcity through sustainable industrial practices in Region 12 (SOCCSARGEN). A survey of 378 licensed master plumbers was conducted to assess awareness, adoption, and effectiveness of water conservation strategies. Findings reveal that water-efficient fixtures, rainwater harvesting, and leak detection technologies significantly reduce industrial water consumption, yet their adoption is hindered by high installation costs, limited technical expertise, and lack of policy enforcement. Correlation analysis indicates a strong positive relationship between awareness and adoption of sustainable practices, while financial constraints negatively impact implementation. ANOVA results further highlight variations in adoption across different industrial sectors. To address these challenges, the study recommends policy enhancements, training programs, and financial incentives to promote sustainable plumbing solutions in industrial settings, ensuring long-term water security and environmental sustainability.

DOI: /10.61463/ijset.vol.13.issue2.238

Experimental Study on Structural Audit of an Old Building in Amravati District: A Review

Authors- Samruddhi Ashok Pokale, Assistant Professor H. B. Dahake

Abstract-Concrete structures are subjected to environmental variations that lead to defects affecting their strength, durability, and service life. This study evaluates the impact of such defects on structural and non-structural elements using non-destructive testing methods, including the Rebound Hammer Test and Ultrasonic Pulse Velocity (UPV) Test. The Rebound Hammer Test results on a school building revealed significant variations in compressive strength across different columns, with some showing signs of deterioration. In non-structural elements, cracks (47%), dampness or leakage (28%), colour peeling (17%), and vegetation growth (8%) were identified as major defects, emphasizing maintenance concerns. The findings highlight the importance of periodic inspections, proper material selection, and effective curing techniques to enhance the longevity and performance of concrete structures. Timely interventions and repair strategies are recommended to prevent further deterioration and ensure structural integrity.

Hands-Free Virtual Mouse with Gesture, Eye, and Speech Recognition Using Machine Learning Algorithms

Authors- Assistant Professor Mr. M. Prathap, Birru Ganesh Vardhan Yadav, Mogadala Prem Kumar, Bemuni Palli Naga Suvarna, Tankasala Anjali

Abstract-The advancement of human-computer interaction (HCI) has led to the development of hands-free control systems, enabling accessibility and convenience in digital environments. This paper presents a Hands-Free Virtual Mouse utilizing gesture, eye, and speech recognition powered by machine learning algorithms to enhance user experience and accessibility. The system integrates multiple recognition techniques to simulate conventional mouse functions. Gesture recognition is achieved using computer vision and deep learning models, allowing users to perform cursor movements and clicks through predefined hand gestures. Eye tracking is implemented via facial landmark detection and gaze estimation, providing an intuitive pointer control mechanism. Additionally, speech recognition employs natural language processing (NLP) models to execute commands such as “click,” “scroll,” and “drag.”Machine learning techniques, including Convolutional Neural Networks (CNNs) for gesture detection, Support Vector Machines (SVMs) for gaze estimation, and Deep Neural Networks (DNNs) for speech command classification, are used to ensure high accuracy and responsiveness. The system is trained on extensive datasets to improve adaptability across different users and environments. This hands-free virtual mouse can significantly benefit individuals with physical disabilities, virtual reality (VR) applications, and smart home interfaces, reducing the dependency on traditional input devices. The results demonstrate that combining multiple recognition modalities enhances usability, making the system a robust alternative to conventional pointing devices.

DOI: /10.61463/ijset.vol.13.issue2.239

Automating Alzheimer’s Disease Prediction Using AI and Machine Learning Boosting Machines

Authors- Research Scholar Ms.A. Kamatchi, Associate Professor Dr. V. Maniraj

Abstract-Alzheimer’s disease (AD) is a progressive neurodegenerative condition that primarily affects the elderly. Although its symptoms start off mildly, they worsen as time goes on. While there is no cure for AD, early diagnosis can significantly mitigate its adverse effects. This study proposes a methodology called SMOTE-RF for AD prediction, utilizing machine learning (ML) algorithms. The performance of three algorithms—decision tree (DT), extreme gradient boosting (XGB), and random forest (RF)—is evaluated for this purpose. The experiments are conducted using the Open Access Series of Imaging Studies (OASIS) longitudinal dataset, which is available on Kaggle. To address class imbalance in the dataset, the synthetic minority oversampling technique (SMOTE) is applied. The experiments are carried out on both imbalanced and balanced datasets. On the imbalanced dataset, DT achieved an accuracy of 73.38%, XGB reached 83.88%, and RF obtained the highest accuracy of 87.84%. After balancing the dataset using SMOTE, DT achieved 83.15% accuracy, XGB reached 91.05%, and RF achieved the highest accuracy of 95.03%. The highest accuracy of 95.03% was attained with the SMOTE-RF model.

DOI: /10.61463/ijset.vol.13.issue2.240

The Strength of Blockchain Technology as an Advanced Security System for the Nigerian Financial Sector (Fintech) in Financial Transactions

Authors- Sulaiman Zakariyya Yakubu, Muhammad Auwal Ishaq

Abstract-Introduction: Blockchain technology is revolutionizing the financial technology sector (FinTech) by increasing security and transparency. In Nigeria, its adoption has the potential to revolutionize financial services, reducing the risks associated with traditional banking. This study examines the role of blockchain in securing financial transactions, examining its benefits and challenges in the Nigerian context. However, challenges related to regulation, scalability, and education must be addressed to ensure effective adoption. The study emphasizes blockchain’s ability to create a secure, transparent, and inclusive financial ecosystem, while recommending steps to overcome the identified challenges for successful implementation. Methodology: The methodology employed in this study involves a comprehensive review of existing literature, a qualitative analysis of current blockchain applications in the FinTech sector, and a case study approach to assess the impact of blockchain technology in Nigeria. The gender gap in blockchain and FinTech participation calls for initiatives to promote diversity and inclusivity. The age distribution shows that blockchain technology appeals primarily to individuals in the middle stages of their careers, likely due to their familiarity with digital advancements. The high representation of bachelor’s degree holders underscores the need for advanced education in blockchain-related fields. Implications: This study demonstrates the potential of blockchain technology to improve security, reduce fraud, improve financial inclusion and contribute to economic growth, in the Nigerian financial sector. For these benefits to be fully realized, it is necessary to address challenges related to regulation, scale and training. Blockchain can play an important role in building a safe and efficient financial ecosystem and move Nigeria towards economic progress.

DOI: /10.61463/ijset.vol.13.issue2.241

Hybrid AI: Combining Deep Learning with Neuromorphic Computing

Authors- Shreeyash Gharat

Abstract-Introduction: Hybrid AI is a new method in artificial intelligence that seeks to integrate deep learning and neuromorphic computing in order to enhance efficiency, flexibility, and energy consumption. Deep learning has achieved significant success in computer vision and natural language processing, but it comes with serious disadvantages such as high computational expenses, energy inefficiency, and poor scalability. By contrast, neuromorphic computing, which replicates the form and operation of biological neural networks, offers a low-power option in the form of spiking neural networks (SNNs), which supports real-time processing with reduced energy requirements. This paper explores the theoretical underpinnings of Hybrid AI, particularly how strong pattern recognition properties of deep learning can be properly merged with neuromorphic processing to generate more efficient AI models. It contains a comparison among deep learning, neuromorphic computing, and Hybrid AI systems based on their accuracy, energy efficiency, and computational speed. The research method includes applying a Hybrid AI model based on cutting-edge neuromorphic hardware and deep learning paradigms, and testing it for performance on actual tasks.

Foodies: A Web-Based Food Delivery Application

Authors- Aashutosh Yadav

Abstract-The food delivery industry has seen exponential growth with the rise of online platforms, enabling customers to order food from restaurants with ease. Foodies is a web-based food delivery application designed to provide an efficient and user-friendly experience. It leverages modern technologies such as React with Vite for frontend development, Spring Boot for backend processing, MongoDB for database management, and AWS S3 for image storage. The system is tested using Postman, and Maven is used as the build tool. This paper elaborates on the system architecture, features, technology stack, and performance aspects of Foodies, along with its impact on the food delivery market. Additionally, it highlights challenges faced during development and how they were overcome.

DOI: /10.61463/ijset.vol.13.issue2.242

Smart Lab System

Authors- Mr.Kunalsing M. Tomar, Mr. Pratik A. Pathak, Mr. Mandar V. Tayade, Mr. Ashwin P. Suradkar, Mr. Sagar B. Rangari, Assistant Professor Mr. A.V. Saywan

Abstract-This thesis presents a fingerprint-based smart home automation system that enhances security and convenience by allowing authorized users to control home appliances and access premises securely. The system employs biometric authentication to ensure only registered users can operate devices, integrating fingerprint recognition, microcontroller technology, and IoT connectivity. By replacing traditional keys with biometric access, the system improves security while offering a user-friendly interface for home automation. Additionally, voice-controlled Bluetooth automation is included to assist the elderly and injured in managing household utilities. The proposed system ensures efficient, secure, and accessible home automation with multiple authentication layers.

Exploring the Social and Economic Impacts of 6G Networks and their Potential Benefits to Society

Authors- Ibrahim Muhammad Hassan, Inusa Sani Maijama’a, Abdulrauf Adamu, Suleiman Bashir Abubakar

Abstract-6G networks are set to revolutionized society and the economy by providing ultra-fast connectivity, low latency and global coverage. From a societal perspective, 6G will bridge the digital divide by connecting remote and underserved areas, enabling virtual learning, remote health care and smart cities. It will improve public security by monitoring disasters in real time and improve the quality of life in an artificial intelligence-driven, sustainable environment. From an economic point of view, 6G will drive Industry 4.0, enabling smart factories, self-driving systems and efficient supply chains, boosting innovation and productivity. It will create jobs, stimulate investment and promote sustainable farming, education and transport. This paper focus on Potential benefits which include equal access to health and education, smarter urban living and global cooperation. However, realizing these benefits requires addressing ethical, regulatory and security issues in order to ensure an inclusive and sustainable take-up. Essentially, 6G promises to transform society and the economy, driving progress and improving lives around the world.

From Comments to Insights: Sentiment Analysis of YouTube Videos

Authors- Sujal Lothe, Komal Gulhane, Srushti Tembhare, Sujal Bhimanwar, Ashish Sawant

Abstract-Sentiment analysis is a method used to extract and interpret user opinions and perspectives on products or services. YouTube, as one of the foremost video-sharing platforms, generates millions of views and a vast array of comments. These comments provide valuable insights that can be utilized to enhance both the ratings and the overall quality of the content uploaded. Through natural language processing (NLP) and machine learning (ML) models, these comments can be analyzed in a highly effective manner. Research on sentiment analysis has explored numerous classification techniques, from binary models (positive or negative) and ternary models (positive, negative, neutral) to multi-class systems (such as joy, sadness, fear, surprise, and anger). However, identifying the most effective and precise method remains an ongoing challenge. In addition, sentiment analysis is frequently used to evaluate the emotional nuances expressed in comments on YouTube videos. This paper delves into various approaches and methodologies employed for sentiment analysis within the realm of YouTube videos. It systematically categorizes these techniques, offering significant insights for researchers engaged in the fields of data mining and sentiment analysis.

DOI: /10.61463/ijset.vol.13.issue2.243

Seasonal Effect on Resting Behavior of Haryana Calves under Different Flooring Systems

Authors- Assistant Professor Mamta, Associate Professor Rajneesh Sirohi, Associate Professor Yajuvendra Singh, Professor Brijesh Yadav, Associate Professor Mukul Anand, Assistant Professor Ajay Kumar, Srashti Dixit

Abstract-Livestock markets then began producing different kinds of flooring mats and bedding materials, which, along with increased demand, claimed to be of utmost significance in cow growth and health. Although the claims of rubber mat are very high yet regarding indigenous dairy animals scientific studies that too season wise, are meager. Few studies have been carried out in India to observe the impact of flooring on animal behavior. Considering these facts as gap the present study was undertaken to study the season wise impact of concrete, rubber and cow dung bed flooring on resting behavior of Haryana calves at the Livestock Farm complex, DUVASU, Mathura on eighteen Haryana calves in two seasons April to June for summers and December to February for winter. The selected calves were randomly divided into three groups, containing six animals in each group: (T1-Control) Concrete flooring, (T2) Compost/cow dung bed flooring and (T3) Rubber mat installed flooring. Behavioural attributes of calves were investigated by using Video recording system. The overall lying duration was found significantly different between the treatment groups as T1<T3

Different Methods of Monitoring Heat Stress in Dairy Animals

Authors- Srashti Dixit, Associate Professor Rajneesh Sirohi, Associate Professor Yajuvendra Singh, Assistant Professor Mamta, Assistant Professor Ajay Kumar, Vishakha Singh Gaur, Lavish Chelani, Vinayak Jaswal, Akshat Kaushik

Abstract-Heat stress nowadays has become a grave issue, especially in a country like India. Heat stress can lead to decreased output, decreased animal welfare, decreased fertility, increased susceptibility to illness, and in severe situations, increased mortality. Rising temperatures, an increase in the number of production animals retained globally, and the advancement of agriculture, particularly in emerging economies, present a serious challenge to the global dairy industry. Addressing heat stress proactively can help maintain dairy animal health and productivity, ensuring the well-being of the animals and the efficiency of dairy operations. Invasive methods of heat stress monitoring involve a hematological profile, blood-biochemical profile, a thermometer for rectal temperature and stethoscope for heart rate and pulse rate. Non-invasive methods having no human interference include Bioclimatic indices (THI, BGHI, HLI, THIadj, ETI, CCI, ITSC) which use environmental parameters for heat stress assessment. Infrared thermography uses thermal images to detect the surface temperature of animals, Biosensors (Accelerometers, Ear tag, Bolus sensors, Data loggers) fitted with transmitters can transmit the data to a distant place. Mobile applications are also in use which can transmit data via Bluetooth technology. Cortisol concentration in the feaces and hair of animals and fecal microbiome estimation is also indicative of heat stress. Monitoring heat stress with these approaches and responding appropriately and timely can significantly reduce its negative effects. The main focus of this review is to summarize the data related to initiatives aimed at creating livestock heat stress detection systems that take into account various physiological and behavioral reactions.

Food Calorie Tracker: A Deep Learning Based Full-Stack Web Application for Real Time Food Recognition and Calorie Estimation

Authors- Professor Dr. P. Vara Prasad, P. Venkata Vamsi Madhav, Y. Guruprasanth Reddy, P. Subhahan, A. Harikrishna

Abstract-Obesity, a severe and growing chronic disease, has been worsened by the increasing convenience of food delivery services. As access to food has expanded, so too has concern over nutritional habits and health. The main goal of this project is to accurately identify various food items and estimate their calorie content in real-time, offering users smart and personalized diet monitoring capabilities. This project proposes a Deep Learning-based solution as a Full Stack Web Application for food image recognition and calorie estimation. Developed with a React-based frontend and a Python-powered backend, the system allows users to upload images of food items for real-time classification and calorie analysis. By leveraging advanced deep learning architectures such as YOLO (You Only Look Once), the application effectively tackles the challenges of accurate classification and feature extraction in large and diverse datasets, such as the IndianFoodNet-30 dataset. This integration ensures precise recognition and efficient calorie estimation, offering users a seamless and intelligent diet monitoring experience. This full-stack solution demonstrates the integration of modern web technologies with robust deep learning models, delivering a scalable and user-friendly tool for real-time food recognition and calorie monitoring. The frontend offers an intuitive interface for uploading food images and visualizing results, while the backend efficiently handles image processing, deep learning inference, and calorie computation.

DOI: /10.61463/ijset.vol.13.issue2.244

Next Word Prediction

Authors- Anas Khan, Aazain Khan, Abhinav Chaudhary, Ahmed Mehdi Zaidi, Mr. Amit Kumar Saini

Abstract-Next-word prediction models greatly improve user engagement and are essential in applications such as voice typing, virtual assistants, and predictive text systems. In order to enhance next-word prediction in speech-driven applications, this study presents a hybrid technique that blends voice recognition with Long Short-Term Memory (LSTM) networks. A voice recognition system that translates audio into text is used by the model to process spoken input. LSTMs, which are well-known for their ability to capture sequential patterns and long-term relationships in language, are then used to examine the text data. Accurate context comprehension and prediction are guaranteed by this two- layered structure. The system can process spoken inputs in real time and provide precise next-word predictions even in loud and dynamic contexts thanks to the combination of speech recognition with LSTMs. Its versatility and resilience are demonstrated through experimental assessment on a variety of datasets. Next-word prediction models are essential for improving human-computer interaction, especially in voice-enabled technologies and text input systems. In order to develop a reliable next-word prediction model specifically for spoken language, this study investigates the combination of voice recognition and Long Short-Term Memory (LSTM) networks. Speech recognition converts audio input into text, which LSTM networks subsequently examine to provide contextually aware word predictions. Even in dynamic and chaotic situations, the model performs exceptionally well at capturing linguistic connections in real-time. The system’s accuracy and adaptability are demonstrated through evaluation on a variety of datasets, underscoring its potential for use in voice typing, virtual assistants, and accessibility aids for hands-free operation. Intelligent computers are now able to accurately anticipate text thanks to developments in natural language processing. This paper presents a next-word prediction model that combines voice recognition with Long Short-Term Memory (LSTM) networks. After converting spoken input into text, LSTM analyzes the string and uses contextual information to anticipate the following word. In speech-driven applications, the system outperforms conventional text-based models with notable accuracy gains.

Server Inventory Management

Authors- Ms. Theekcika. M, Assistant Professor Mr.Yuvaraj. V

Abstract-Server inventory management is a critical aspect of maintaining and optimizing IT infrastructure in organizations. This system involves the systematic tracking, monitoring, and management of servers, hardware components, and related software, ensuring efficient resource utilization, cost management, and operational continuity. The implementation of an effective server inventory management solution allows businesses to maintain accurate records of server configurations, and server locations. It also facilitates proactive maintenance, fault detection, and timely upgrades, reducing downtime and enhancing overall system performance.

DOI: /10.61463/ijset.vol.13.issue2.245

Train Accident Prevention: A Comprehensive Approach to Safety Enhancements

Authors- Harsh Malik, Mohd.Zubair, Ajay Kumar

Abstract-The safety of train transportation is critical in preventing fatal accidents and ensuring the safe movement of passengers and goods. This paper investigates key factors contributing to train accidents and proposes preventive measures to enhance safety. The study focuses on human error, mechanical failure, and infrastructure deficiencies as the main causes of train accidents. We employ a combination of historical data analysis, simulation modeling, and expert interviews to identify critical risk points and suggest improvements. The findings highlight the role of automation, advanced signaling systems, and regular maintenance schedules in reducing accidents. Additionally, better driver training and safety protocols are recommended to mitigate human error. The results suggest that integrating AI-powered monitoring systems can significantly improve early detection of anomalies, reducing response times and preventing accidents. The implications of these findings underscore the importance of technological innovation and strict adherence to safety regulations in preventing train accidents.

Design and Analysis of Steering Arm for Minimum Error Condition for Off Road Commercial Vehicles

Authors- Mr. B. Bharath Kumar, Syed Mujahid Hussain, Pynda Sarath, Balla Abhiram, Boddu Durga Prasad

Abstract-Steering systems are a critical component in vehicles, enabling drivers to change direction while maintaining stability, control, and safety. They play an essential role in ensuring the vehicle’s directional stability, allowing for smooth turns without loss of traction. Steering systems must be designed to provide precise handling and reliability under various operating conditions. The most commonly used steering geometries include Davis and Ackermann, which offer specific advantages in terms of wheel alignment and maneuverability. Popular types of steering systems include trapezoidal, double trapezoidal, and rack-and-pinion mechanisms, each with distinct applications depending on the vehicle’s requirements. Steering mechanisms are generally categorized into two types: direct and indirect. Direct steering involves rotating the wheel directly, while indirect steering uses push-pull or up-down mechanisms. A twin lever steering mechanism, for example, is a push-and-pull type that utilizes advancements in science and technology for enhanced control. This mechanism is controlled primarily by bi-articular muscles, offering ergonomic advantages in certain applications. In this project, the focus is on analyzing the steering arm, a vital component that connects the steering system to the wheel hub and transmits the force from the steering wheel to rotate the wheels directly. The design of the steering arm is critical for achieving the desired steering performance and ensuring structural integrity under applied loads.

DOI: /10.61463/ijset.vol.13.issue2.246

Optimization and Comparision of Linear Regression Model of Process Parameters Using Anova and Taguchi Methods

Authors- Mr. K. Sree Ramachandra Murthy, Yallaboyina Siva Shankar Veera Prasad, Pepakayala Dinesh, Bhargav Manikanta Meenavali, Thirumalasetty Veera Venkata Satya Sai

Abstract-In the Material Extrusion (MEX) process, optimizing key parameters such as layer height, line width, and print speed is crucial for enhancing print efficiency, reducing material consumption, and minimizing production time. This study employs ANOVA (Analysis of Variance) and Taguchi Methods to develop and compare a linear regression model that predicts print time and material usage. Additionally, the predicted values are compared with Cura 5.9.0 software results to validate the model’s accuracy. The Taguchi method, a robust statistical optimization tool, is used to determine the most influential factors affecting the print time and material consumption by analyzing the signal-to-noise ratio. ANOVA is applied to assess the significance and interaction of each parameter, ensuring reliable optimization. Layer height impacts print resolution and time, line width influences extrusion accuracy and material flow, and print speed directly affects the total build time. A linear regression model is developed based on experimental data, predicting output responses under different parameter settings. The model’s results are then compared with Cura’s estimated print time and material consumption values, highlighting any deviations and potential areas for improvement. The findings indicate that Cura’s slicer predictions align closely with experimental results but exhibit minor discrepancies due to software-based assumptions in deposition dynamics.

DOI: /10.61463/ijset.vol.13.issue2.247

Design of Mechanical Grab For Lifting Plates

Authors- Mr. B. Hari Krishna, Uppala Pavan Sai, Shaik Habibul Rehman, Yarra Durga Venkata Krishna, Thota Dhanush

Abstract-This report will provide valuable information to a mechanical engineer to build up an engineering approach in designing a purposeful motorized grab. Through reading of this report, an individual can know how to precede a project of designing the grab and which parameters should be of prime focus in research and development phase these motorized mechanical grabs can be classified into different types. Motors with electromagnetic brakes are used. This is an important safety feature that prevents the unit from accidentally opening under load. To improve the efficiency, safety and quality of material handling functions, Mechanical Grabs can be used in the heavy industry where material handling is quite frequent as it not only ensures safety but also allows consumer to save time and costs. The Automated Telescopic action in the mechanical grab not only provides control to the user but also provides ease for lifting sheets of variable sizes resulting in optimized work flows.

DOI: /10.61463/ijset.vol.13.issue2.248

3d Modelling and Simulation of Industrial Robot Arm Under Static Loading Condition

Authors- Mrs. K. Tulasi, Lokareddy Durga Sri Venkata Sai, Sana Vinay Kumar, Mattaparthi Kishore, Sattineedi Pavan Naga Venkata Sai Kumar

Abstract-Computer Aided Design (CAD) and Computer Aided Engineering (CAE) is the most important and essential tool in product development process. Huge challenge is faced by the companies while integrating CAD and CAE in their design process. The previous studies do not clearly give the impact of CAD and CAE on product development process and particularly its impact on cost and time of development. The study is carried out to show the importance of CAD and CAE as a tool of product development and its effect on the development cost and time when implemented early in the process. Computer -aided engineering (CAE) for systems analysis is needed to address conceptual and preliminary design for a broad class of products. System analysis encompasses many engineering disciplines and the associated CAE technology will play a key role in product development and design. Improved products can be designed by employing easy-to-use model building tools and sophisticated mathematical algorithms, which accurately predict performance and cost of future products. Lessons learned from CAD/CAM, particularly in the graphics and database management areas, combined with recent algorithm research, and supported with sufficiently powerful hardware will enable CAE systems to achieve marked productivity improvements. The authors describe their experience in computer-aided engineering for control system design. This discussion forms the basis of identifying future CAE needs and their realizations.

DOI: /10.61463/ijset.vol.13.issue2.249

Face Centered Central Composite Design to Optimize The Material Extrusion Process

Authors- Mr. A. Yeswanth, Penke Jagan Mohan, Pichika Balaji Sai Manikanta, Badugu Likhilbabu, Kalvakolanu Raj Kumar

Abstract-Material Extrusion (MEX), commonly used in Fused Deposition Modeling (FDM), requires precise process optimization to balance printing efficiency, surface quality, and material usage. This study employs the Face-Centered Central Composite Design (FCCCD), a statistical Design of Experiments (DOE) method, to optimize shell thickness, layer width (line width), and travel speed, aiming to minimize printing time while enhancing surface finish. The FCCCD approach allows a systematic evaluation of parameter interactions by generating a response surface model. Shell thickness affects part rigidity and material consumption, layer width influences bonding strength and print resolution, while travel speed impacts deposition accuracy and total fabrication time. Experiments were conducted using Fusion 360’s process simulation tools, followed by statistical analysis to identify optimal settings. Results indicate that optimized shell thickness and line width contribute to a better surface finish, while increasing travel speed reduces printing time but can affect deposition accuracy. The response surface methodology (RSM) derived from FCCCD enables efficient parameter tuning, ensuring a balance between production speed and part quality. This study demonstrates that FCCCD is a reliable approach for optimizing MEX process parameters, leading to improved efficiency and reduced waste. The findings provide a scientific foundation for enhancing FDM-based manufacturing, making it more cost-effective and sustainable.

DOI: /10.61463/ijset.vol.13.issue2.250

Cloud Computing Adoption for Entrepreneurship Development in Africa: Risk, Benefits and Good Practice

Authors- Ibrahim Muhammad Hassan, Muktar Hussaini, Asmau Isyaku Dutse, Abdulrauf Adamu

Abstract-Cloud computing has become a game changer, providing scalable, affordable, and easy-to-use solutions for businesses worldwide. The use of cloud computing can transform Africa, where entrepreneurship is booming and where it offers previously unheard-of opportunities for growth and innovation. However, in order to achieve their full potential, African business owners need to be aware of the risks, benefits and best practices associated with the adoption and implementation of the Directive. This paper therefore examines and analyses the factors influencing the uptake of CC through various theoretical models of adoption, which help to understand the factors influencing its uptake and uptake. The models used in the study are the TOE (Technology-Organization-Environment) framework, the DI (Diffusion of Innovation) theory, the UTAUT (Unified Technology Acceptance and Use Theory) and the TAM (Technology Acceptance Model). These theories and models provide insight into the uptake and use of technologies by individuals and organizations. In addition to offering guidance on how to successfully adopt and use new technologies, they provide frameworks for examining the factors influencing decision-making on adoption. The study’s findings concluded that by taking prudent risks and making optimal use of the benefits of cloud computing, companies can use cloud technology to drive innovation, productivity and growth, while avoiding potential risks.

Design and Simulation of Scaffold Design for Tissue Regeneration

Authors- Mr. A. Phani Bhaskar, Nandikolla Satish, Pithani Veera Naveen, Koppisetti Bhaskara Manikanta, Randi John Wesly

Abstract-Tissue engineering scaffold is a biological substitute that aims to restore, maintain or to improve tissue functions. The scaffold for orthopedic surgeries need to be designed very accurately for absolute benefit to the patient. In this study various unit library structure designs for scaffold are presented for improvement in scaffold characteristics and advancements in its applications. These structures are designed to achieve the specific mechanical structure and properties which are superior to the available conventional techniques of scaffold fabrication. As the requirements for a better scaffold are met adequately through better and appropriate designing techniques, the prospects of fabricating a more successful engineering scaffold also improves. In this study, Finite element analysis was performed to investigate the characteristic features of the unit library structure designs to evaluate and examine their mechanical properties including porosity, effective modulus, compatibility with other designs and stress distribution. DICOM images for human foot were processed and reconstructed using image processing tools and 3D reconstruction software. The validated design of unit library structure was made in the bulk form and given a shape of reconstructed bone. Fused Deposition Modelling was used to manufacture the validated design.

DOI: /10.61463/ijset.vol.13.issue2.251

Taguchi Grey Scale Optimization of FDM Process Parameters for Reducing Printing Cost by Using Doe Software

Authors- Mr. P. Ram Prasad, Pitchuka Pavan, Sai Varma Konduri, Madhavarapu Vijay Kumar, Abbireddy Veera Prabhas

Abstract-Fused Deposition Modeling (FDM) is a widely used additive manufacturing process due to its cost-effectiveness and versatility. However, optimizing process parameters is essential to achieve a balance between printing cost, time, and part strength. This study employs Taguchi Grey Scale Optimization (TGSO) to optimize key FDM parameters—orientation, layer thickness, and printing speed—using DOE software to minimize printing cost while maintaining structural integrity. The Taguchi method was used to design experiments with multiple parameter variations, analyzing their influence on printing time and mechanical strength. Grey relational analysis (GRA) was then applied to evaluate multiple performance characteristics and determine the optimal settings. Orientation affects material usage and support requirements, layer thickness influences surface finish and strength, and printing speed directly impacts manufacturing efficiency and cost. Experimental results indicate that optimized process parameters significantly reduce printing time and material consumption while preserving part strength. The Grey Taguchi method proved effective in identifying the most cost-efficient parameter combination, offering a structured approach to multi-objective optimization in FDM. This study highlights the importance of process parameter optimization in reducing operational costs and enhancing FDM efficiency. The findings serve as a valuable guideline for manufacturers and researchers seeking to improve the economic and mechanical performance of 3D-printed components while ensuring sustainability and resource efficiency in additive manufacturing.

DOI: /10.61463/ijset.vol.13.issue2.252

Design, Geometrical and Material Optimization of Engine Fins for Maximum Heat Transfer

Authors- Mrs. P. Gayathri, Siriki Abhiram, Sidda Satya Anantha Kumar, Thogaru Durga Mani Govind, Kona Murali Krishna

Abstract-The Engine cylinder is one of the major automobile components, which is subjected to high temperature variations and thermal stresses. In order to cool the cylinder, fins are provided on the surface of the cylinder to increase the rate of heat transfer. By doing thermal analysis on the engine cylinder fins, it is helpful to know the heat dissipation inside the cylinder. We know that, by increasing the surface area we can increase the heat dissipation rate, so designing such a large complex engine is very difficult. The main aim of the present paper is to analyze the thermal properties by varying geometry of cylinder fins using Ansys work bench. The 3D model of the geometries are created using SOLIDWORKS and its thermal properties are analyzed using Ansys workbench R 2024. The variation of temperature distribution over time is of interest in many applications such as in cooling. The accurate thermal simulation could permit critical design parameters to be identified for improved life. Presently Material used for manufacturing cylinder fin body is Aluminium Alloy AA 6061 which has thermal conductivity of 160 – 170 W/mk. presently analysis is carried out for cylinder fins using this material.

DOI: /10.61463/ijset.vol.13.issue2.253

Robotic Process Automation: ID Card Generation System

Authors- Ashwini Rahude, Priyal Patil, Shruti Pednekar, Sanika Savekar

Abstract-The ID Card Generator System, developed using UiPath Studio, is designed to automate the process of generating identification cards for members, staff, and students within an organization. This system eliminates the need for manual card creation by providing an efficient and user- friendly platform for ID card generation. Key features include automated data collection and standardized card design, ensuring uniformity and accuracy. By streamlining the ID card issuance process, the system enhances operational efficiency while maintaining security and reliability. This solution serves as a robust and effective option for organizations seeking to optimize and automate their identification card management. In this paper, we have studied and discussed about the different tools used for developing software robots, traditional methods of completing the task without RPA being used and also how the task is automated after the use of software robots in various fields, viz., banking, e- commerce, educational institutions, human resources, health care and office. We have also discussed the pros and cons of automating the task using robotic process automation. In short, this paper will give an overview of the emerging technology – RPA.

Exploration and Chemical Constituent’s Analysis of Flowers in Around Coimbatore District

Authors- Sriram R, Adarsh A, Manikandan S, Shadhashree K, Mareeswaran M, Sabapathy J.K, Kalaivani M, Assistant Professor Dr. S. Vinodhini

Abstract-The Ayurvedic medical system is a treasure of medicinal preparations, both single and combined formulations, and herbal preparations represent the majority of them. Flowers play a significant role in our daily lives, both directly and indirectly. Flowers are one of the useful parts of the plant source. Hence, the purpose of this study was aimed at collecting the classical references of commonly available flowers that have been mentioned in Ayurveda for their medicinal uses. Results show that, as there are several flower drugs in the Ayurvedic system of medicine that are useful for treating different types of disease conditions/ailments, a review has been made focusing on the therapeutic uses, brief descriptions, major chemical constituents, and compound formulations contributing to the popularization of various important medicinal uses of these flowers. As a result, it may be inferred that flowers are one of the most important components of a plant in disease treatment, and ‘flowers as medicine’ will be a good method that adds value to Ayurveda.

A Robust Online Application for Streamlined Bookkeeping, Secure Payments, and Reliable Peer-to-Peer Exchanges

Authors- Assistant Professor Shivangi Sharma, Devansh Gautam, Sanjana Rajput, Purva Pardhi, Ritik Ghosh

Abstract-This paper presents the designing of a web-based financial application that implements digital bookkeeping and payment management features considering the needs of small and medium-sized businesses (SMBs). Identified financial management challenges for SMBs to track their expenses and debts as well as to make UPI-based payments are considered. The added feature of the app is the sound peer-to-peer payment with state-of-the art technologies including usage of biometric authentication, face recognition, and near-field communication. This will enable smooth financial transactions and, consequently, enhance security thereby reducing reliance on intermediaries and risk of frauds. It then delves into technical architecture, security features, and user experience, extending that with business implications of the application in a scenario of commission-based benefits. The thesis then discusses market potential and scalability of the app .It is the upgraded version in which we used NFC.

AI-Powered Cybercrime Detection: A Comprehensive Review of Machine Learning and Deep Learning Techniques

Authors- Mrs. G.V Rajeswari, I.Charmila, B.Dileep Kumar, L.Tulasi Guru Charan, P.Sri Aditya, A.Neha Sai Sritha

Abstract-This study investigates the application of deep learning and machine learning techniques in cybersecurity for detecting and mitigating cyber threats. It specifically focuses on utilizing Long Short-Term Memory (LSTM) networks and Graph Neural Networks (GNN) to analyse network traffic and system logs, identifying patterns, anomalies, and potential malicious activities. The objective of this research is to strengthen network security and enhance cyber defence mechanisms by leveraging Artificial Intelligence (AI) to address the challenges posed by constantly evolving cyber threats.

DOI: /10.61463/ijset.vol.13.issue2.254

DeepCrackNet: MobileNet-Driven Deep Learning Framework for Image-Based Crack Detection Using Transfer Learning

Authors- B.L.K.Vamsi, K.Sravya, K.V.V.Sai Karthik, L.Niha Jyothi, Y.P.S.Bhavaraja Praneeth, Mrs.P.Satyavathi

Abstract-Crack in infrastructure present significant challenges to public safety, necessitating timely detection for effective maintenance. This research introduces Deep Crack, a novel deep-learning approach for image-based crack prediction. Utilizing the power of Convolutional Neural Networks (CNNs), the study employs the Rfcn_b architecture as its backbone, leveraging transfer learning to enhance crack detection capabilities. The proposed methodology incorporates extensive data preprocessing, including image augmentation techniques to mitigate data scarcity. The model is trained and validated on a diverse dataset, effectively distinguishing images containing cracks from those without. A custom top layer is integrated into the classifier, featuring global average pooling and dense layers to optimize performance. To evaluate the model’s effectiveness, a comprehensive assessment framework is introduced, including confusion matrices and classification reports. The results demonstrate that Deep Crack achieves high accuracy, precision, and recall, validating its potential for real-world infrastructure monitoring and maintenance. This research contributes to the intersection of deep learning and image processing, offering an innovative solution for automated crack detection. The proposed methodology not only highlights the effectiveness of Rfcn_b and transfer learning but also underscores the broader applications of deep learning in civil engineering and infrastructure management.

DOI: /10.61463/ijset.vol.13.issue2.255

Deep Learning-Powered Android Malware Detection and Prevention Leveraging Real-Time Twitter Threat Intelligence

Authors- Mrs.K.S.R. Manjusha, R.Sai Hrushi, K.Devi Naga Sri Aditya, U. Siva Tarun, K.Miriyam Nissy, D.Adhithi

Abstract-Smartphones have become an integral part of daily life, making security and privacy paramount, particularly for the Android operating system, which dominates the market. However, its widespread adoption also makes it a prime target for malware attacks, posing significant threats, especially to IoT devices dependent on Android applications. This paper presents a multilayer approach to Android malware detection, combining real-time data extraction from Twitter with deep learning techniques. Our method continuously updates a malware hash database every 48 hours using Twitter data, ensuring the latest threats are identified. Additionally, a deep learning model based on a Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM) is employed to analyse application permissions, achieving a 94% detection accuracy. This integrated approach offers a robust and adaptive solution to enhance Android OS security.

DOI: /10.61463/ijset.vol.13.issue2.256

Brain Tumor Detection and Classification Using Vision Transformers, Ensemble Learning, and Transfer Learning with Explainable AI Insights

Authors- Mrs. L.Yamuna, D.Surya Naga Laxmi Sahithi, K.S.S. Durga Hari Prasad3, V.Subhash, M.Hari Abhilash, Shaik Farha

Abstract-The abnormal development of either malignant or benign tissues in the brain results in long-term damage to its function. Magnetic Resonance Imaging (MRI) is a widely used technique for detecting brain tumors. Upon receiving MRI images, specialists physically examine the filters to assess whether a patient has a brain tumor. However, the interpretations of MRI images may vary between different experts, potentially leading to inconsistent results as professionals may have differing evaluation methods. Moreover, simply identifying the presence of a tumor is insufficient; it is also critical to determine the tumor type to initiate treatment promptly. This paper focuses on the multiclass classification of brain tumors, as much research has been conducted on binary classification. To enhance the speed, accuracy, and reliability of tumor detection, we explored the effectiveness of several deep learning (DL) architectures, including VGG16, InceptionV3, VGG19, ResNet50, InceptionResNetV2, and Exception. Based on this analysis, we propose a transfer learning (TL)-based multiclass classification model, IVX16, which leverages the top three performing TL models. The dataset used consists of 3,264 images, and after conducting thorough experiments, we obtained the following peak accuracies: 95.11% for VGG16, 93.88% for InceptionV3, 94.19% for VGG19, 93.88% for ResNet50, 93.58% for InceptionResNetV2, 94.5% for Exception, and 96.94% for IVX16. Additionally, we applied Explainable AI techniques to assess the performance and validity of each DL model and incorporated the recently developed Vision Transformer (ViT) models. We compared the results obtained from these ViT models with those from the TL and ensemble models.

DOI: /10.61463/ijset.vol.13.issue2.257

Smart Crop Recommendation: A Hybrid AI System Integrating Machine Learning and Deep Learning for Precision Agriculture

Authors- Mr.A.Janardana Rao, R.Usha, K.Bala Venkata Adithya, P.Anil Kumar, Ch.E Naga Sai Priya, S.Satya Kumar

Abstract-Recent advancements in agriculture have led to the development of crop recommendation systems, offering a new approach to optimizing crop yields and the efficient use of resources. In this study, we introduced an enhanced “Crop Recommendation System” designed to assist farmers in selecting suitable crops based on various soil and environmental factors. Our system leverages machine learning (ML) and deep learning (DL) models to enhance agricultural productivity and crop yields. The dataset used for training and evaluation was sourced from Kaggle’s repository, featuring key parameters such as nitrogen (N), phosphorus (P), potassium (K), rainfall, pH, temperature, humidity, and crop labels corresponding to 22 different crop types. The study incorporates several ML and DL algorithms, including Decision Trees, Random Forest, XGBoost, Support Vector Machine, K-Nearest Neighbors, Naive Bayes, Artificial Neural Networks, Deep Neural Networks, and Temporal Convolutional Networks. According to performance metrics, Random Forest and TCN delivered the highest accuracy, achieving 99.2% and 99.9%, respectively. Other models also performed impressively, with accuracy ranging between 93.8% and 98.7%. Although Support Vector Machine (SVM) showed slightly lower performance, with 93.4% accuracy, it still yielded satisfactory results. The project also explores parameter tuning for XGBoost and TCN, with TCN outperforming XGBoost after optimization. The findings of this research indicate that both machine learning and deep learning models are highly effective in crop recommendation systems, wisth TCN providing accurate and efficient recommendations. Furthermore, this study aids precision agriculture by offering a web-based interface for farmers, helping them select crops based on environmental and soil conditions.

DOI: /10.61463/ijset.vol.13.issue2.258

AI-Driven Time Series Forecasting for Financial Markets: Leveraging Machine Learning for Smarter Predictions

Authors- Mrs.V.Anantha Lakshmi, S.Venkata Basavayya, Y.S.Santosh Kumar, P.Bhanu Divyasri, V.Sandeep, S.Likhita

Abstract-Financial markets, including stock prices, exchange rates, and commodity prices, are inherently volatile and influenced by numerous factors, making their prediction a challenging yet essential task. Accurate forecasting of market trends is crucial for investors, financial analysts, and policymakers, as it helps in making informed decisions and mitigating risks. In this study, we explore the use of Support Vector Machine (SVM), a powerful machine learning algorithm, for time series forecasting of financial market trends. Traditional forecasting methods often struggle with financial data due to its non-linear and dynamic nature. However, SVM is well-known for its ability to handle high-dimensional data and capture complex patterns, making it a suitable choice for financial market prediction. Our approach leverages historical price and volume data to train the SVM model, enabling it to recognize patterns and predict future market movements. The study evaluates how effectively SVM adapts to changing market conditions, demonstrating its ability to model non-linear relationships within financial time series. Additionally, we consider external economic factors that may influence market behavior, further validating the robustness of the model. The findings highlight the potential of SVM in financial forecasting, offering a reliable alternative to traditional methods. Future work may involve integrating hybrid models combining SVM with deep learning techniques or incorporating macro-economic indicators to further enhance prediction accuracy. This research contributes to the growing field of AI-driven financial analysis, paving the way for more sophisticated and data-driven investment strategies.

DOI: /10.61463/ijset.vol.13.issue2.259

Interpretable AI-Driven Wireless Capsule Endoscopy Image Classification: Advancing Explainability in Medical Diagnostics

Authors- Dr.A.Radha Krishna, Ch.Raja Rajeswari, Y.V.S.Kartheek, L.Bhuvan Sai Ram, R.Arya Vinayaka Venkata Siva Sai, K.Teja Ramyasr

Abstract-Deep learning has significantly advanced medical imaging and computer-aided diagnosis (CAD), providing powerful tools for disease detection. While various deep learning (DL) models exist for medical image classification, further analysis is needed to better understand their decision-making processes. To address this, several explainable AI (XAI) techniques have been proposed to improve the interpretability of DL models. In the field of endoscopic imaging, medical professionals primarily rely on visual inspections for preliminary disease diagnosis. However, integrating automated deep learning systems can enhance both efficiency and accuracy in medical assessments. This study aims to increase the reliability of model predictions in endoscopic imaging by implementing multiple transfer learning models on a balanced subset of the Kvasir-Capsule dataset, a widely used wireless capsule endoscopy (WCE) imaging dataset. The dataset subset includes the top 9 classes for training and testing. Our results demonstrate that the Vision Transformer model achieved an F1-score of 97% ± 1%, outperforming previous studies on the same dataset. Other models, such as MobileNetV3Large and ResNet152V2, also performed exceptionally well, achieving F1-scores above 90%. These results mark a significant improvement over prior benchmarks. To enhance model interpretability, we employ multiple XAI techniques, including Grad-CAM, Grad-CAM++, Layer-CAM, LIME, and SHAP, to generate heat maps that highlight key decision-making regions.

DOI: /10.61463/ijset.vol.13.issue2.260

Hybrid Intrusion Detection System Using SVM for Anomaly and Misuse Detection in Networks

Authors- Professor& HOD Jayshree Boadh, Seema Narware

Abstract-The growing complexity and volume of cyber threats demand efficient and reliable intrusion detection systems (IDS) to safeguard network environments. This research presents a hybrid intrusion detection system that leverages Support Vector Machines (SVM) to address both anomaly and misuse detection in networks. The proposed system integrates the strengths of anomaly-based detection, capable of identifying novel and zero-day attacks, with misuse-based detection, which excels at recognizing known attack patterns. By employing SVM, the system ensures high accuracy in classifying network traffic while minimizing false positives and negatives. The hybrid approach involves preprocessing network traffic data to extract relevant features, which are then classified using SVM. Anomaly detection identifies deviations from normal network behavior, while misuse detection applies pre-defined signatures of known threats. Experimental evaluations on benchmark datasets demonstrate the system’s robustness, achieving superior performance in terms of accuracy 97.1 , precision, recall , and F1 score compared to standalone anomaly or misuse detection methods. This study highlights the potential of SVM in hybrid intrusion detection frameworks to provide a comprehensive and scalable solution for modern network security challenges. Future work will focus on incorporating adaptive learning to enhance the system’s ability to detect evolving attack vectors.

Smart Pest Control System: Deep Learning Algorithms for Pest Detection and Pesticide Selection

Authors- Maheswarreddy B, Madhurajavardhan Reddy T, Venkata Lakshmi Reddy A, Mrs.Kohila R

Abstract-Agriculture is an important basic industry worldwide, and pests can cause huge losses to crop production in every country. According to research, nearly half of global crop production will be impacted to varying degrees due to pests every year, which seriously affects the regional economy and people’s daily lives. Crop pests are a major agricultural problem worldwide because the severity and extent of their occurrence threaten crop yield. However, traditional pest image segmentation methods are limited, ineffective and time-consuming, which causes difficulty in their promotion and application. Deep learning methods have become the main methods to address the technical challenges related to pest recognition. When pests infect a field, they must be recognized in time; therefore, farmers can deliver timely treatment and avoid the spread of pests. Still, a traditional pest recognition schemes have numerous limitations. Most of the generally used techniques are manual study. Here the farmers or experts manually check the field monthly, daily, and weekly for pest or diseases. We propose an improved deep convolution neural network to better recognize crop pests in a real agricultural environment. The proposed network includes parallel attention mechanism module and residual blocks, and it has significant advantages in terms of accuracy and real-time performance compared with other models. And also extend the framework to predict the fertilizers or pesticides based on identified pest.

Nutrigenius System Intelligent Food Classifier for Personalized Nutrition Analysis

Authors- Vamsi Krishna J, Sanjay Dath P, Vinod Babu P, Mrs. Kohila R

Abstract-The people across the universe are becoming more sensitive to their diet. Unbalanced diet can cause many problems like weight gain, obesity, sugar, etc. So different systems were developed so as to analyse food images to calculate calorie, nutrition level etc. Food is one of the most important requirements of every living being on earth. The human beings require their food to be fresh, pure and of standard quality. The standards imposed and automation carried out in food processing industry takes care of food quality. Over the last few years, the food-as-medicine concept has gained momentum- thanks to physicians’ and practitioners’ awareness about making food an integral part of treatment for chronic illnesses along with their medications. Measuring the food is very important for a successful healthy diet. Measuring calorie and nutrition in daily food is one of the challenging methods. Smartphone plays a vital role in today’s technological world using this technique which enhances the issue in intake of dietary consumption. In this project the food image recognition system for measuring the calorie and nutrition values have been developed. The user has to take the picture of the food image, this system will classify the image to detect the type of food and portion size and the recognition information will estimate the number of calories in the food. In this system the food area, size and volume are used to calculate the calories and nutrition in accurate way.

Skills Students Develop When They Engage in a Cross Disciplinary Approach of Scientific Knowledge and the Role of Digital and Blended Learning Environments

Authors- Asimakopoulos Z., Professor Smyrnaiou Z.

Abstract-Modern science curricula in secondary education are oriented towards the cultivation of knowledge and skills that are linked to the context of everyday life, aiming to connect science with society, the environment and culture. However, it is observed that despite the contribution and implementation of modern pedagogical frameworks and the extensive use of digital technologies in the learning process, the results in terms of the diffusion of science in society and the choice of scientific careers are not as expected. In this research we attempt to study in which way the application of a Cross disciplinary approach to science teaching with the contribution of modern pedagogical frameworks as well as the support of digital and blended learning environments can enhance students’ interest and contribute to the development of skills that can lead to the construction of scientific knowledge.

Natural Language Query System for Excel Data

Authors- Dinesh Kumar G, Ajith T, Gokulnath G, Jainul Abedeen A

Abstract-In today’s data-centric environment, accessing and analyzing information efficiently is crucial for effective decision-making. However, traditional spreadsheet tools like Microsoft Excel often require users to possess technical skills in formulas, functions, and data manipulation techniques. This paper proposes a Natural Language Query (NLQ) System for Excel Data that enables users to interact with spreadsheet data using plain English queries. The system leverages Natural Language Processing (NLP) techniques to interpret user queries, identify relevant data elements, and convert them into structured operations executable on Excel files. By bridging the gap between natural language and spreadsheet functionalities, this system empowers non-technical users to perform data filtering, aggregation, and analysis without prior programming or Excel formula knowledge. The proposed system enhances data accessibility, reduces dependency on technical expertise, and promotes user-friendly interaction with tabular data.

DOI: /10.61463/ijset.vol.13.issue2.261

Multi-Modal AI Evaluator: A Unified Framework for Evaluating Instruction-Following across Text, Image, and Audio Modalities

Authors- Guhaa Priyan GP, Methun R, Prasanna SN, Jeevak A

Abstract-Artificial Intelligence has really come a long way in handling all sorts of content, whether it’s text, images, or audio. However, in our everyday lives, we often need to make sense of and reason through these different types all at the same time. This paper presents the Multi-Modal AI Evaluator, a framework designed to assess how effectively AI models can follow instructions that involve various input types. The system boasts a user-friendly web interface created with Streamlit, allowing users to input text, images, and audio. It integrates the Google Speech-to-Text API for transcription, GPT-2 for language generation, and a human-in-the-loop feedback system to ensure a comprehensive evaluation of how well an AI model produces responses. The evaluator provides real-time scoring based on accuracy, contextual relevance, and user feedback. This work marks a significant advancement toward establishing standardized evaluations for multi-modal AI systems and fostering the development of artificial intelligence that better meets human needs.

DOI: /10.61463/ijset.vol.13.issue2.262

Crop Yield Prediction Using Machine Learning

Authors- Bhima Yojitha, K.Ram Charan Teja Reddy, Chiranjeevi Althi, Kunda Venkata Chandrika, Paulraj Amulya

Abstract-This paper presents a machine learning-based system for predicting crop yield by integrating historical agricultural data, weather patterns, and soil properties. The model applies regression techniques, decision trees, and deep learning to assess key factors such as temperature, rainfall, soil nutrients, and farming methods. By utilizing real-time data processing and predictive analytics, the system provides valuable insights to assist farmers and policymakers in resource optimization and sustainable agricultural management. This approach helps reduce uncertainties, strengthen food security, and improve overall farm productivity.

DOI: /10.61463/ijset.vol.13.issue2.263

Gender and Climate Change Adaptation among Smallholder Farmers in Nigeria

Authors- Abba Bala Ibrahim, Shamsuddeen Bawale, Zaitun Sunusi Bakabe

Abstract-This paper investigate the intersections of gender and climate change adaptation among smallholder farmers in Nigeria taking a case study from Garin Itace Viillag of Yobe State Damaturu local government ,the study focuses on how gender roles influence adaptation strategies, the challengess faced by female farmers in adapting to climate change. The study Utilizes a mixed-methods approach and integrates qualitative data from interviews and focus group discussions with quantitative data derived from surveys conducted across various farming communities in Garin Itace Vilage of Yobe State Nigeria. Findings from the study reveal that traditional gender roles significantly alter adaptation strategies, with male farmers more likely to access resources such as improved seeds, fertilizers, and extension services, while female farmers depend on indigenous knowledge systems and social networks. Moreover, the study also highlight that female smallholder farmers face disproportionate limitations to accessing climate adaptation resources, including limited land ownership, financial constraints, inadequate access to information, and value system. The study also indicated that the ineffectiveness of current policies in addressing these gender disparities, and emphasize the need for inclusive policies that recognize the unique vulnerabilities of women and enhance their adaptive capacities. Recommendations made from the study include improving access to resources, promoting gender-responsive agricultural policies, and integrating gender considerations into climate change adaptation frameworks. The research made a profound contribution to the broader discourse and debate on gender and climate change by providing empirical evidences and localized insights on the gender-specific impacts of climate change and proposed actionable solutions to enhance resilience among smallholder farmers in Yobe State Nigeria.

DOI: /10.61463/ijset.vol.13.issue2.263

Electric Bicycle Design with BLDC Motor Drive: A Practical Approach

Authors- Vikas Gore, Babu Satwekar, Prasad Wable, Pratik Gadade

Abstract-The world is witnessing a surge in personal car usage. cars are increasingly polluting urban areas, releasing large amounts of carbon dioxide and other climate-altering greenhouse gases into the atmosphere, while also consuming vast quantities of petroleum. Alarmingly, the rate of automobile usage is growing much faster than the human population. if current trends continue, it is estimated that over 3 billion vehicles will be in operation by 2050.additionally, the price of petroleum continues to rise, prompting people to seek alternative sources of energy to power vehicles particularly electricity. in this context, bicycles have transformed from being seen as old-fashioned recreational items to becoming less-polluting, compact, and ultra-light personal mobility tools. electric bicycles, or e-bikes, are emerging as a sustainable pillar of urban transportation in major cities around the world. an e-bike is essentially a traditional bicycle equipped with an electric motor to assist with propulsion. it is an eco-friendly and urban-friendly mode of transportation, powered by a rechargeable battery. e-bikes first gained popularity in china, where in 1991 the government identified their development as a key technological goal (hongyong, 2016). by 2016, china had become the world’s largest manufacturer and consumer of e-bikes. the increased demand in china and many other asian countries has largely been driven by economic growth and rising household incomes. in recent decades, sales have grown at a rate of approximately 10% per annum (weiss et al., 2015). currently, around 90% of the world’s e-bikes are sold in china, with an estimated 170 million users commuting daily.

DOI: /10.61463/ijset.vol.13.issue2.264

AI-Driven Predictive Maintenance: Deep Learning for Mechanical Parts Life Estimation and Health Monitoring

Authors- Mrs. M. Mani Deepika., M.Yaswanth., K. Abhishek, CH.Rohan., J.Sanjana., A. Pavani

Abstract-To enhance the accuracy of predictions and enable real-time monitoring of mechanical parts, a deep learning-based approach was developed. First, a Convolutional neural network (CNN) was designed to extract key features from mechanical components. These extracted features were then processed through a fully connected layer for information fusion and classification, enabling precise life prediction and health status monitoring. The trained deep learning model was subsequently integrated into a monitoring system, forming a comprehensive framework for mechanical parts’ life prediction and condition assessment. Continuous optimization and updates were performed to improve prediction accuracy, real-time responsiveness, and adaptability to varying working conditions and environmental factors. Experimental results demonstrated that the model achieved a Mean Absolute Error (MAE) of 2.1, a Root Mean Squared Error (RMSE) of 2.5, and a Mean Absolute Percentage Error (MAPE) of 10%. These findings highlight the model’s excellent performance and its potential to provide significant technical support for engineering applications.

DOI: /10.61463/ijset.vol.13.issue2.265

AI-Powered Loan Risk Assessment: Predicting Risk Levels Using Machine Learning and Customer Profiling

Authors- MMr. Y. Ravi Bhushan., C.Kavya Sri., B.Praharshitha., C.Sriyashaswi., S.V.Tejasri., A.Prudhvi Kedar.

Abstract-Banks are essential to the global financial system, relying on loan interest as a key source of income, but loan defaults can lead to significant losses. To mitigate this risk, machine learning algorithms can efficiently predict the likelihood of default before approving a loan. In this study, six models—Decision Tree, Random Forest, Support Vector Machine (SVM), Multi-layer Perceptron (MLP) Artificial Neural Network, Naive Bayes, and a stacking ensemble model—were trained using a dataset with 20 key factors commonly found in loan applications. The stacking ensemble model achieved the highest accuracy at 78.75%, while the Random Forest model performed similarly with 78.15% accuracy but greater efficiency. Key predictors of credit risk included credit amount, checking account status, customer age, loan duration, and loan purpose. These findings highlight the potential of machine learning models to enhance loan approval decisions and minimize financial risk for banks.

DOI: /10.61463/ijset.vol.13.issue2.266

AI-Driven Phishing Detection: Advanced Machine Learning Strategies for Identifying Malicious Websites

Authors- Mrs. T. Sankaramma, I.V.B. Sai Sandeep, B. Bhuvana Chandrika, K.V.V.S.Sai Manohar, R.Joseph Hosanna., T.Madhuri

Abstract-In recent years, website phishing attacks have significantly increased, posing a major cyber security threat. While various software solutions have been developed to detect phishing websites, they are not entirely effective in identifying all fraudulent activities. Several challenges remain in accurately distinguishing between legitimate and malicious sites. To enhance detection accuracy and computational efficiency, integrating machine learning algorithms into phishing attack detection has emerged as a promising approach. Machine learning not only improves accuracy but also helps address limitations in existing models. This research focuses on training the ENSEMBLE Machine Learning (ML) model using a collected dataset to detect phishing websites effectively.

DOI: /10.61463/ijset.vol.13.issue2.267

Machine Learning-Based Analysis and Prediction of Electric Vehicle Costs

Authors- Mrs. R. Veera Meenakshi., J.S.N.Srinivas Sattwik., K.K.Satya Prasanna., M.Thanmai Sswechitha., P.V.S Raghavendra Nehru, N.Shanmukhi Satya

Abstract-Electric vehicles (EVs) offer significant environmental benefits by reducing emissions, yet their widespread adoption is largely influenced by cost. Machine learning (ML) algorithms provide a promising approach for predicting EV prices. This study aims to evaluate and compare the performance of various well-established ML algorithms to determine the most effective model for price prediction. To identify key factors influencing EV costs, we conducted a literature review analysing the elements that contribute to pricing. We then theoretically assessed different ML algorithms and validated our findings through comparative analysis and simulation results.

DOI: /10.61463/ijset.vol.13.issue2.268

Comparative Study on Partial Replacement Of Waste Foundry Sand As A Fine Aggregate And Granite Waste As A Course Aggregate: A Review

Authors- Dnyaneshwari Kailasrao Kadu, Prof. H.B. Dahake

Abstract-The possibility of granite waste and waste foundry sand (WFS) as sustainable substitutes for fine and coarse aggregates in M40-grade concrete is examined in this study. A by-product of the ferrous and non-ferrous metal casting industries, WFS was added at replacement levels of 5%, 10%, 15%, and 20% by weight of fine aggregate. The highest limit was chosen at 20% since strength decreases beyond this percentage. At the same substitution levels, granite waste a byproduct of the stone processing industry—was also used as a partial substitute for coarse aggregates. To investigate their combined impacts on concrete, two experimental scenarios were examined: one with a constant 20% WFS replacement and different percentages of granite waste, and another with simultaneous fluctuations in both WFS and granite waste content.

DOI: /10.61463/ijset.vol.13.issue2.269

Self-Supervised Deep Learning for Maritime Surveillance: Detecting Intentional AIS Shutdowns in Open Seas

Authors- Mrs.A. Srujana Jyothi., V.Siva.Sai Ram., K.Chandra Mouli., U.T.S.K.Krishnam Naidu., P.Isaac Ben Joseph., N. Likhita Pravallika

Abstract-Maritime traffic surveillance plays a vital role in detecting illegal activities such as unauthorized fishing or the trans shipment of illicit goods, making it a critical responsibility for coastal administrations. In open waters, authorities rely on Automatic Identification System (AIS) messages transmitted by on board transponders, which are captured by surveillance satellites. However, vessels engaged in illegal activities often intentionally disable their AIS transponders to evade detection. Distinguishing between intentional AIS shutdowns and signal loss due to protocol limitations, adverse weather conditions, or satellite positioning constraints remains a significant challenge. This paper introduces a novel approach for identifying abnormal AIS missing receptions using self-supervised deep learning techniques and transformer models. By leveraging historical AIS data, the trained model predicts whether a message should be received in the next minute, flagging anomalies by comparing the predictions with actual outcomes. Our method is designed for real-time processing, capable of handling over 500 million AIS messages per month, covering the movements of more than 60,000 ships. The approach has been tested on a year’s worth of real-world data from four Norwegian surveillance satellites, demonstrating effectiveness by successfully rediscovering previously identified intentional AIS shutdowns based on existing research findings.

DOI: /10.61463/ijset.vol.13.issue2.270

A Comprehensive Review of Machine Learning Techniques for Skin Lesion Detection and Classification

Authors- Mr.N.V.S.Gopalam., D.Gopireddy., R.Rama Naga Ganesh., P.Venkata Satish., R.V.S.Lakshmi Aishwarya., G.Naga Dinesh.

Abstract-This research explores the application of machine learning (ML) in dermatology to enable faster and more accurate identification and classification of skin injuries. Traditional diagnostic methods rely on visual examination, which can be subjective and prone to varying interpretations. ML offers a potential solution by analyzing vast amounts of data and recognizing patterns that enhance diagnostic precision. The study examines current advancements in machine learning, focusing on real-world validation and addressing dataset variability. While deep learning (DL) has shown promising results, the research highlights the advantages of traditional ML techniques in terms of interpretability and processing efficiency. Various approaches for automating skin lesion analysis, including feature engineering, rule-based techniques, and conventional ML algorithms, have been explored. To overcome existing challenges, the study proposes leveraging advanced transfer learning methods, integrating genetic and clinical data, and improving AI explainability. The future of dermatological diagnosis hinges on collaboration between ML experts and dermatologists to develop real-time diagnostic tools that enhance accessibility and accuracy. By combining medical expertise with ML capabilities, this integration has the potential to revolutionize dermatology, offering scalable solutions for rapid lesion diagnosis and improved patient outcomes. The ongoing advancements in automated dermatological diagnostics are expected to pave the way for more personalized treatment approaches, ultimately transforming patient care.

DOI: /10.61463/ijset.vol.13.issue2.271

Micro-Expression Recognition for Lie Detection Using Image Processing

Authors- Dr. Pankaj Malik, Akshit Harsola, Somya Sharma, Srashti chouhan, Ananya Subramanya Rao, Akanksha Bhadouriya

Abstract-Micro-expressions are involuntary facial expressions that reveal a person’s true emotions, often occurring in high-stakes situations such as interrogations and security screenings. Detecting these subtle expressions is crucial for lie detection, as they can indicate concealed emotions. This research proposes a novel micro-expression recognition system using image processing techniques to enhance the accuracy of deception detection. The methodology involves facial feature extraction using Local Binary Patterns (LBP) and Optical Flow analysis, followed by classification using a Convolutional Neural Network (CNN). Experiments were conducted on benchmark micro-expression datasets, including CASME II and SAMM, to evaluate system performance. The proposed model achieved an accuracy of 89.4%, outperforming traditional handcrafted feature-based methods. The results demonstrate that deep learning-based image processing techniques can significantly enhance the recognition of micro-expressions for lie detection applications. The findings suggest potential applications in forensic investigations, security protocols, and psychological analysis. Future work will focus on real-time implementation and improving recognition across diverse demographic groups.

DOI: /10.61463/ijset.vol.13.issue2.272

Efficient AI-Based Career Platform for Personalized Job Recommendations Using NLP and ML

Authors- Adish H, Abdul Akram A, Jeberson nishith A, Ahamed Akhil MN

Abstract-Artificial intelligence (AI) has revolutionized various areas, including recruitment, by optimizing candidate selection, automating screening for curriculum vitae, and improving workplace accuracy. Traditional attitudinal processes often require addressing inefficiencies such as subjective bias, long configuration cycles, and overwhelming application volumes. These challenges necessitate AI-driven solutions that efficiently process large-scale recruitment data and provide accurate job recommendations. Machine Learning Methods to improve resume analysis, job data extraction, and candidate job adaptation. The system uses several AI models such as TFIDF, Word2VEC, and BET bin, and combines collaboration filtering and content-based recommendation algorithms to analyze job openings and curriculum with high accuracy. By using these techniques, SkillSync ensures job seekers receive relevant personalized job recommendations and enable recruiters to efficiently identify the right candidates. Scraping framework and AI control ranking list mechanism. Evaluation metrics are also discussed, including accuracy, recall, F1 scores, and intermediate mutual ranks (MRR), which include evaluations of workplace systems’ performance. Experimental results show that the accuracy of recommendations is significantly improved, surpassing traditional keyword-based and rule-based contract systems. This paper also examines the ethical challenges of AI-driven recruitment, including bias mitigation strategies, while ensuring compliance with data protection regulations such as GDPR. The proposed employment recommendation system aims to enhance recruitment efficiency, reduce setup time, and improve labour market accessibility. models, and explainable AI (XAI) techniques to enhance interpretability and transparency in decision-making. These advancements will further strengthen the role of AI in the evolving landscape of intelligent recruitment technologies.

Stay Safe Security App with Scream Alert Detection System

Authors- Sarang Kale, Pratiksha Bhokare, Vishakha Omar, Yash Gawande Shuddhodhan Ramteke, Ajay Gade

Abstract-The Stay Safe Security App with Scream Alert Detection System is a mobile application designed to enhance personal safety by detecting distress signals through predefined audio patterns. The system recognizes screams and other emergency cues, triggering an automatic alert to emergency contacts with real-time location sharing. This paper explores the technical implementation, accuracy, and real-world applications of the system. The proposed system aims to improve safety measures by providing a hands- free emergency alert mechanism. Testing results demonstrate the effectiveness of predefined audio pattern detection in various environmental conditions.

DOI: /10.61463/ijset.vol.13.issue2.273

Social Smart Voting System Through Fingerprint Sensor

Authors- Rutuja Dharme, Samiksha Ramteke, Rutik Hade, Subodh Sawai, Prof S. A. Meshram

Abstract-The main objective of our project is to introduce a new voting system which uses fingerprint for verification of a voter. Our proposed system can be used to conduct elections at different levels such as that of the Parliament, Panchayat and so on, on the same day which will reduce the cost of conducting elections on different dates. The Smart Voting System is an android application which enables user to vote in his smart phone using fingerprint. This is more advanced compared to the present system because it doesn’t need any man power. And voter doesn’t need to visit the polling booth. The voting can be done from android mobile phone from anywhere. Since voting can be done from anywhere, it will speed up the voting process for the voters as well as government as a result the security in voting can be ensured by preventing fake votes. Our system will provide a more convenient way for voting for people. This project provides the specification and requirements for E-Voting using an Android platform. This technology helps the user to cast the vote without visiting the polling booth. The application follows proper authentication measures in order to avoid fraud voters using the system.

DOI: /10.61463/ijset.vol.13.issue2.274

Smart Voting System Using Machine Learning

Authors- Asst.Prof. Pooja Dongare, Asst.Prof. Nimisha Rai

Abstract-the increasing need for secure, transparent, and efficient electoral processes has led to the adoption of technology-driven solutions. this paper presents a smart voting system using machine learning to enhance election integrity, voter authentication, and fraud detection. the primary objective of this project is to develop a digital smart voting system that eliminates the vulnerabilities of traditional manual voting methods, such as paper-based ballots, which are prone to fraud and multiple voting attempts. the proposed system integrates multi-layered authentication mechanisms to ensure security and reliability. a biometric fingerprint scanner is incorporated to verify voter identity by cross-checking fingerprints against a registered database. once a voter’s identity is confirmed, they can proceed to cast their vote through an interactive panel, selecting their preferred candidate. the final vote is displayed on a liquid crystal display (lcd) to provide transparency and voter satisfaction. additionally, to enhance automation and security, the system sends a warning notification to individuals who have not yet cast their vote. this autonomous and secure voting system aims to improve electoral transparency, eliminate fraud, and enhance voter confidence in the election process.

DOI: /10.61463/ijset.vol.13.issue2.275

Home Automation System Using ESP8266

Authors- Y. Karthik Raj, Harshit Singh, Yogesh Kumar Yadav, Kiran Dewangan

Abstract-This paper presents the design and implementation of a comprehensive home automation system leveraging the ESP8266 microcontroller and the Blynk IoT application. The system integrates five key sensors: a gas sensor, PIR sensor, flame sensor, humidity sensor, and water level sensor. By utilizing these sensors, the proposed system enhances home safety, security, and convenience through real-time monitoring and control via a user- friendly mobile interface.The gas sensor (MQ2) detects hazardous gases, providing early warnings for gas leaks. The PIR sensor enhances security by detecting motion, while the flame sensor offers fire detection capabilities. Finally, the humidity sensor (DHT11) monitors environmental humidity, ensuring comfort and protecting electronic devices.

AI-Powered Scheme Navigator for Tamilnadu Government Schemes

Authors- Rasu A, Sundharesh M, Manoj M, Sanjai R, P. Assistant Professor Arul Selvam. M.E

Abstract-Government schemes encompass diverse areas such as education, healthcare, agriculture, social welfare, and infrastructure. However, a lack of awareness and difficulty in accessing accurate information often prevents individuals from availing of these benefits. To address these challenges, this project proposes the development of a Natural Language Processing (NLP)-based chatbot designed to provide seamless access to information about Tamil Nadu government schemes. The chatbot leverages advanced NLP frameworks, such as spaCy and Hugging Face Transformers, to process and interpret user queries, delivering precise and relevant responses. Comprehensive data on government schemes is collected, preprocessed, and used to train the model. Integrated into a user- friendly interface, the chatbot ensures effortless interaction, allowing users to inquire about various initiatives and obtain real-time information. The system incorporates testing and monitoring mechanisms to ensure accuracy and adaptability to a wide array of user inputs. Regular updates are planned to reflect policy changes and maintain the chatbot’s relevance. Additional features include user authentication for personalized assistance and provisions for human support to handle complex queries. By offering an intelligent conversational interface, this project aims to enhance accessibility and engagement, empowering citizens to make informed decisions and utilize available government resources effectively.

DOI: /10.61463/ijset.vol.13.issue2.276

CNN-Based Image Classification for Medical Diagnosis: A Case Study on X-ray Images

Authors- Sejal Rahul Sanas

Abstract-Medical image analysis has become a critical component in the early detection and treatment of diseases, particularly with the integration of artificial intelligence (AI) techniques. This research focuses on leveraging Convolutional Neural Networks (CNNs) to automate the classification of chest X-ray images for the diagnosis of conditions such as pneumonia, tuberculosis, and lung opacity. CNNs, a powerful class of deep learning models, are capable of recognizing intricate patterns in medical images with high accuracy. In this study, a CNN model was developed and trained on a labeled dataset of chest X-ray images. The methodology encompasses data preprocessing, model architecture design, training, and performance evaluation. Key metrics including accuracy, precision, recall, and F1-score were employed to assess the model’s effectiveness. The results indicate that CNN-based models can substantially enhance diagnostic accuracy and speed, particularly in resource-limited settings where access to expert radiologists may be constrained. The paper also addresses challenges such as limited dataset size, computational demands, and the importance of model interpretability in clinical applications. Overall, this study underscores the potential of deep learning as a supportive tool in healthcare, contributing to timely diagnosis and improved patient outcomes.

Voice Command Based System Controller Using Natural Language Processing

Authors- Divyani R. Sadafale, Prof Aakansha P. Tiwari, Nikita D. Kalbande, Swapnil I. Sirsat, Om P. Bobade

Abstract-This paper introduces a desktop-based intelligent voice assistant capable of controlling Windows operating systems through speech commands. By integrating artificial intelligence (AI) and natural language processing (NLP), the system fosters seamless human-computer interaction. Users can manage system functionalities, including launching software, modifying settings, and handling files, using voice instructions. The architecture leverages a hybrid approach combining deep learning with conventional NLP techniques to ensure high recognition accuracy and responsiveness. The design, implementation, and evaluation of the system are presented, highlighting its potential in accessibility, automation, and improving user productivity.

DOI: /10.61463/ijset.vol.13.issue2.277

Smart Poultry Farm Cooling Management

Authors- Associate Professor Dr.P.Vijayalakshmi, Hariharan J, Jagadeesh S, Abiraj G, Gowtham T

Abstract-A Chicken Incubator Monitoring System develop to complete and conditioned automatic incubator for chickens under defined ranges of temperature, humidity, and feeding schedules. A temperature sensor manages a mechanism responsible for the pumping of water through which the incubator is maintained with steady heat and moisture. With use of humidity sensor, a warming lamp is controlled, which provides the required conditions for egg incubation. Thus, by virtualizing all ambient conditions, the efficiency of the hatching is enhanced along with the entire incubation process. It also constitutes a timer-based food feeding mechanism to supply food to hatched chicks at specific times, thus reducing the burden or effort of manual feeding and helping to ensure the chicks are uniformly fed. This feeding process comprises an automated food distribution system designed to dispense fixed measured quantities of food over predestined time schedules so that the chicks benefit from healthy growth during their early phases considering the particular needs of their growth spurts. Otherwise, there is a feed operation whereby a premium quantity is given according to specified time schedules. Optimize incubator environmental conditions and improve efficiency in poultry farming by having the temperature and humidity control under automated feeding. With the use of IoT monitor and control mechanisms provides distance monitoring and control over incubation settings at the farmer’s convenience, thus minimizing human error and making hatch success rates sing. This system is particularly important for small-scale and large-scale poultry farmers looking into very high productivity and high best practices in livestock management.

DOI: /10.61463/ijset.vol.13.issue2.278

Smart Poultry Farm Cooling Management

Authors- Dr G Rama Subba Reddy, K.V.Sunil, B.Vishnu Deva Reddy, M.V.Siva Sai Kumar, B.Bharath Kumar

Abstract-The judicial system is an important component of the nation’s democratic system, protecting the rights and liberties of its residents while preserving the rule of law. With time, many necessary changes are made in the judiciary system to maintain peace, trust, and order in the country. But court hearings and proceedings take too much time for a decision. Despite the fact that the legal industry is developing faster than ever with the aid of developing technologies, there are still many unexplored areas and there is always potential for improvement. In this paper1, we present a simplified approach, “AI in Law Practice”. The model is developed by utilizing, the most disruptive technology – Machine Learning. The dataset is stored in IPFS for legal and ethical considerations. The algorithm can forecast case results, giving departments and attorneys useful information. Predicting the case output with accuracy is the model’s problem. The F1-score, accuracy, precision, recall, support, and recall are used to evaluate the system’s performance, which also demonstrates the system’s practical applicability. The model is trained on 3304 U.S. Supreme Court cases and achieves 95% accuracy2. The model that is currently being built will be utilized by legal practitioners, law departments, end users, etc.

DOI: /10.61463/ijset.vol.13.issue2.279

IoT-Based Real time Health Monitoring and Alert System Using Arduino

Authors- Dr Y Subba Reddy, Sugumanchi Maheswari, Palakolanu Dharani, Chennuru Naga Sharanya, Lomada Sandeep Kumar Reddy

Abstract-In recent years, the integration of Internet of Things (IoT) technology in the healthcare sector has enabled continuous and remote monitoring of patient health. This project presents the design and implementation of an IoT-based real-time health monitoring and alert system using Arduino, aimed at enhancing patient care and emergency response. The system utilizes sensors such as a heart rate sensor, temperature sensor, and oxygen (SpO₂) sensor to continuously monitor vital signs. These parameters are collected by an Arduino microcontroller and transmitted via a Wi-Fi module (such as the ESP8266) to a cloud server or web dashboard in real-time. If any parameter crosses a critical threshold, the system instantly triggers an alert via SMS, email, or buzzer, ensuring timely medical intervention. The proposed solution is cost-effective, portable, and scalable, making it suitable for home care, hospitals, and remote health centers. It contributes to the growing need for smart healthcare solutions by offering real-time insights and immediate alerts, potentially saving lives through faster response times.

DOI: /10.61463/ijset.vol.13.issue2.280

Industrial Parameters Monitoring and Control Using IoT

Authors- Gollapudi Venkata Ramana Rao, Samudrala Kavya, C Vijaya Lakshmi, Putha Sai Rohith Reddy, Barika Rajeswari

Abstract-This paper presents an IoT-based industrial environment monitoring and control system using Arduino UNO as the master controller and ESP32 for cloud connectivity via ThingSpeak IoT platform. The system integrates sensors to measure temperature, humidity, light intensity, smoke, and fire. Based on predefined thresholds, it activates a cooling fan, LED, buzzer, and CPU fan for automated responses. The LCD displays real-time sensor data, while ESP32 transmits it to the cloud for remote monitoring. This smart system enhances safety and automation, making it suitable for industrial applications. The proposed solution provides real-time alerts and efficient environmental control.

DOI: /10.61463/ijset.vol.13.issue2.281

Investigate the Behaviour of Variously Located Tall Structures with Distinct Seismic Loads

Authors- K. Saroja Rani, N. S.S. Malleswari, K.S.S. Reddy, S. Haranadh Raja, V. N. V. K. Dinesh

Abstract-Now a days with the increase of population in cities, the area available for residence as well as commercial purpose is very congested and people are facing problem in parking, so for parking we left ground story openly known as soft storey. Which doesn’t consist in filled masonry walls to resist lateral forces which are induced on building due to earthquake, while damage and collapse due to soft storey are most often observed in buildings. In the present project, we want to analyses these kinds of buildings for different seismic zones of India for same load conditions. We will do the comparative study of collapse conditions in different zones of India (i.e., Bending moment, Shear force, Lateral displacement and Support reactions), with the help of Staad. Pro V8i software.

DOI: /10.61463/ijset.vol.13.issue2.282

Study of Behaviour of Wind Evaluation of Multi-Storey Building With and Without Floating Columns Structures in Staad.Pro

Authors- L. Praveen Kumar, M. Ajay Kumar, V. Saranya Sahithi, K.S.S. Karthik, Y. Jayakanth

Abstract-– Tall buildings often face space constraints due to the current surge in urbanization. The structure can be affected by wind gusts in both directions. These gusts have the potential to impact the structure from both directions. Over the past few years, the structure has experienced effects from these gusts in both directions. These designs aim to enhance the visual perspective of the projects they undertake. The variability in floor height causes a discontinuity in the stiffness of the structure at the level of the soft story. This phenomenon is caused by floor height fluctuations. In the even If winds expose this discontinuity, it could potentially cause buildings to This study aimed to perform a static analysis of three-dimensional building frames, which included G+7 storeys, floating columns, and soft storey elements. elements. The other sixty-four examples feature floating columns at a single level, with the soft storey varying directly from the ground (G) story to the G+7 storey. Eight of the instances include centre floating columns on any one of the storeys, while sixty-four of the other cases have floating columns at a certain level. This instance considers a total of seventy-three instances. Furthermore, we construct a simple example where neither the storeys nor any of the column’s float, adhering to the previously stated conditions. We conducted the analysis using the maximum node displacements (resultant), maximum moments, maximum shear force, maximum axial force, and maximum storey drift. It is necessary to do an analysis of the findings in order to arrive at technical conclusions.

DOI: /10.61463/ijset.vol.13.issue2.283

Planning, Design and Structural Behaviour of Water Tank Using Etabs

Authors- V. Tanuja, G. Gowthami Sri, V. Priya, K. Teja Swaroop , U. Himani Josh

Abstract-– Circular overhead water tanks are large storage containers constructed for storing water supply at certain height to pressurize the system of water distribution. It comprises of a heavy water mass at the top of a slender staging which is utmost critical parameter consideration for the collapse of the tank. A detailed understanding of the performance of the structures under dynamic forces is necessary to meet the safety objectives during construction and maintenance. Other modes of failures considered are sloshing damage at roof, buckling, inlet or outlet pipe breaks. From previous studies it was clear that inadequately designed tanks were damaged heavily due to improper design. This may be due to the lack of knowledge regarding the behavior of supporting system of the tank, and also due to improper selection of geometry of staging patterns. From the study it is concluded that the primary mode shape of circular tank is torsion which needs to be eliminated. To eliminate the torsion mode shape in circular elevated water tank, orientations of columns are modified. Circular water tanks are constructed in order to provide required head so that the water will flow under the influence of gravity the construction practice of water tanks is as old as civilized man. The water tanks project have a great priority as it serves drinking water for huge population from major metropolitan cities to the small population living in towns and villages.

DOI: /10.61463/ijset.vol.13.issue2.284

Analysis and Design of Different Framed Structures G+30 Tall Building Under Dynamic Loading Behaviour

Authors- V. Tanuja, P. Sri Naga Devi, Y. Rakesh, J. Chandu, J. Likith Satya Sai

Abstract-– The increasing demand for high-rise buildings has led to the necessity of designing structures that can efficiently withstand dynamic loads such as wind and seismic forces. This study focuses on the analysis and design of different framed structural systems for a G+30 tall building under dynamic loading behavior. The primary objective is to compare the performance of various framing systems, including moment-resisting frames, braced frames, and shear wall – frame systems, to determine the most effective structural configuration in terms of stability, strength, and serviceability. The analysis is carried out using finite element software such as ETABS or STAAD.Pro, following relevant design standards like IS 875 (Wind Loads), IS 1893 (Seismic Loads), and IS 456 (Concrete Design Code).The study employs Response Spectrum Analysis (RSA) and Time History Analysis (THA) to assess seismic behavior, while wind effects are evaluated based on terrain categories and wind speed variations. Key parameters such as story drift, base shear, fundamental time period, lateral displacement, and inter-story drift are examined to compare the performance of different structural systems.

DOI: /10.61463/ijset.vol.13.issue2.285

A Structural Investigation on Multi-Storeyed Structures with Dynamic Performance in a Seismic Zone Using Different Bracings

Authors- B. Sri Datta Subrahmanyam, K. Harshasri, Ch. Meghana, Ch. SVV Ram Kishore, K. Manikanta

Abstract-– Buildings often provide individuals’ shade. Construction of large structures is necessary because most people prefer solitude. Tall buildings have the drawback of being less earthquake-resistant. Therefore, tall buildings constructed in seismically active areas may sustain serious damage or even die. Therefore, a building’s construction must be capable of withstanding the lateral and gravitational forces generated by earthquakes. Multi-story buildings employ a specific method to withstand lateral loads. The storey’s RC-framed construction used three distinct bracing methods. We have examined and assessed the G+14 stored RC frame design for seismic zone IV. We use computer-aided software, STAAD Pro V8i, to analyse the RC-framed models using the Response Spectrum Method. We examined the structural behaviour using a range of bracing methods, including X-bracing, inverted-V bracing, and V-bracing. We placed these techniques at different positions on the outer faces and all four sides of the constructions. We examine the parameters of time, base shear, storey drift, storey displacement, bending moment, and peak storey shear for both braced and unbraced models. The braced frame’s base shear value increases compared to an unbraced frame, while its storey displacement, storey drift, bending moment, and duration decrease. The braced frames have higher peak story shear values. The model with X-bracing on the mid-bays of the buildings’ exterior sides significantly increases the structure’s stiffness when compared to the alternative bracing technique.

DOI: /10.61463/ijset.vol.13.issue2.286

A Comparative Analytical Study of Multi-Story RCC, Steel and Composite Building Cost and Seismic Response

Authors- L Praveen Kumar, C Surya Teja, S.V Chakra Raju, B. Greshma, G Dayanidhi

Abstract-– Engineers are familiar with the troubles that arise even as developing metallic or concrete structures, considering every fabric has its own set of traits. Because metallic components are often made up of thin plate factors, they are prone to buckling both locally and laterally. As a end result, they’re examined for buckling and instability screw ups, at the same time as concrete contributors are usually thick and tough to buckle, but they’re susceptible to creep and shrinkage over time. As a result, a steel-concrete composite construction has been industrialised to take advantage of each substance. Steel-concrete composite systems are the maximum value-effective solution to the numerous technical layout requirements for stiffness and electricity, combining the incredible traits of each metal and urban with lower costs, faster creation, and fire safety, amongst other blessings. In a number of locations. This kind of production has become a well-known issue in multi-tale metallic frame structures. A bare metallic frame with commonplace H-type section columns supports I-kind section beams, which in turn guide the overlying composite ground slab in the simplest form of composite structures. The composite floor slab, alternatively, is made of bloodless-shaped profiled steel sheets that serve as both the everlasting formwork and the vital tensile reinforcement for an in situ solid concrete slab. This powerful structural technique is suitable for systems that should face up to seismic forces. In this study, ETABS 2015 version 15.2.2 incorporated building layout software program was used to simulate all three sorts of systems defined above, namely steel, concrete, and composite multi-tale buildings. The Static seismic coefficient method and Dynamic Response spectrum evaluation technique are used to examine all three types of systems.

DOI: /10.61463/ijset.vol.13.issue2.287

Detection of Drivers behavior using Real time drowsiness classification in Machine Learning Methods

Authors- Mr J Jagadeswara Reddy, Swathi Devarangula, Harsha Priya, Veera Raja Mohan Reddy, Birru Madhu

Abstract-– Real-time drowsiness classification, aiming to improve the detection of drowsiness in environments where safety is critical, such as in driving or machine operation. The model processes various physiological and behavioral signals, including eye movements, facial expressions, and heart rate, by representing them as nodes in a graph. The edges between nodes capture the inter dependencies among these signals, allowing the model to better understand the complex patterns associated with drowsiness. By incorporating connectivity-awareness, the CAGNN enhances the detection process by prioritizing key sensor interactions and adapting to changing input conditions. This approach enables more accurate classification, even in the presence of noisy or incomplete sensor data, which is a challenge for traditional methods. By leveraging the relationships between multiple signals, the CAGNN is able to provide more robust and dynamic real-time drowsiness assessments. Experimental results show that the CAGNN outperforms traditional methods in both accuracy and computational efficiency. The model is designed to deliver low-latency predictions, ensuring fast and reliable real-time performance. This approach provides an effective solution for drowsiness detection, improving safety and reducing risks associated with drowsiness in various operational settings.

DOI: /10.61463/ijset.vol.13.issue2.288

Multiple Disease Prediction Using Machine Learning

Authors- Associate Professor Dr.R.Jayaraj, Kirubkaran A

Abstract-– The global burden of disease continues to be a major contributor to disability and death, largely due to the challenges posed by disease prediction. This is primarily because of the overlapping nature of symptoms between different diseases, making accurate and timely diagnosis difficult. Current healthcare systems often focus on predicting and diagnosing a single disease at a time, which limits their scope and ability to identify multiple conditions concurrently. This work aims to design and develop an intelligent multiplatform software system capable of simultaneously predicting various diseases, utilizing modern machine learning techniques such as Convolutional Neural Networks (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) networks. The integration of Binary Equivalent Simplification Optimization (BESO) in the system ensures that the model is both computationally efficient and highly accurate. The approach involves several key stages, including data collection and preprocessing, feature extraction through CNN, sequence modeling using BiLSTM, and optimization through BESO. CNNs are employed to capture critical features from the input data, which are then processed by BiLSTM to account for the temporal dependencies in disease progression. This system is expected to significantly reduce intervention costs, improve diagnostic accuracy, and ultimately enhance patient quality of life by enabling early, simultaneous detection of multiple diseases. By enabling the simultaneous prediction of multiple diseases, healthcare providers can more accurately identify at-risk patients, reduce the time spent on diagnosis, and optimize treatment plans. This system can also be particularly beneficial in resource-limited settings, where quick, accurate predictions are crucial but the availability of specialist care may be restricted. Moreover, as the software is designed to be a web-based application, it ensures wide accessibility and ease of use for both healthcare professionals and patients, making it a versatile solution for modern healthcare challenges. The long-term goal is to scale the system, incorporating additional diseases and datasets, and continuously refine its algorithms to enhance predictive power and healthcare outcomes on a global scaleelection process.

DOI: /10.61463/ijset.vol.13.issue2.289

Green Synthesis of Metallic Nanoparticles: The Multifaceted Role of Plant-Based Biological Entities

Authors- Utibe B. Orok, Eno A. Moses, Emmanuel I. Uwah, Godwin N. Enin, Dele P. Fapojuwo, Oluwatosin Audu, Archibong S. Eric, Nathaniel S. Essien, Nnamso D. Ibuotenang, Idongesit B. Anweting, Edidiong L. Udokang, Tijesuni J. Adeoye, Solomon E. Shaibu

Abstract-– The rapidly evolving field of nanotechnology is witnessing a shift towards environmentally friendly synthesis of nanomaterials. This review examines the role of plant-based biological entities in the green synthesis of metallic nanoparticles (MNPs). Plant components such as phytochemicals, enzymes, proteins, and nucleic acids offer a wide range of biomolecules that enable the eco-friendly synthesis of MNPs. The study aims to provide an in-depth understanding of various plant-derived biological entities and their multifaceted roles in MNP synthesis. It specifically explores their reducing and stabilizing functions and their influence on nanoparticle properties. The findings indicate that plant-derived biological moieties provide a sustainable, cost-effective, and eco-friendly method for nanoparticle synthesis. Their innate capabilities allow for both reduction and stabilization of MNPs in a streamlined one-pot synthesis process. These unique bioactive components facilitate the production of nanoparticles with distinct physicochemical characteristics, promoting technological and medical advancements. Future research focusing on the biochemical pathways of these entities will pave the way for more customized and efficient nanoparticle production.

Design of Mechanical Grab for Lifting Plates

Authors- Mr. B. Hari Krishna1, Uppala Pavan Sai2, Shaik Habibul Rehman3, Yarra Durga Venkata Krishna4, Thota Dhanush

Abstract-– This report will provide valuable information to a mechanical engineer to build up an engineering approach in designing a purposeful motorized grab. Through reading of this report, an individual can know how to precede a project of designing the grab and which parameters should be of prime focus in research and development phase these motorized mechanical grabs can be classified into different types. Motors with electromagnetic brakes are used. This is an important safety feature that prevents the unit from accidentally opening under load. To improve the efficiency, safety and quality of material handling functions, Mechanical Grabs can be used in the heavy industry where material handling is quite frequent as it not only ensures safety but also allows consumer to save time and costs. The Automated Telescopic action in the mechanical grab not only provides control to the user but also provides ease for lifting sheets of variable sizes resulting in optimized work flows.

DOI: /10.61463/ijset.vol.13.issue2.290

3d Modelling and Simulation of Spindle Using Static Structural Analysis

Authors- Mr. M. V. V. Siva Krishna, Pediredla Venkatesh, Nellimarla Sai Rajesh, Cheerla Balaji Srinivasu, Thota Krupa Jyothi

Abstract-– This report will provide valuable information to a mechanical engineer to build up an engineering approach in designing a purposeful motorized grab. Through reading of this report, an individual can know how to precede a project of designing the grab and which parameters should be of prime focus in research and development phase these motorized mechanical grabs can be classified into different types. Motors with electromagnetic brakes are used. This is an important safety feature that prevents the unit from accidentally opening under load. To improve the efficiency, safety and quality of material handling functions, Mechanical Grabs can be used in the heavy industry where material handling is quite frequent as it not only ensures safety but also allows consumer to save time and costs. The Automated Telescopic action in the mechanical grab not only provides control to the user but also provides ease for lifting sheets of variable sizes resulting in optimized work flows.

DOI: /10.61463/ijset.vol.13.issue2.291

Condition Design and Material Optimization of Axial Gas Turbine Under Static Loading

Authors- Mrs. B. Anusha Srikanta, Markanda Raghaveandra, Matta Sri Ammayya, Tirukkovalluri Sri Jaiyanth Kowshik, Yedidha Chaitanya, Challa Sai Durga Kishore

Abstract-– Micro turbines are becoming widely used for combined power generation and heat applications. Their size varies from small scale units like models crafts to heavy supply like power supply to hundreds of households. Micro turbines have many advantages over piston generators such as low emissions less moving parts, accepts commercial fuels. Gas turbine cycle and operation of micro turbine was studied and reported. Brief description on CAD software and CATIA studied and reported. Different parts (Inlet. Storage, Nozzle, Rotor, coupling, outlet, clip, housing) of turbine are designed with the help of CATIA (Computer Aided Three Dimensional Interactive Analysis) software. Then they were assembled to a single unit and coupled to a generator to produce power. The turbine is of Axial input and axial output type.

DOI: /10.61463/ijset.vol.13.issue2.292

Design and Optimization of A High Speed Vehicle Using Fluent Tool / CFT

Authors- Dr. G. Avinash1, Pampana Devi Veera Venkata Ramakishore2, Bade Venkata Shiva Karthik3, Mohammad Abubakar Siddiq Khan4, Karri Ranbabu

Abstract-– Aerodynamic characteristics of a racing car are of significant interest in reducing car-racing accidents due to wind loading and in reducing the fuel consumption. Sports cars are most commonly seen with spoilers, such as Ford Mustang, Subaru Impreza, and Chevrolet Corvette. Even though these vehicles typically have a more rigid chassis and a stiffer suspension to aid in high-speed manoeuvrability, a spoiler can still be beneficial. One of the design goals of a spoiler is to reduce drag and increase fuel efficiency. Many vehicles have a fairly steep downward angle going from the rear edge of the roof down to the trunk or tail of the car. Air flowing across the roof tumbles over this edge at higher speeds, causing flow separation. The flow of air becomes turbulent and a low pressure zone is created, thus increases drag. Adding a spoiler at the very rear of the vehicle makes the air slice longer, gentler slope from the roof to the spoiler, which helps to reduce the flow separation. Reducing flow separation decreases drag, which increases fuel economy; it also helps keep the rear window clear because the air flows smoothly through the rear window. The limitations of conventional wind tunnel experiment and rapid developments in computer hardware, considerable efforts have been invested in the last decade to study vehicle aerodynamics computationally. This thesis will present a numerical simulation of flow around racing car with spoiler positioned at the rear end using commercial fluid dynamic software ANSYS FLUENT.

DOI: /10.61463/ijset.vol.13.issue2.293

Energy Recovery From Boiler Blowdown Water at Edible Oil Refine

Authors- Mr. K. Anil Kumar, Kapavarapu Uma Mahesh, Abbireddy Sai, Gajji Uma Maheswara Rao, Jayavarapu Naga Rajesh

Abstract-– Blow down water is the part of water that is purposely drained during the boiler operation to limit the level of impurities in boiler water to an acceptable level. So it is contains large quantity of heat energy. The aim of the present work is to improve energy efficiency of steam boilers in the oil refinery company. This aim has been achieved through designed and manufactured of a heat exchanger consists of a shell and tube unit to recover heat from surface blow down water and reducing indirect losses. The blow down water (hot fluid) is supplied to the heat exchanger at atmospheric pressure by passing it through the tube side and the feed water in the Shell side. These were done as counter flow. A flow control valve is used to control the flow rate of hot blow down water inside the heat exchanger. The experiments are done at the blow down water and feed water flow rates ranging between (0.06-0.14) m3 /h with 0.02 m3 /h interval, and between (0.1-0.5) m3 /h with 0.1m3 /h interval, respectively.

DOI: /10.61463/ijset.vol.13.issue2.294

Design and Static Structural Analysis of Sapcecraft Motor Head

Authors- Mr. N. Raghuveer, Kadali Sanjay Kumar, Dasari Rambabu, Illa Divya Prakash, Chappidi Sarath Kumar

Abstract-– Pyrogen igniters are small rocket motors used to ignite large solid rocket motors (SRMs). The pyrogen igniter is integrated with the SRM, with its head-end fastened using M12 fasteners to ensure a leak-proof joint, sealed with O-rings to withstand high-temperature gases. The primary function of the pyrogen igniter head-end is to endure the igniter’s peak pressure for approximately one second and the rocket motor pressure for nearly 90 seconds. The head-end is fabricated using 15CDV6 steel, a high-strength alloy suited for extreme conditions. During manufacturing, it is subjected to 1.2 times the rocket motor’s Maximum Expected Operating Pressure (MEOP), equivalent to 88 ksc. To monitor strain levels, three strain gauges are strategically positioned at critical areas on the flange.

DOI: /10.61463/ijset.vol.13.issue2.295

Process Parameter Optimization of FDM Process Using DOE Methods

Authors- Mrs. U. Chaitanya Vardhini, Lella Lakshmi Satyanarayana, Guree Ashok, Paka Madhu Manohar, Pragada Jagadeeswarudu

Abstract-– Fused Deposition Modelling (FDM) is widely used for fabricating functional and prototype parts due to its cost-effectiveness and material versatility. However, optimizing process parameters is crucial to improving printing efficiency, material consumption, mechanical properties, and surface quality. This study employs Design of Experiments (DOE) methods to investigate the impact of infill pattern (concentric, quarter cubic, cubic subdivision), infill density (30%, 60%, 90%), and top & bottom closing layers (2, 4, 6 layers) on the overall performance of FDM-printed parts.A series of experiments were conducted using a statistical DOE approach, where the selected parameters were systematically varied to evaluate their effects on printing time, material consumption, surface roughness, and mechanical strength. The infill pattern plays a significant role in internal structure formation, influencing load distribution and energy absorption. The infill density directly impacts part strength and weight, while the top and bottom layers affect surface smoothness and structural integrity. The results indicate that cubic subdivision infill at higher densities (60%-90%) enhances mechanical strength but significantly increases printing time and material usage. The quarter cubic infill provides a balance between strength and weight reduction, making it suitable for lightweight applications. The concentric pattern demonstrated poor mechanical properties due to its lack of uniform load distribution. Additionally, increasing the number of top and bottom layers improved surface quality but also contributed to higher material consumption and extended print time.

DOI: /10.61463/ijset.vol.13.issue2.296

Influence and Optimization of Infill Pattern on Printing Time and Material Consumption Using Stastical Software

Authors- Ms. G. B. T. S. Alekhya, Gonnabathula Naga Mani Ramya, Komarada Anil Kumar, Allu Gayathri Devi, Mohammed Abdul Khadar Gilani

Abstract-– In Material Extrusion (MEX) additive manufacturing, the infill pattern significantly affects printing time, material consumption, and mechanical properties of the printed parts. This study investigates the impact of different infill patterns on these key performance indicators using statistical optimization techniques. The research aims to identify an optimal infill configuration that balances strength, weight, and print efficiency while minimizing material waste. Various infill patterns such as rectilinear, honeycomb, gyroid, and concentric were analyzed, and their effects on printing time and material consumption were evaluated using Design of Experiments (DOE) and Taguchi optimization methods. The study also examines the role of infill density and pattern complexity in determining the overall build efficiency. Results indicate that complex lattice structures (e.g., gyroid) tend to provide better strength-to-weight ratios but increase printing time and material usage. Conversely, simpler patterns like rectilinear offer faster prints with reduced material consumption but may compromise mechanical integrity.Additionally, the porosity between layers due to infill selection was analyzed. process.

DOI: /10.61463/ijset.vol.13.issue2.297

Design and Static Structural Analysis of Rocket Stand

Authors- Mr. Ch. Satya Prakash, Danthuluri Ashish Varma, Donam Kavya Sri, Gubbala Ravi Sai Kiran, Mangina Tharun

Abstract-– The aim of this project is to design a Solid Rocket Motor (SRM) test stand using advanced computer-aided design (CAD) software, SolidWorks. Rocket launching is an intricate engineering challenge due to the complexity of the propulsion system, where the malfunction of even a single component can lead to catastrophic failure of the entire mission. To mitigate such risks, the design and testing of rocket motors are critical processes that contribute significantly to the success of a launch. The SRM test stand plays a vital role in this context by providing a controlled environment for developing, characterizing, and testing solid rocket engines.

DOI: /10.61463/ijset.vol.13.issue2.298

Enhanced Intrusion Detection System Based on Deep Learning

Authors- Saddam Hussain, Prof. Santosh Nagar, Prof. Anurag Shrivastava

Abstract-– With the rapid growth of digital networks, cyber security threats have become increasingly sophisticated, making traditional Intrusion Detection Systems (IDS) less effective. To address this challenge, we propose an Enhanced Intrusion Detection System (IDS) based on deep learning techniques, significantly improving threat detection accuracy and response efficiency. The proposed system leverages Deep Neural Networks (DNNs), Convolutional Neural Networks (CNNs), and Long Short-Term Memory (LSTM) to analyze network traffic and detect malicious activities in real time. By integrating feature selection, the model optimizes classification performance, reducing false positives while maintaining high detection rates. The system is trained and evaluated on the NSL-KDD dataset, achieving 97.6% accuracy, 97.2% precision, 96.7% recall, and an F1-score of 96.9%, demonstrating its superior performance in identifying cyber threats. These results highlight the potential of deep learning in enhancing network security by providing intelligent, real-time IDS capable of effectively mitigating emerging cyber threats. The proposed approach sets a strong foundation for developing highly accurate and adaptive intrusion detection systems in modern cyber security environments.

DOI: /10.61463/ijset.vol.13.issue2.299

Design of Hydraulic Power Pack for Vertical Turning Lathe Machine

Authors- Mr. P. Chinna, Bellapu Vasu, Pothula Venkateswara Rao, Dimmala Siddardha, Palla Hemanth Kumar

Abstract-– The project is aimed at the Design of hydraulic Power pack for vertical turning lathe. This Hydraulic Power pack is used to obtain the various motions of the Vertical Turning Lathe Clamping and unclamping is also done with hydraulic system. Hence versatility and reliability of hydraulics is prime importance. The Power pack is an integral supply unit usually containing a pump, reservoir, relief valve and direction control valve, Pressure control valve. For the purpose of design of hydraulic Power pack, the component is to be designed are pump, reservoir, heat exchanger and an electric motor. Hydraulic drives and controls have become more important due to automation and mechanization. Many of the modern and Powerful machinery are controlled partly or completely by hydraulics. Hydraulic system is less complicated and has fewer moving parts. Today drive and control system engineering is inconceivable without hydraulics. Special emphasis is made on design of Power pack in which the elements, maintenance aspects and trouble-shooting methods is dealt with.

DOI: /10.61463/ijset.vol.13.issue2.300

3d Modelling and Simulation of Industrial Robot Arm Under Static Loading Condition

Authors- Mr. M. Anil Kumar, Lokareddy Durga Sri Venkata Sai, Sana Vinay Kumar, Mattaparthi Kishore, Sattineedi Pavan Naga Venkata Sai Kumar

Abstract-– Computer Aided Design (CAD) and Computer Aided Engineering (CAE) is the most important and essential tool in product development process. Huge challenge is faced by the companies while integrating CAD and CAE in their design process. The previous studies do not clearly give the impact of CAD and CAE on product development process and particularly its impact on cost and time of development. The study is carried out to show the importance of CAD and CAE as a tool of product development and its effect on the development cost and time when implemented early in the process. Computer -aided engineering (CAE) for systems analysis is needed to address conceptual and preliminary design for a broad class of products. System analysis encompasses many engineering disciplines and the associated CAE technology will play a key role in product development and design. Improved products can be designed by employing easy-to-use model building tools and sophisticated mathematical algorithms, which accurately predict performance and cost of future products. Lessons learned from CAD/CAM, particularly in the graphics and database management areas, combined with recent algorithm research, and supported with sufficiently powerful hardware will enable CAE systems to achieve marked productivity improvements.

DOI: /10.61463/ijset.vol.13.issue2.301

Characterisation of Bio-Diesel Blends for Combustion in Swirl Burner

Authors- Ademola Samuel Akinwonmi, Moses Oluwatobi Fajobi

Abstract-– This study utilized commercially available diesel, palm kernel and groundnut oils to produce biodiesel. Raw palm kernel oil (PKO) and groundnut oil (GO) were purchased from the local producers. Each of them was blended with conventional diesel in the ratios 1:1, 1:2, 1:3 and 1:4 respectively and each sample subjected to a proximate analysis by using American Society for Testing and Materials (ASTM) standards to determine the density, kinematic viscosity, flashpoint and specific gravity respectively. Findings showed that the best ratio that yielded very low biodiesel density was 1:1 for the AGOPKO blend with 0.901 g/ml. While that of the AGOGO blend is the sample in ratio 1:2 with a density value of 0.8894 g/ml. None of the kinematic viscosity recorded was lower than the limit set by the ASTM standard because all of them have far above 1.9 cSt recommended. Only the ratio 1:4 of DPKO met the minimum standard biodiesel flashpoint of 132 °C. While the same value was obtained for the two samples of blends in ratio 1:4 which is 0.905 each for AGOPKO and AGOGO, implying that specific gravity is inversely proportional to density, kinematic viscosity and flashpoint respectively. Furthermore, the optimum calorific values (45.35 and 45.40 MJ/kg) were obtained at ratio 1:2 of AGOPKO and ratio 1:1 of AGOGO respectively. The specific biofuels samples when combusted in a swirl burner gave a maximum temperature of (1020, 992, 966 and 950)oC and (1055,1022,997 and 969)oC respectively. Therefore, groundnut-palm kernel oil based fuel is suitable for combustion.

DOI: /10.61463/ijset.vol.13.issue2.302

Deep Learning Ophthamology: Predicting Eye Diseases Using Pre-Trained Neural Network Algorithm

Authors- Siva Prasad J, Sathish Reddy Cm, Lakshmi Narayana U

Abstract-– Retinal imaging has emerged as a valuable tool in the early detection and diagnosis of various systemic diseases. This study presents a novel approach for the simultaneous prediction of multiple diseases utilizing retinal images. The proposed methodology involves the collection of a diverse dataset comprising retinal images labelled with the presence or absence of multiple diseases, including but not limited to diabetic retinopathy, age-related macular degeneration, glaucoma, and hypertensive retinopathy. Preprocessing techniques are applied to ensure data consistency and remove noise, followed by feature extraction from retinal images using advanced deep learning architectures. Machine learning models, including multi-label classification and multi-task learning, are trained on the extracted features to predict the presence of multiple diseases simultaneously. The performance of the models is evaluated using rigorous validation techniques, including accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve (AUC- ROC). Clinical validation is conducted to assess the effectiveness of the predictive system in real-world healthcare settings. The integration of the predictive system into clinical workflows is discussed, emphasizing seamless interaction with healthcare professionals and compliance with regulatory standards. The study concludes with insights into ongoing research and development efforts aimed at further improving the accuracy and scope ofmulti- disease prediction using retinal images using VGG16 framework in deep learning framework. This research represents a significant step towards leveraging retinal imaging for comprehensive disease diagnosis and management, with the potential to enhance early intervention and improve patient outcomes.

DOI: /10.61463/ijset.vol.13.issue2.303

Blockchain-Based Secure Patient Data Management System

Authors- Jatin Verma, Aadit Khanolkar, Jatin Jain, Aditya Suresh Kasar, Dr. Preeti Godbole, Dr. Shailender Aote

Abstract-– In the digital age, managing patient records efficiently and securely is critical to healthcare systems. The increasing vol- ume of healthcare data, coupled with stringent regulatory require- ments, necessitates innovative solutions to ensure confidentiality, integrity, and availability of patient records. Traditional electronic health record (EHR) systems rely on centralized databases, making them vulnerable to cyberattacks, unauthorized modifications, and single points of failure. This paper explores the implementation of a patient data management system using blockchain technology, leveraging Hyperledger Fabric, Chaincode smart contracts, and secure APIs for seamless data integration. The proposed system ensures tamper-proof records, decentralized data access, and com- pliance with global healthcare regulations such as HIPAA and GDPR.

Utility of Polarity Analysis Software in Uncomplicated Urinary Tract Infections (UTI) in Women Using Uti-Siq Questionnaire

Authors- Assistant Professor Dr Divya Parmar

Abstract-– Background: Urinary tract infections (UTIs) are among the most prevalent infections worldwide, particularly affecting women and often leading to significant morbidity and substantial antibiotic use, which contributes to the alarming rise in antimicrobial resistance. Despite the progress made in healthcare, UTIs continue to pose a significant challenge today. Homeopathy, which focuses on treating the patient holistically by considering the individual’s overall health, constitutional type, and emotional well-being, employs various tools to enhance the precision of remedy selection. One such tool is Polarity Analysis, developed by Dr. Heiner Frei, which modernizes Bönninghausen’s traditional approach by using Bayesian statistical principles to refine remedy selection based on polar symptoms. This method addresses the concept of contraindicating symptoms, where a remedy’s polar symptoms contradict the patient’s symptoms, rendering the remedy unsuitable. Aim: This review and case series article investigates the utility of Polarity Analysis (PA) software in the homeopathic treatment of uncomplicated UTIs in women, emphasizing its role in selecting individualized remedies through polar symptom analysis and its potential benefits in managing these infections. In this scenario, patients were examined and diagnosed. Then patients were asked to fill the UTI_SIQ questionaries which served. Next, the patient filled out a Checklist (available at www.heinerfrei.ch resources) to record the modalities and polar symptoms they have observed. Finally, the most suitable remedy is determined through repertorization using the PA software. Results: The traditional challenges of subjectivity in homeopathic prescriptions were addressed effectively through the structured and statistically backed framework of PA, enhancing precision in identifying suitable remedies. Conclusion: The PA software appeared to enhance the accuracy of remedy selection, leading to notable reductions in UTI symptoms and high patient satisfaction, indicating its potential efficacy in homeopathic practice. Additionally, exploring its application in managing other medical conditions could further establish its role in clinical practice.

DOI: /10.61463/ijset.vol.13.issue2.304

Intelligent Attendance Tracking System VIA Webcam

Authors- Mohammad Ubaid, Zaid Khan, Mohammed Maarij Ahmad, Sufiyan Ali, Professor A. D. Sawant

Abstract-– Attendance tracking is a crucial aspect of educational institutions and workplaces. Traditional attendance systems, such as manual entry and RFID-based methods, are often prone to errors, inefficiencies, and proxies. This paper presents an intelligent attendance tracking system utilizing webcam-based facial recognition technology. The system leverages computer vision and machine learning techniques to automate attendance marking, ensuring accuracy and reducing human intervention. The implementation is done using Python, employing libraries such as OpenCV, Face Recognition, Tkinter, and Excel manipulation tools.

DOI: /10.61463/ijset.vol.13.issue2.305

Agriculture Equipment Rental System to increase the Income and Decrease the Suicide Ratio of Farmer

Authors- Vipul Sonone, Siddesh Zade, Shruti Wanare, Achal Shinde, Professor C.S.Gadge

Abstract-– The agricultural sector in many countries is facing significant challenges, including limited access to resources, inadequate infrastructure, and fluctuating market prices. These challenges have led to a decline in farmers’ income, resulting in increased debt, poverty, and, tragically, a rising number of farmer suicides. This paper proposes the development of an agriculture equipment rental app, designed to increase farmers’ income and reduce the suicide ratio. The app aims to provide farmers with access to affordable and high-quality agricultural equipment, enhancing their productivity and reducing costs. By promoting sustainable agriculture practices and providing a viable solution to farmers’ problems, the app has the potential to positively impact the livelihoods of millions of farmers worldwide.

DOI: /10.61463/ijset.vol.13.issue2.306

Enhanced Seismic Performance of Irregular Multi-Story Buildings through Strategic X-Bracing: A Review

Authors- Janhavi Banarase, Assistant Professor I. B. Dahat

Abstract-– Seismic resilience is a critical design consideration for multi-story buildings, particularly those with irregular geometries, due to their heightened vulnerability to damage from uneven mass distribution, stiffness, and height variations. Traditional reinforcement techniques often prove insufficient for these complex structures under seismic loads, necessitating innovative approaches. This study investigates the efficacy of X-bracing—a reinforcement method using steel or concrete braces arranged in an “X” shape—in improving the seismic performance of irregular multi-story buildings. By redistributing lateral forces, X-bracing enhances lateral stiffness, limits drift, and absorbs seismic energy, making it a promising solution for mitigating earthquake-induced damage.

The Detecting and Mitigating Fake News Using Machine Learning Techniques

Authors- Sarthak Thakre, Prathamesh Yenurkar, Sandesh Sabale, Shreyash Kude, Aniket Hiwarkar, Mrs. Harshita D. Jain

Abstract-– The Fake News Detection System with Integrated Chatbot Summarization is an intelligent application designed to enhance digital media literacy and curb misinformation by detecting and classifying fake news in real-time. The system leverages machine learning algorithms to evaluate the credibility of news content while a chatbot interface provides users with concise summaries and credibility scores. This paper discusses the technical architecture, implementation of natural language processing techniques, and real-time user interaction through the chatbot. The proposed solution offers a user-friendly and accessible method to verify news, thereby empowering individuals to make informed decisions. Evaluation results confirm the system’s accuracy in diverse news categories and its practical utility in daily media consumption.

Spam Detection in Social Networks

Authors- Sathishkumar N, Marikannan P, Mohanraj D, Aravind R, Assistant Professor Dr. S Mohan

Abstract-– This project focuses on spam detection in social networks, which has become an essential task due to the large volume of user- generated content. Natural Language Processing (NLP) techniques are employed to detect spam on social media platforms, leveraging various NLP libraries to process and analyze textual data. Libraries such as NLTK (Natural Language Toolkit), pandas. NLTK are used for tasks like tokenization, part-of-speech tagging and sentiment analysis to extract meaningful features from the text. Additionally, machine learning algorithms, implemented through Scikit-learn library utilized to build classifiers that identify spam posts. Techniques such as Naive Bayes and Support Vector Machines (SVM) are explored for their ability to distinguish between spam and non-spam content.

Optimization and FFT Comparison of Multistage Cascaded OP-AMPs in 0.18µm SCMOS

Authors- Poojashree Chouksey, Professor Dr.Anamika Singh, Professor Dr. Soni Changlani

Abstract-– The design presented in this paper focuses on the design and frequency analysis of two-stage, three-stage, and multistage CMOS operational amplifiers. The study investigates the influence of various parameters on amplifier characteristics, designed with a 5V power supply using C5 process CMOS technology. The key performance parameters include Gain, Phase Margin, Gain Bandwidth, and Slew Rate, analyzed through AC, DC, and transient response simulations. The development process starts at the schematic design phase and continues into simulations to optimize performance metrics. The circuit has been implemented based on a multistage CMOS transistor design. Electric CAD for VLSI version 9.07 has been utilized for circuit schematics, whereas LTSpice is used to run simulations and validate the obtained results. According to literature, various operational amplifier configurations exist that have been typically divided into four types: single-stage, two-stage, three-stage, and multistage amplifiers. Here, special emphasis is put on two-stage and three-stage topologies since these are useful in low-voltage high-performance applications.

Design and Simulation of Steady State Thermal Analysis of Exhaust Engine Valve

Authors- Mrs. B. Kranthi1, Puli Hari Shankar2, Kadiyala Rupananda Ganesh Kumar3, Ramavath Arun Kumar Naik4, Sai Durga Nishith Medisetti

Abstract-– The design presented in this paper focuses on the design and frequency analysis of two-stage, three-stage, and multistage CMOS operational amplifiers. The study investigates the influence of various parameters on amplifier characteristics, designed with a 5V power supply using C5 process CMOS technology. The key performance parameters include Gain, Phase Margin, Gain Bandwidth, and Slew Rate, analyzed through AC, DC, and transient response simulations. The development process starts at the schematic design phase and continues into simulations to optimize performance metrics. The circuit has been implemented based on a multistage CMOS transistor design. Electric CAD for VLSI version 9.07 has been utilized for circuit schematics, whereas LTSpice is used to run simulations and validate the obtained results. According to literature, various operational amplifier configurations exist that have been typically divided into four types: single-stage, two-stage, three-stage, and multistage amplifiers. Here, special emphasis is put on two-stage and three-stage topologies since these are useful in low-voltage high-performance applications.

DOI: /10.61463/ijset.vol.13.issue2.307

Unveiling Land Use Pattern: A Comprehensive Approach to Satellite Image Analysis

Authors- Durgesh Bhatkar, Prathamesh Chandurkar, Aryan Gohatre, Dr. Vinay Rajgure, Sushant Dhone

Abstract-– Classifying Land Use and Land Cover (LULC) is essential for urban planning, environmental assessment, and resource management. This study introduces an automated LULC classification technique that employs image processing tools using OpenCV and Flask. The method involves segmenting images with the HSV color model and identifying built-up areas, natural vegetation, and other land categories. The findings highlight an effective classification process, offering valuable insights into urban and ecological landscapes.

DOI: /10.61463/ijset.vol.13.issue2.308

Enhancing Student Engagement: A Comprehensive IT Department Portal for Resources and Community

Authors- Mr. K. Nagaraju, Mr. Chetan kajge, Ms. Pratiksha Barde, Ms. Rutuja Junghare, Mr.Chetan Chaudhari

Abstract-– The Information Technology Department website serves as a central hub for accessing resources, services, and information related to the department. It provides details about the IT department’s objectives, available courses, faculty members, research activities, and events. Additionally, the site offers tools for students, staff, and faculty to access academic materials, submit IT-related requests, and stay updated on departmental announcements and news.

DOI: /10.61463/ijset.vol.13.issue2.309

FixMate: A Smart Digital Platform for Booking Skilled Household Technicians

Authors- Shahbaz Aazmi, Sufiyan Khan, Sufiyan Ahmed, Saqueef Ahmed, Dr. S.S. Agrawal

Abstract-– FixMate is an online platform that helps people easily find and hire skilled technicians and household workers. It connects users with verified professionals like plumbers, electricians, carpenters, AC mechanics, painters, cleaners, and appliance repair experts. Many people face difficulty in finding reliable help for household repairs, and FixMate solves this problem using technology. The platform ensures that all workers are verified, skilled, and rated by previous customers. Users can book services with just a few clicks and get help at their doorstep. FixMate also supports workers by giving them access to more job opportunities and fair payments. It uses a secure, user-friendly interface that works on both web and mobile. The platform is designed to be fast, reliable, and scalable to serve many users. FixMate also contributes to improving the informal workforce by offering them a professional space to grow. This paper discusses the idea behind FixMate, how it works, and the impact it can make in everyday life.

DOI: /10.61463/ijset.vol.13.issue2.310

Real Time YOLO Based Mycobacterium Tuberculosis Detection Model

Authors- Arul Selvam P, Sivadhandapani S, Udhaya Barathi U, Kishore M, Balakumar k

Abstract-– Accurate identification of Mycobacterium tuberculosis (M. tuberculosis) is a critical step in the diagnosis of tuberculosis. Existing object detection methods struggle with the challenges posed by the varied morphology and size of M. tuberculosis in sputum smear images, which makes precise targeting difficult. To solve these problems, an improved YOLOv8s model is proposed. Specifically, an additional detection head is added to focus on small target information. Second, a multi- scale feature fusion module is introduced to adapt the model to different sizes of M. tuberculosis. In addition, a convolutional layer is added to the Coordinate Attention (CA) module to extract more advanced semantic features. Finally, a self-attention mechanism is added after the CA module to enhance the model’s ability to accurately understand and localize the varied morphology of M. tuberculosis. Our model performed well with an average precision of 85.7 % when tested on a publicly available dataset. This clearly demonstrates the effectiveness of our proposed model in M. tuberculosis detection.

High-Performance 3D-Printed Materials for Aerospace Jet Engine Applications

Authors- Renuka Suryawanshi, Radha Tadas, Nakul Thote

Abstract-– The aerospace sector encounters considerable chal- lenges in the production of jet engine parts due to high ex- penses, lengthy manufacturing timelines, and material waste linked to traditional techniques. Moreover, conventional materials frequently fall short of the stringent strength-to-weight ratio, thermal durability, and longevity necessary for jet engines. These constraints underscore the necessity for advanced materials and innovative manufacturing methods to boost efficiency, perfor- mance, and sustainability. This research investigates the potential of Additive Manufacturing (AM), also known as 3D printing, for use in aerospace jet engine applications. The main objective is to examine how AM can enhance material properties, improve manufacturing efficiency, and lower total costs. The research reviews recent progress in nickel-based superalloys, titanium alloys, and ceramic matrix composites, which provide greater mechanical strength, heat resistance, and lightweight character- istics compared to traditional materials. A structured approach is utilized, incorporating experimental analysis, computational modeling, and process optimization to evaluate the mechanical performance, fatigue resistance, and thermal stability of 3D- printed jet engine components. The methods include selective laser melting (SLM), electron beam melting (EBM), and directed energy deposition (DED), alongside post-processing techniques such as hot isostatic pressing (HIP) and surface treatments. The results indicate that 3D printing facilitates the creation of lightweight, durable, and aerodynamically efficient engine com- ponents, resulting in improved fuel efficiency, lower emissions, and enhanced sustainability.

Next Generation EV Wireless Power Autonomous Station with Smart Billing System

Authors- Ms.S.Vasanthiriya, S.Mahendra, P.Venkateswarlu, U.Gomadhan

Abstract-– This system presents the most promising wireless power transfer (WPT) technique for charging electric vehicle (EV) batteries, namely inductively coupled power transfer (ICPT). In order to solve the inconvenient problem of electric vehicles (EVs) wired charging mode, the high-power and low-frequency resonant wireless power transfer (WPT) charging system is studied and implemented. Compared with wired charging mode, WPT charging mode can provide power supply for Evs more flexibly and more conveniently. Here we design a wireless charging platform which consists of a wireless power transmitter unit with a transmitter coil fitted under the base of a parking platform. When a vehicle is parked on the platform the vehicle get detected and charged automatically.

Smart Waste Management Using IoT and AI – Optimizing Garbage Collection and Recycling

Authors- Aditya Mundhe, Akshar Patil, Harshal Patil

Abstract-– Effective trash management is an essential compo- nent of urban sustainability, yet traditional collection systems frequently result in overflowing bins, poor route design, and insufficient recycling efforts. This study describes an innovative smart waste management system that combines IoT sensors and AI analytics to improve garbage collection, reduce pollution, and increase recycling efficiency. The system uses ultrasonic, weight, and gas sensors to monitor waste levels in real time, while AI- powered route optimization reduces fuel usage and operating costs. IoT sensors collect data, which is subsequently analyzed in the cloud for predictive improvement. The suggested method is useful in urban garbage collection, industrial waste management, and public health efforts since it prevents waste overflow and streamlines collection routes. By merging AI, cloud computing, and IoT-enabled tracking, this system delivers a scalable, cost- effective, and eco-friendly solution that promotes sustainability and smart city initiatives. The originality of this research is in its potential to support real-time decision-making using AI-driven optimizations and cloud-based data, enabling authorities a more efficient and advanced waste management framework.

Real-Time Vision-Based Object Detection and Tracking

Authors- Bhagyashri Sharad Pisal

Abstract-– This research focused on evaluating object detection and tracking methods for real- time vehicle monitoring in urban environments, with a specific emphasis on parking spot tracking. Various versions of the YOLOv8 detection model were investigated to understand the trade-offs between speed and accuracy. The smaller YOLOv8n model offered high frame rates and low computational requirements, making it well- suited for real-time scenarios. In contrast, the larger YOLOv8l model delivered improved detection accuracy but with slower inference speeds, making it preferable for applications where precision is more critical. The research also compared tracking algorithms, specifically OC-SORT and StrongSORT. StrongSORT demonstrated higher tracking accuracy and better performance in handling occlusions and non-linear object movement, while OC- SORT provided faster processing and performed effectively in simpler tracking tasks. Based on the findings, the optimal setup for real-time vehicle and parking spot monitoring involves using YOLOv8n for efficient detection and StrongSORT for robust tracking. The study also highlighted the importance of high-resolution input, such as 4K video, in improving detection detail and enabling accurate license plate recognition.

Installing A Single Plc For Home Automation of An Entire Society

Authors- Kunal Liladhar Patil, Dhananjay Rathod

Abstract-– Home automation is reshaping contemporary living environments by improving efficiency, security, and energy control. This study introduces a novel method for automating an entire residential community with a single Programmable Logic Controller (PLC). In contrast to conventional home automation systems that depend on various microcontrollers or IoT modules, this system centralizes management, lowering complexity and maintenance expenses. The PLC serves as a central hub, effectively managing various tasks including lighting, security, water distribution, energy optimization, and predictive maintenance. To facilitate smooth automation, the system integrates modular input-output setups, enabling it to adjust to different needs within the residential community. The PLC utilizes Modbus and MQTT protocols for communication, allowing for real-time data transfer and remote access. Furthermore, a Human-Machine Interface (HMI) is incorporated, enabling administrators to oversee and manage different parameters via an interactive dashboard. The suggested system not only boosts security through automated gate controls, biometric verification, and CCTV monitoring, but also increases resource efficiency by optimizing energy use and water distribution according to real-time requirements. Additionally, machine learning algorithms can be implemented to forecast system failures, minimizing downtime and maintenance expenses. This research emphasizes the practicality, benefits, and drawbacks of utilizing one PLC for extensive residential automation, showcasing its ability to transform home automation by providing a scalable, economical, and smart solution for intelligent communities.

DOI: /10.61463/ijset.vol.13.issue2.311

Automated Deep Learning for Threat Detection in Luggage from X-ray Images

Authors- Sanjay Varma K, Maruthachalaprabu R, Mohamed Nazeem Ibrahim M, Gunalan R, .Anupriya.k

Abstract-– Luggage screening plays a crucial role in airport security, helping to identify potential threats and ensure passenger safety. Automating the detection of dangerous objects in X-ray scans can make this process faster and more efficient. In this study, we explore different machine learning algorithms to detect firearm parts in baggage scans, with a particular focus on identifying steel barrel bores—an essential component of a weapon. We worked with a dataset of 22,000 dual-view X-ray images, containing both harmless and potentially dangerous objects. First, we applied standard image processing techniques, such as noise reduction, edge detection, and grayscale thresholding, to clean and refine the images. Then, we used advanced deep learning models—Convolutional Neural Networks (CNNs) and Stacked Autoencoders—to classify objects in the scans. To compare performance, we also tested simpler machine learning models, including shallow Neural Networks and Random Forests. To validate our findings, we tested our models on a second dataset containing X-ray scans of courier parcels. Our results highlight the strengths of deep learning in accurately identifying firearm parts, and we discuss the advantages of our approach compared to traditional methods.

Identity Centric Cloud Security

Authors- Sonam Priya, Subarna Shakya

Abstract-– Cloud computing’s inherent scalability and flexibility necessitate robust Identity and Access Management (IAM) systems to ensure secure access to resources. This paper focuses on developing an identity-centric cloud security framework, with the primary goal of streamlining user authentication and access control. The project involved integrating Zendesk with the AWS console and implementing Single Sign-On (SSO) based on Role-Based Access Control (RBAC). AWS Cognito was employed to manage user identities, enabling secure authentication and authorization, while distinct roles (admin and basic) were defined to enforce granular access permissions. The implemented system successfully demonstrates secure user authentication and the effective use of RBAC to control resource access. By leveraging AWS services, the architecture provides scalability and facilitates real-time monitoring through CloudWatch, enhancing the ability to detect and respond to security incidents. The findings underscore the importance of an identity-centric security model in cloud environments and highlight the benefits of centralized identity management. The paper emphasizes the necessity of robust authentication and authorization mechanisms to secure cloud-based applications. The study advocates for the adoption of Identity Providers (IdPs) like AWS Cognito, which offer scalable and secure solutions for managing user identities. Furthermore, it recommends the careful definition of roles and permissions aligned with organizational security policies and continuous monitoring to maintain a strong security posture in the cloud.

DOI: /10.61463/ijset.vol.13.issue2.312

Analyzing NASA’s 2023-2024 Asteroid Close Approaches

Authors- Aakash Hivrale, Pallavi Hivrale, Dr. Mahender Kondekar

Abstract-– Asteroids, remnants from the early solar system, often pass near Earth. While most pose no threat, analyzing their trajectories is vital for planetary defense. This study investigates asteroid close approaches using NASA’s 2023–2024 dataset. Employing data analysis and machine learning techniques, this research identifies patterns, builds predictive models, and visualizes results to enhance preparedness and public awareness. The results support strategic planetary defense and offer a framework for future asteroid monitoring initiatives.

DOI: /10.61463/ijset.vol.13.issue2.313

Optimizing Regression Test Suites for Automotive Embedded Systems: A Risk-Based Approach

Authors- Vikas Gore, Pranjal Dhengale

Abstract-– The In the realm of automotive embedded systems, ensuring the reliability and safety of software is paramount. Regression testing plays a critical role in maintaining software quality by verifying that recent code changes have not adversely affected existing functionalities. However, the extensive nature of regression test suites can lead to significant time and resource consumption. This paper presents a risk-based approach to optimizing regression test suites for automotive embedded systems. By prioritizing test cases based on the risk associated with different software components, we aim to enhance testing efficiency while maintaining high standards of safety and reliability. Our methodology involves identifying high-risk areas through a combination of historical data analysis, expert judgment, and automated risk assessment tools. The proposed approach is validated through a series of case studies, demonstrating its effectiveness in reducing test execution time and resource usage without compromising the detection of critical defects. Automotive embedded systems are growing in complexity, requiring efficient regression testing to meet safety standards like ISO 26262. Traditional approaches often result in redundant test executions, increasing time and resource usage. This paper proposes a risk-based method to optimize regression test suites by prioritizing and minimizing test cases based on software criticality, change impact, and historical fault data. Risk is assessed using a scoring model incorporating ASIL, change frequency, and defect density. Applied to a simulated ECU, the approach shows improved testing efficiency while maintaining fault detection and system reliability. It supports faster development cycles and robust software delivery, especially in CI/CD and OTA update scenarios.

DOI: /10.61463/ijset.vol.13.issue2.314

Osteoarthritis Severity Level Detection Using X-ray Images

Authors- Dr.D. Satheesh Kumar, Mohit Chandran, Mukesh G, Shreeman MM, Karthick P

Abstract-– Osteoarthritis (OA) is a chronic degenerative joint disease affecting millions worldwide, leading to joint pain, stiffness, and limited mobility. Early diagnosis is critical for effective treatment and disease management. Traditional OA diagnosis relies on radiologists interpreting X-ray images, which is a time-consuming and subjective process. The integration of artificial intelligence (AI) in medical imaging offers a promising solution to automate and enhance the diagnostic process. In this study, we propose an AI-driven approach for osteoarthritis severity level detection using X-ray images. We utilize a fine-tuned EfficientNet model, trained on the Kaggle Osteoarthritis dataset. The model employs convolutional neural networks (CNNs) with dropout layers to improve generalization and prevent overfitting. Preprocessing techniques such as normalization, augmentation, and resizing enhance the robustness of the model. The AI model is optimized using the Adam optimizer and evaluated using binary cross-entropy loss. Experimental results demonstrate that our approach achieves high classification accuracy, outperforming traditional methods.

Transformer Models in NLP

Authors- Ayush Sharma, Harsh Mehta, Abhishek Gautam, Priyanka Rajpal, Ashima Thakur, Harleen Kaur

Abstract-– Natural Language Processing (NLP) has been transformed by transformer models, which introduce self-attention mechanisms that effectively capture long-range interdependence. Transformers, as opposed to recurrent networks, allow for parallel processing, which boosts efficiency for jobs like text production and translation. This architecture, was first presented by Vaswani et al. (2017) [1], opened the door for models such as GPT (autoregressive generation) and BERT (bidirectional understanding) [2]. NLP applications are further expanded by more advanced versions like T5 and GPT-4 [6]. Despite their achievements, interpretability and high computing costs remain obstacles [7]. To guarantee that transformers continue to innovate in AI-driven language processing [14], ongoing research focuses on increasing efficiency and minimizing biases.

DOI: /10.61463/ijset.vol.13.issue2.315

Auto Machine Learning

Authors- Shubham Gadhave, Akanksha Barge, Abhishek Dongre

Abstract-– The AutoML Platform is an innovative web-based solution designed to automate and simplify the machine learning lifecycle. It allows users to upload datasets, perform automated preprocessing, train models, and gain actionable insights with minimal technical intervention. Leveraging modern technologies like Flask, scikit-learn, and advanced algorithms such as XG- Boost and LightGBM, the platform supports classification and regression tasks. It integrates the Google Gemini API to provide AI-driven recommendations for data profiling, feature engineer- ing, hyperparameter tuning, and model performance analysis. Key features include secure user management, intelligent dataset analysis, cross-validation, and model deployment with robust error handling and security measures. By streamlining complex workflows and offering AI-enhanced insights, the platform de- mocratizes machine learning, making it accessible, efficient, and scalable for diverse users and applications.

Cryptocurrency Tracker for Real Time Code Analysis and Monitoring

Authors- Ankita Nathe, Bali Donadkar, Vaishanvi Digraskar, Shyam Nimje, Samrudhi Solunkey

Abstract-– The word crypto comes from the idea of hiding, so people often think of cryptocurrency as hidden mon- ey. But it’s important to understand that cryptocurrency is not a physical type of money you can’t touch it or hold it like coins or paper bills. Its value goes up when more people start using it. Cryptocurrency is a very popular topic right now. It’s a type of digital money that allows for quick, safe, and easy transac- tions online. It works through a system called blockchain, which is like a public record where all trans- actions are kept safe and visible to everyone, but the users’ identities remain private. Blockchain helps keep everything secure and makes sure the information is stored across many different systems.

A Comparative Survey on Sentiment Analysis Using NLP and Machine Learning

Authors- Bonwale Priyanka Mahadev

Abstract-– Sentiment Analysis (SA), or opinion mining, is a crucial aspect of Natural Language Processing (NLP) that focuses on determining the sentiment expressed in textual data—whether it is positive, negative, or neutral. With the rapid growth of user- generated content on social media, forums, and review sites, SA has become an essential analytical tool for various applications. This paper presents a comprehensive comparative study of sentiment analysis techniques, contrasting classical Machine Learning (ML) algorithms with advanced Deep Learning (DL) models. Traditional models such as Naïve Bayes, Support Vector Machines (SVM), and Logistic Regression are evaluated against DL models like Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Long Short-Term Memory networks (LSTMs), and Transformer-based architectures including BERT and RoBERTa. The study discusses critical NLP processes including preprocessing, feature engineering, and vectorization strategies—ranging from Bag-of-Words (BoW) and TF-IDF to Word2Vec and contextual embeddings. Experimental results using multiple benchmark datasets underscore the superiority of Transformer models in most contexts while highlighting the continued relevance of classical models in constrained environments. Finally, ongoing challenges and future research directions are presented, including explainability, zero-shot learning, multilingual support, and multimodal sentiment analysis.

A Comprehensive Review of Optimizing Microgrid Design and Operation: Challenges, Issues, and Techniques

Authors- Sagar S. Chaugule, Assistant Professor Dr. P. V. Paratwar

Abstract-– The increasing levels of carbon emissions and greenhouse gases have heightened the demand for renewable energy sources (RESs). This shift is driven by the rapid depletion of natural resources, necessitating the integration of renewable energy into power systems. Microgrids, with their decentralized control configurations, represent a transformative approach to energy distribution, moving away from traditional centralized systems. These systems, installed on the demand side of distributed generation, are poised to reshape electricity markets at both the generation and distribution levels in the coming years. Microgrids also facilitate diversified investments across sectors such as generation operations. This review explores the fundamental concepts of microgrids, including their definitions, classifications, associated challenges, and optimization techniques aimed at enhancing grid stability and security.

Machine Learning Framework for the Detection of Mental Stress Via Webcam

Authors- Prasad U. Giridhar, Yash J. Hirudkar, Prashant K. Joshi, Shravani M. Karne, Professor Gargeya M.5

Abstract-– This paper presents a Machine Learning Framework for Face Emotion Recognition (FER) aimed at detecting mental stress in real-time during online digital learning. The system is designed to enhance e-learning experiences by identifying student emotions using a deep convolutional neural network (CNN) trained on the FER-2013 dataset. With a training accuracy of 77% and validation accuracy of 64%, framework integrates Flask for web deployment and OpenCV for video capture, enabling real-time emotion classification across seven emotion categories. Insights generated through emotion detection can assist educators in understanding student engagement, detecting stress indicators,and optimizing teaching strategies based on emotional cues, ultimately contributing to more responsive and supportive digital learning environments.

DOI: /10.61463/ijset.vol.13.issue2.316

Performance Analysis of a Solar PV and Wind Hybrid Energy System with Energy Storage System

Authors- Saket Bihari, Professor Ashish Kumar Rai

Abstract-– This paper presents a detailed performance analysis of a hybrid renewable energy system comprising Solar Photovoltaic (PV), Wind Energy Conversion System (WECS), and an integrated Energy Storage System (ESS). The objective is to enhance power reliability, reduce intermittency, and ensure continuous energy supply in a standalone or grid-connected environment. A dynamic simulation model is developed, and system performance is evaluated under varying solar irradiance, wind speed, and load demand profiles. Results show that the hybrid system with ESS effectively balances supply-demand mismatches and ensures voltage stability and frequency regulation, even during fluctuating weather conditions.

Leveraging IoT-Based Technology for Precision Water Management in Agriculture

Authors- Krushna Varat, Sudhanshu Wagh

Abstract-– Inefficient irrigation makes water scarcity worse in farming, leading to waste, soil damage, and lower crop yields, especially in dry areas. using IoT technology to manage water better is still limited because of technical, cost and practical issues. This research aims to create an IOT based system that improves water usage in Irrigation by using real-time data from soil sensors, weather forecasts, and crop needs. The study uses a multiphase approach, collecting data with IOT sensors, analyzing it with machine learning, and implementing a smart irrigation system. The system features an easy-to-use interface. allowing farmers to control irrigation schedules and it is adaptable to different crop, soil type and weather Condition. This research presents an innovative IoT based precision irrigation solution that tackles kay Issues like water conservation, sustainability and food security in agriculture.

DOI: /10.61463/ijset.vol.13.issue2.318

Intelligent Engine Management System for Fuel Efficiency and Emission Control

Authors- Rounak Sayyad, Siddhi Hande, Ayushi Tayde

Abstract-– Modern cars must have improved performance, reduced emissions, and increased fuel efficiency. Conventional engine management systems (EMS) sometimes have trouble adapting to various driving situations, which leads to inefficiency. In order to increase engine efficiency, this study presents an advanced EMS that makes use of sensors, Electronic Control Units (ECUs), and Machine Learning (ML). In order to adjust fuel injection and ignition timing, the system uses adaptive control techniques and collects real-time data from sensors that track engine characteristics including temperature, pressure, and RPM. The suggested EMS contributes to the development of intelligent and ecologically friendly transportation systems with noteworthy enhancements, such as fuel savings of 10.

DOI: /10.61463/ijset.vol.13.issue2.319

Enhancing Vehicle Safety with Anti-Braking System (ABS)

Authors- Aakriti Tickoo, Ketaki Varade, Siddhi Chipade

Abstract-– This research paper examined the Anti-lock Braking System (ABS) for enhanced vehicle safety and braking capa- bility. The primary problem it focused on was how sudden braking wheel lockup would make drivers lose control and prolong stopping distances. The objective of the study was to observe how well ABS avoids wheel lock, stabilizes vehicles, and minimizes accidents.The findings indicated that ABS enabled cars to stop faster on slippery and wet roads while maintain- ing the ability to steer.For practical use, ABS was combined with modern electronic control systems and traction control to better manage braking force. This study added new value by identifying weaknesses in traditional braking systems and suggesting improvements using advanced sensors and smart braking technology. Data analysis also supported that ABS improved braking, particularly in emergency situations.

DOI: /10.61463/ijset.vol.13.issue2.320

Robust Inventory Control under Intermittent Demand and Supply Disruptions: A Fuzzy Logic Approach

Authors- Purnima Raj, Nitish Kumar Bharadwaj

Abstract-– Inventory management in the presence of intermittent demand and unpredictable supply disruptions presents significant challenges to traditional deterministic models. This study proposes a robust inventory control framework using fuzzy logic to effectively manage uncertainty in both demand and supply chains. The model incorporates fuzzy demand forecasts, probabilistic supply interruptions, and flexible replenishment policies. By utilizing fuzzy inference systems and robust optimization techniques, the proposed approach enhances decision-making under uncertain environments. Numerical simulations and comparative analyses with conventional models demonstrate the superiority and practical viability of the proposed fuzzy logic-based inventory control system. Furthermore, sensitivity analysis and graphical illustrations support the robustness of the model across varying scenarios of disruption intensity and demand irregularity.

DOI: /10.61463/ijset.vol.13.issue2.323

Plane Symmetric Cosmological Model and Quadratic Equations of State Solutions in f(R, T) Gravity

Authors- L.S. Ladke, S.D.Mhashakhetri, V.P. Tripade</p

Abstract-– In this paper, we have studied the perfect fluid matter distribution using three equation of states for plane symmetric space time in),( TRf gravity also, obtained solution of field equations in gravity assuming shear scalar  is proportional to the expansion scalar . Some kinematical properties of the model have been studied.

DOI: /10.61463/ijset.vol.13.issue2.321

Waste Segregation and Recycling Guide Using Machine Learning

Authors- Bingi Aswini, Panyam Ramachandra Reddy, M.Chandana Priya, Vaddi Nikith Raj, B. Tejeswar Reddy

Abstract-– The increasing volume of waste generated due to rapid urbanization poses significant challenges to environmental sustainability and public health. Traditional waste segregation methods are often manual, inefficient, and error-prone, highlighting the need for intelligent automation. This paper presents Greencore, an AI-powered system designed to automate waste classification and promote eco-friendly disposal practices. The system employs lightweight machine learning models such as MobileNet and YOLO to accurately classify over 80 types of waste—including recyclable, biodegradable, and hazardous materials—through images captured via a mobile application. Integrated with geolocation services using the Google Maps SDK, the application guides users to the nearest appropriate disposal or recycling facility within a 50-meter radius. Built on a robust backend using FastAPI and MongoDB, the platform also incorporates user behavior tracking, JWT- based authentication, and real-time performance monitoring. Data preprocessing techniques such as normalization and augmentation enhance the accuracy and robustness of the models. The proposed system not only simplifies household and industrial waste management but also contributes to smart city initiatives and circular economy goals by promoting awareness, accountability, and sustainable practices through technology.

DOI: /10.61463/ijset.vol.13.issue2.322

Optimized Data Compression for Efficient Cloud Storage

Authors- Vedashri Chorge, Dhiraj Jambhale, Aman Jain, Soham Jagtap, Professor Rashmi Jolhe

Abstract-– The exponential growth of digital data demands smarter and more secure cloud storage solutions. This project introduces an intelligent framework that integrates lossless data compression, AES encryption, and NLP-based PDF summarization to optimize storage, enhance security, and improve document usability. Compression reduces file size and bandwidth usage without compromising integrity, while encryption ensures safe transmission and access control using AWS S3 for scalability and efficient retrieval. To enhance document accessibility, NLP-powered summarization enables users to extract key insights from lengthy PDFs, reducing reading time and boosting productivity. Transformer-based models help streamline content management for research, corporate, and cloud-based applications. By combining compression, encryption, and intelligent summarization, the system delivers a secure, scalable, and efficient cloud storage solution. Performance evaluations confirm improved retrieval speed and reduced storage needs. Future upgrades may include AI-driven adaptive compression, real-time optimization, and advanced data security for enhanced cloud performance.

DOI: /10.61463/ijset.vol.13.issue2.324

IoT-Driven Boiler Monitoring and Control System

Authors- Satish Baban Arjun, Omkar Sanjay Kaldhone, Abhayraje Somnath Gaikwad

Abstract-– This project focuses on the development of an IoT-based Boiler Automation System using the ESP32 microcontroller, designed to remotely monitor and manage key boiler parameters. The system employs a DHT11 sensor to measure ambient temperature and humidity, while an ultrasonic sensor is used to detect the water level inside the boiler. Sensor data is transmitted to the Blynk IoT platform via Wi-Fi, where real-time values are visualized through interactive gauges on a user-friendly interface. This solution enhances boiler safety, efficiency, and enables remote supervision and control.To ensure automated operation, a heater and water pump will be controlled through the Blynk platform, allowing remote activation based on predefined conditions. A relay module will be used to switch the heater and pump. For demonstration purposes, a small 5-liter container will be used as the boiler. This project improves boiler efficiency, safety, and operational convenience by enabling remote monitoring and control. Future enhancements may include the integration of features such as automatic alerts, energy usage optimization, and predictive maintenance capabilities to further increase system intelligence and reliability.

DOI: /10.61463/ijset.vol.13.issue2.325

Green Cloud Computing: Advancing Sustainable Solutions for Energy-Efficient Digital Infrastructure

Authors- Dhaval Amane

Abstract-– Cloud computing’s growing energy requirements present serious environmental issues as it becomes the foundation of contemporary digital infrastructure. By optimising energy use, lowering carbon footprints, and utilising renewable energy sources, green cloud computing—a game-changing strategy—integrates sustainability into cloud services. This study examines cutting-edge techniques for improving cloud infrastructures’ environmental efficiency, such as AI-driven workload management, energy-conscious virtualisation, and dynamic resource allocation. The goal of this research is to close the gap between ecological responsibility and performance scalability by analysing existing architectures and suggesting an adaptable, eco-friendly framework. The results demonstrate that the future generation of ecologically conscious IT systems cannot function without sustainable cloud computing.

Principal Availability Management System

Authors- Assistant Professor Ms.B.Manjubashini, M.S.K.Mallisavitha, D.Divya, A.S.Dhanesha, S.DhanaShalini

Abstract-– This study introduces a web-based platform specifically crafted to enhance the processes of appointment scheduling and grievance management in academic institutions. The system features a user-centric interface that allows Heads of Departments (HODs) and Directors to efficiently check the Principal’s availability in real time and request appointments without manual intervention. In addition, the platform incorporates a grievance tracking system that enables users to submit, monitor, and follow up on complaints, ensuring timely resolutions. Essential functionalities include dynamic updates, notification services, and access control tailored to user roles. The system not only reduces administrative complexity but also encourages transparency and responsiveness, fostering a well-coordinated academic environment by automating and optimizing institutional workflows.

DOI: /10.61463/ijset.vol.13.issue2.345

Predecting the Stages of Dementia Using OASIS

Authors- Vempalli Fazil, Kaluva Bhumika Reddy, Chiranjeevi Althi, Vankadari Lakshmi Bhargavi

Abstract-– Dementia is a debilitating neurodegenerative disorder characterized by progressive cognitive decline, memory impairment, and behavioral changes. Alzheimer’s disease (AD), the most common form of dementia, accounts for 60- 80% of cases worldwide. Early and accurate diagnosis is crucial for effective intervention, yet traditional diagnostic methods often fail to detect early-stage dementia due to their reliance on subjective clinical assessments. This study leverages the Open Access Series of Imaging Studies (OASIS) dataset, a comprehensive repository of neuroimaging data, demographic information, and clinical assessments, to develop a machine learning (ML)-based predictive model for dementia staging. We employ a multi- algorithm approach, including Support Vector Machines (SVM), Random Forest (RF), and Deep Learning architectures (CNN, MobileNet, ResNet), to classify individuals into distinct stages: Normal Cognitive Function Mild Cognitive Impairment (MCI) Early-Stage Alzheimer’s Disease Advanced Alzheimer’s Disease Our results demonstrate that Random Forest achieves 92% accuracy, while CNN and ResNet models excel in detecting subtle neuroanatomical changes. This research underscores the potential of AI-driven diagnostics in revolutionizing dementia care.

DOI: /10.61463/ijset.vol.13.issue2.326

Automated Railway Platform Management System with Track Switching

Authors- Kaustubh Deshpande, Pratik Dahale, Vishwajeet Hampalle

Abstract-– Due to the rapid expansion of railway networks and rising train traffic, traditional platform management and track routing systems are experiencing significant strain. Delays, human errors, and safety risks frequently result from the manual handling of train schedules, platform assignments, and track switching. This research paper presents the design and implementation of an Automated Railway Platform Management System with Track Switching, aiming to modernize railway station operations through automation and intelligent decision-makin Based on train arrival times and current platform occupancy, the proposed system automatically allocates platforms by utilizing real-time sensor data, microcontrollers, and programmable logic. In parallel, the system includes a motorized track switching mechanism to safely and effectively direct arriving trains to their respective platforms. By integrating technologies such as IR sensors, Arduino boards, servo motors, and wireless communication modules, the system achieves autonomous control over train routing and platform management. Our solution makes the most of the infrastructure that is already in place and improves passenger convenience by reducing delays and ensuring that platform information is updated promptly. Real-time operation logic, hardware and software integration, and the system architecture are all discussed in detail in this paper. The implementation results demonstrate that automation in platform management and track switching can significantly reduce human dependency, improve safety, and support the future of smart railway transportation systems.

DOI: /10.61463/ijset.vol.13.issue2.327

A Secure Smart Surveillance System for Crowd Behavior Recognition

Authors- Assistant Professor Dr. R Jayaraj, Aswin S Mohan, Ajal VC, Aravind Raj Padmanabhan

Abstract-– In an era of growing security demands, especially in public and crowded environments, intelligent surveillance systems are critical for preventing and responding to anomalous events. Traditional systems lack real-time analysis, behaviour recognition, and security in data handling. This research proposes a secure, smart surveillance system capable of detecting abnormal crowd behaviours such as stampedes, fights, and unauthorized gatherings. The core of the system is a hybrid Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) model for spatiotemporal behaviour recognition. Advanced Encryption Standard (AES) ensures secure video data transmission. Experiments using benchmark datasets demonstrate a detection accuracy of 92.5%, with low latency and robust encryption. The system offers a scalable and practical solution for real-time deployment in high-risk public areas.

Improving Plant Disease Classification with Deep-Learning-Based Predicton Model

Authors- Anzil Mohannad Muthalib, Raashid M, Vishnuprasad P, Vishnu VK, Dr. R Jayara

Abstract-– Plants are the major contributors to food provision worldwide. Different environmental conditions cause plant diseases which leads to great production loss. But manual identification of plant diseases is a laborious and prone-to-error activity. It may be an unreliable means of detecting and controlling plant diseases. Embracing cutting-edge technologies like Machine Learning (ML) and Deep Learning (DL) can solve these issues by making it possible to identify plant diseases at an early stage. In this article, the latest trends in the application of ML and DL methods for plant disease identification are discussed. The study the articles published from 2015 to 2022, and the experiments in this research illustrate how these methods work to enhance plant disease detection efficiency and accuracy. This research also discusses challenges and limitations when employing ML and DL for the identification of plant disease, including limitations in data availability, image quality, and distinction between diseased and healthy plants. The study gives significant information for practitioners, researchers of plant disease detection, and industry experts through solutions to the limitations and challenges that they present and through offering a broad vision of where this field of study stands in relation to its present research state, identifying the strengths and weaknesses of these methods, and presenting some possible solutions that can address their challenges of implementation.

DOI: /10.61463/ijset.vol.13.issue2.328

Machine Learning-Driven Prediction Model for Early Detection of Heart Disease

Authors- Associate Professor Mr.T.K.P.Rajagopal, Ms.Poorana Jothi A, Ms.Thilagam S, Ms.Vedhashini S, Mr.Vijay S

Abstract-– A reliable decision support system for heart failure prediction is essential to meet the rising need for accurate early diagnosis and risk assessment. Heart failure impacts millions globally, and precise outcome predictions can enhance patient care. Key predictive factors include demographic and physiological data such as age, gender, cholesterol, blood pressure, and ECG findings. Current methods often lack early detection accuracy, underscoring the need for advanced models. Using the Random Forest algorithm, this system addresses missing data, balances classes, and optimizes features to achieve 85.5% accuracy across four datasets. By offering personalized risk insights, it aids clinicians in making informed treatment decisions, supporting better patient outcomes.

IoT Based Energy Monitoring and Management System for Real-Time Appliances

Authors- Dr. S Shanthala, Ashwini, Avani Shridhar Ambilkar, Manasa P S, Dhanush K Gowda

Abstract-– IoT Based Energy Monitoring and Management System for Real-Time Appliances is designed to revolutionize power usage tracking, analysis, and optimization. The system integrates smart sensors, interconnected devices, and advanced analytics to capture detailed energy consumption data, providing users with comprehensive insights. By identifying inefficiencies and transforming complex data into intuitive visualizations, the system empowers users to monitor trends, detect anomalies, and implement targeted strategies for improved energy efficiency. With remote accessibility and user-friendly interfaces, both individuals and businesses can make data-driven decisions to reduce energy wastage and operational costs. Leveraging intelligent algorithms for pattern recognition and anomaly detection, the system promotes sustainable energy practices. Future advancements include integration with smart grids and renewable energy solutions, contributing to a smarter and greener energy ecosystem.

DOI: /10.61463/ijset.vol.13.issue2.329

The Influence of Music on Purchasing Behavior of Consumers on Supermarket

Authors- Khyal Jadav, Debashish Roy

Abstract-– Music plays a crucial role in shaping consumer behavior, particularly in retail environments like supermarkets. This study explores the impact of background music on customer shopping behavior, specifically whether a pleasant atmosphere created by music influences the time spent in the store. A well-curated musical ambiance has been linked to enhanced customer experiences, increased engagement with products, and potential boosts in sales. By analyzing consumer responses to different musical styles, tempos, and volumes, this research aims to determine if music encourages prolonged shopping durations and affects purchasing decisions. Data will be collected through surveys and observational methods to assess the correlation between music and consumer behavior. The findings of this study will provide valuable insights for retailers on how to optimize their store environments using music as a strategic tool to enhance customer satisfaction and drive revenue growth. This research contributes to understanding the psychological and behavioral effects of music in retail settings.

Beyond Screens: The Impact of Social Media Marketing

Authors- Akshay Dwivedi

Abstract-– The exponential rise of social media has fundamentally altered the marketing landscape. Social media platforms have become powerful tools for businesses to engage with their customers, promote products, build brand loyalty, and analyze consumer behavior in real-time. This paper explores the broad and deep impact of social media marketing (SMM) on businesses, focusing on its benefits, challenges, strategies, and future trends. Through a comprehensive literature review, data analysis, and relevant case studies, this study aims to assess how businesses can effectively leverage SMM to achieve sustainable growth, consumer trust, and competitive advantage. The study also emphasizes the behavioral patterns of consumers influenced by social media content, influencer marketing, and digital word-of-mouth.

Data Visualization for Agriculture in Tableau

Authors- Boya Ravi Teja, Muktha Vamshi, Lekkala Abhilash Reddy, Professor Dr Diana Moses

Abstract-– Agriculture remains the backbone of the Indian economy, contributing significantly to the country’s GDP and sustaining the livelihoods of millions. This project presents a comprehensive analysis of agricultural data from Telangana and Andhra Pradesh for the years 2022 to 2024 using Tableau. The dataset includes detailed information on crop production, income generated, and fertilizer usage at the district level. Additionally, the dataset is extended to include five other countries (USA, UK, Germany, Australia, and Canada) to enable comparative geospatial analysis through symbol maps. The dataset is structured in a long format to facilitate efficient visualization and analysis in Tableau. The data is categorized under three key measures: Production, Income, and Fertilizer Usage. Each data point is associated with a detailed time-based hierarchy, including Year, Quarter, Month, and Day to enable time-series analysis and trend forecasting. The project aims to help stakeholders gain insights into crop performance, resource allocation, and income variations over time.

DOI: /10.61463/ijset.vol.13.issue2.330

Dual Axis Solar Tracker Using Embedded Systems

Authors- Buduru Swarna bharathi, Pudi Bhavya Sri, M Chandana Priya, Shaik Ayza Hania, Nellepalli Bhuvaneswari

Abstract-– Solar power is at the forefront of that change as the world moves to cleaner energy sources. However, regular solar panels aren’t mobile. As they’re fixed in one place, they can’t follow the sun across the sky as it moves across the sky. This means that the panels will be less efficient at capturing sunlight, and will therefore miss out on opportunities for doing so. With a dual-axis solar tracker, we are trying to fix that problem. We used an embedded system based on an Arduino Uno to build a tracker that moves the solar panel side to side and up and down, maintaining always its position regarding the sun. Sensors enable the tracker to know where the sun is, and based on that, the panel’s position is moved around throughout the day. The result? A regular fixed solar panel showed us that we got a massive increase on how much energy the panel could capture. It’s simple, cheap, and proves how much opportunity there is to use smart tech to squeeze more out of solar power.

DOI: /10.61463/ijset.vol.13.issue2.331

Novel Helmet Design Integrated For Two Wheeler Management System

Authors- Mrs.E.Nivaditha, Sana Sreenivasulu Reddy, Sangana Raja Reddy, Borra Srihari

Abstract-– Bike riding is a lot of fun, but accidents happen. An accident is a unexpected, and unintended external action which occurs in a particular time and place, with no apparent and deliberate cause but with marked effects. India is one of the busiest countries in the world in terms of road traffic. The Indian road network, spanning over 5 million kilometers, carried almost 90% of the country’s passenger traffic and about 65% of the goods. India in 2019, an average of 414 a day or 17 an hour, according to the transport research wing of the ministry of road transport and highways. People choose motorbikes over car as it is much cheaper to run, easier to park and flexible in traffic. In India, 37 million people are owning two wheelers and 17 million units were sold in 2020. Since usage is high, accident percentage of two wheelers are also high compared to four wheelers. Motorcycles have high rate of fatal accidents than four wheelers. The impacts of these accidents are more dangerous when the driver involves in a high-speed accident without wearing helmet. So, wearing a helmet can reduce this number of accidents and may save the life. We have simple yet ready to use smart helmet system. A module affixed in the helmet, such that, the module will sync with the module affixed on bike and will also ensure that biker has worn Helmet. Additional feature of alcohol detection module will be installed on the helmet.

Review of Automatic Methods for Knowledge Graph Construction in Power Systems

Authors- Mr. Nagesh Kamble1, Dr. Pooja Paratwar

Abstract-– This review examines automatic methods for constructing knowledge graphs in power systems, highlighting their role in integrating heterogeneous operational data and prior knowledge to address the increasing complexity of modern grids. It summarizes key techniques—including knowledge extraction, representation, and reasoning—and explores their applications in fault diagnosis, equipment monitoring, and operational decision support. By systematically analyzing current research and practical implementations, the review demonstrates how knowledge graphs enhance efficiency, reliability, and intelligence in power system management. Future prospects for advancing smart grids and multidisciplinary integration are also discussed, emphasizing the transformative potential of knowledge graph technologies in the energy sector.

DOI: /10.61463/ijset.vol.13.issue2.332

Industrial IoT based Hazardous Gas Leakage Detection and Notification

Authors- C.S.Kannapiran, A.Arul, Associate Professor N.Shanmugavel, T.Santhosh

Abstract-– Nowadays, most of the industries are using Carbon Monoxide (CO) gas for manufacturing their products. These industries chemical synthesis, metal refining, food packaging, detergent manufacturing are handling hazardous gas inside their plants. In case leakage of gas is occurs during the production process, it’s difficult to find out the leakage. Therefore, a more reliable and robust detection is required to increase safety at working place. This article introduces a reliable, instant response and notification for such problem. When this system detects the gas leakage in industry, its turn on a buzzer for labour alert and turn on a exhaust fan for ventilation purpose. This system also sends a notification mail to the monitoring persons and the device is connected to internet so the gas value in the industry will posted to the internet for online monitoring.

Bone Cancer Detection Using Machine Learning

Authors- Katta Varshitha, Chittimani Hareesh, C Vijaya lakshmi, Karanam Sujitha, M.S.Yuvarathna

Abstract-– Bone tumors pose a significant health concern due to their potential malignancy and impact on the skeletal system. Early and accurate detection is critical for effective treatment and prognosis. This project leverages machine learning techniques to classify bone tumors using a publicly available dataset from Kaggle. The dataset includes various features relevant to tumor characteristics, which are analyzed and used to build predictive models. Four powerful classification algorithms Decision Tree, Random Forest, CatBoost, and XGBoost are employed to evaluate and compare performance in terms of accuracy, precision, recall, and F1-score. These models are trained and tested to differentiate between benign and malignant tumors, facilitating automated diagnostic support. The study aims to identify the most effective model for bone tumor classification, contributing to the development of intelligent diagnostic tools in the medical domain. The results highlight the potential of ensemble learning techniques in improving classification accuracy and supporting clinical decision-making.

DOI: /10.61463/ijset.vol.13.issue2.333

Brain Tumor Diagnosis through Deep Learning

Authors- Er. Susil Bhatta, Sharad Kumar Ghimire

Abstract-– Brain tumors are one of the most critical and life-threatening conditions that require prompt and accurate diagnosis for effective treatment. Recent advancements in deep learning have shown promising results in the field of medical diagnostics, particularly in the automated detection and classification of brain tumors from medical imaging data. This studyfocuses on developing a robust and reliable model for brain tumor detection and compare it with pre-trained deep learning models such as InceptionResNetV2. We utilized a comprehensive dataset of magnetic resonance imaging (MRI) scans to train and evaluate various Convolutional neural networks (CNNs). The pre trained models were fine-tuned to adapt to the specific characteristics of the brain tumor dataset, enhancing their ability to accurately identify and classify tumors. The performance of these models was compared using key evaluation metrics, including accuracy, precision, recall, and F1-score. The base model was trained for 20 epochs with a batch size of 32, while the InceptionResNetv2 was trained for 10 epochs under the same batch conditions. Notably, the InceptionResNetv2 exhibited significantly lower training and validation losses of 2% and 1%, respectively, compared to the base model’s losses of 7% and 6%. Both models achieved an overall accuracy of 95%, indicating their comparable effectiveness in classifying instances correctly. However, the base model outperformed the InceptionResNetv2 in precision (96% vs. 94%), suggesting a lower rate of false positives, whereas InceptionResNetv2 demonstrated superior recall (95% vs. 93%), indicating its ability to capture a greater proportion of actual positive cases. Both models attained an identical F1 Score of 94%, reflecting a balanced performance in terms of precision and recall. These findings suggest that while InceptionResNetv2 shows better training and validation performance with enhanced generalization capabilities, the base model is preferable when minimizing false positives is crucial.

DOI: /10.61463/ijset.vol.13.issue2.334

Weather Forecasting System Using python and Django Framework

Authors- Abhishek Shinde, Mangesh Janjal, Professor Shashibala Surapaneni

Abstract-– This paper presents the design and development of a weather forecasting web application using Python and Django, utilizing global meteorological data. The system leverages modern data analytics tools, visualization libraries, and machine learning techniques to analyze and predict atmospheric conditions. A user-friendly interface was created using Bootstrap, enabling users to retrieve real- time weather forecasts efficiently. The study aims to aid agriculture, travel, disaster management, and everyday users by delivering timely and accurate weather updates. In the era of climate variability and increasing demand for reliable weather information, accurate weather forecasting plays a crucial role across various sectors such as agriculture, transportation, and disaster management. This research presents the development of a real-time weather forecasting web application using Python and Django, named WeatherBug. The system integrates meteorological data sourced from Kaggle and Open Weather Map API to analyze and predict weather conditions including temperature, humidity, wind speed, and cloud cover. The application utilizes data preprocessing techniques, exploratory data analysis, and trend identification through machine learning algorithms such as neural networks and support vector machines. Visualizations are generated using libraries like Matplotlib and Seaborn, while the frontend is designed with Bootstrap for responsiveness. The project demonstrates the feasibility of deploying a reliable and user-friendly weather monitoring system, offering valuable insights into weather trends and aiding in timely decision-making. Future enhancements include mobile integration for broader accessibility and push notifications for real-time updates. [3]

A Review Article on the Role of Swasthavritta in Facing Diabetes

Authors- Assistant Professor Dr. Praveen Kumar

Abstract-– In Present time and modren lifestyles have led to a rapid rise in metabolic disorders, with Diabetes, or Madhumeha in Ayurveda, becoming a significant health concern. Diabetes mellitus is becoming fastest considerable diseases in the world. Diabetes affects more than a billion people worldwide, making it a public health priority, especially in developing countries like India, where lifestyle disorders are prevalent. India has being estimated with fastest growing population of Diabetics. Diabetes is a chronic metabolic disease characterized by elevated blood sugar levels, leading to potential damage to the heart, blood vessels, eyes, kidneys. It occurs when the body either doesn’t produce enough insulin or can’t effectively use the insulin it produces. It is a metabolic disorder may result in deficiency or dysfunction of the insulin production. The preventive measures in Ayurveda can prevent the disease. Ayurveda, a holistic health science, addresses both the prevention and treatment of Diabetes through Swasthavritta.

A Multi-Modal Hybrid Search System using RAG and Text2SQL for Semantic Information Retrieval

Authors- Anshul Kachhwal, Dr. Associate Professor Sandeep Gupta, Ankit Jangid, Abhishek Sharma

Abstract-– The Multi-Model Hybrid Search project is a comprehensive solution designed to enable advanced document retrieval and query processing using cutting-edge techniques in Natural Language Processing (NLP) and machine learning. The system integrates multiple models and frameworks, such as LangChain, Ollama, and Chroma, to facilitate document processing, text extraction, embedding generation, and SQL query handling. It allows users to upload various document types (PDF, TXT), extract textual content, generate embeddings, and store them in a vector database for semantic search. A Retrieval-Augmented Generation (RAG) approach enables users to query documents in natural language and obtain structured responses. Additionally, SQL queries can be generated from user input, providing a flexible mechanism to retrieve information from structured databases. The project offers an intuitive web interface built with Django, allowing users to interact with documents, ask questions, and generate structured responses. The integration of these models and tools aims to enhance the efficiency and accuracy of document management, querying, and information retrieval in real-time.

Diffusion-Based Face Generator

Authors- Divya Kanwar, Assistant Professor Anurag Anand Duvey, Rahul Jangid, Samarth Joshi

Abstract-– This research presents a ground-up implementation of a Diffusion-Based Face Generator, developed without the use of any high-level diffusion libraries such as Hugging Face’s diffusers. The model is based on the framework of Denoising Diffusion Probabilistic Models (DDPMs), implementations, this project provides full control over the noise schedule, sampling mechanism, and architecture design — offering greater flexibility and deeper transparency into the generative process.

Effectiveness of Ghana’s Renewable Energy Act, 2011 (Act 832) in Promoting Development and Use of Renewable Energy

Authors- Nana Baah Ampong, Maxwell Wahabu Manpaya

Abstract-– The aim of this study was to assess the effectiveness of Ghana’s Renewable Energy Act, 2011 (Act 832) in promoting the development and use of renewable energy. The study was conducted using a qualitative research design, through in-depth interviews, document analysis, and observations. With a sample size of 10 which was determined using point of saturation and the use of thematic analysis for the analysis of data, the study found that the implementation of the Renewable Energy Act has been ineffective and slow. The implementation of the Act is hindered by several key barriers, including slow approval processes, administrative inefficiencies, and a lack of local expertise, particularly in rural areas. The study also found that the Renewable Energy Act has led to positive environmental outcomes, including a reduction in greenhouse gas emissions and the promotion of cleaner energy sources. The Bui Power Authority’s hydro-solar hybrid system emerged as a successful example of resource optimization with minimal environmental impact. The study further discovered several challenges that impede the full implementation of Ghana’s Renewable Energy Act. Bureaucratic inefficiencies, including delays in permits and approvals, were noted as significant barriers to investment. The lack of technical expertise, particularly in rural areas, was also identified as a major concern affecting the sustainability of renewable energy projects. The study recommended the establishment of capacity- building programs focused on training technicians in the deployment and maintenance of renewable technologies that will empower local communities and enhance project effectiveness. Again, the Renewable Energy Fund should be fully operationalized to provide accessible financing options for developers.

DOI: /10.61463/ijset.vol.13.issue2.335

Satellite Based Iot For Remote Sensing AND Agriculture

Authors- Soham Khulage, Piyush Khonde

Abstract-– Agriculture is a vital sector for the survival of humanity. Various measures have been implemented to improve crop production. However, harsh environmental conditions and frequent pest infestations can lead to significant agricultural losses. In this context, integrating advanced technologies like sophisticated sensors with the Internet of Things (IoT) can boost agricultural output and reduce economic losses. Research conducted worldwide has effectively demonstrated the benefits of using IoT-enabled smart sensors to monitor environmental factors such as moisture, humidity, temperature, and soil compo- sition, all of which are crucial for crop growth. Automated sensors also measure greenhouse gases like carbon dioxide and methane. Additionally, smart farming allows for the assessment of nitrogen levels in the soil, helping farmers determine the appropriate amount of fertilizers to apply. IoT-enabled equipment and drones are valuable for accurately monitoring pest attacks and related diseases in crops. While smart farming holds great potential for the future, it does face challenges such as high implementation costs, data security concerns, and a lack of digital knowledge among farmers. Implementing special economic policies, en- hancing data encryption, and promoting digital literacy could facilitate the adoption of IoT-enabled smart farming in the future.

A Study on Employee Engagement and Its Impact on Job Satisfaction at Merlin Automation Solutions Pvt Ltd

Authors- Mr. S. Inbanathan, Associate Professor Dr. S. Madhiyarsi

Abstract-– This study examines how employee engagement impacts job satisfaction at Merlin Automation Solutions Pvt, Ltd. The research focuses on engagement factors like communication, recognition, organizational support, work-life balance, and growth opportunities. Data was collected from 100 employees through a structured questionnaire using convenience sampling. Statistical tools such as percentage analysis, weighted average, correlation, and ANOVA were used to interpret the findings. The results show a strong positive relationship between employee engagement and job satisfaction. Key factors such as effective communication (average score 4.12), recognition (4.05), and career growth (4.01) were found to significantly influence satisfaction levels. The study concludes that enhancing these engagement drivers can lead to improved employee morale, retention, and overall productivity.

Battery Charging Level Indicator: An Arduino-Based Battery Charging Level Indicator for Low-Power Applications

Authors- Prof. Ravi Bhushan, Fatima Hussain, Isha Agrawal, Ishika Singhal, Faheem Khan

Abstract-– This paper presents the design and implementation of a real-time battery charging level indicator based on the Arduino Uno microcontroller and a 16×2 LCD display. The system offers a low-cost, modular, and user-friendly solution for monitoring the State of Charge (SOC) in rechargeable batteries, particularly for low-power and educational applications. The methodology employs a voltage divider circuit that scales the input voltage from a 12V battery to a safe level readable by the Arduino’s 10-bit analog-to-digital converter (ADC). The measured analog value is then processed and mapped to a corresponding SOC percentage, which is displayed on the LCD interface in real time. Experimental validation using a 12V lead-acid battery shows the system accurately estimates battery levels with a deviation margin of ±0.1V compared to multimeter readings, and SOC estimation within ±2%. The proposed design demonstrates effective monitoring performance for voltage ranges between 11.8V (0%) and 12.6V (100%). The system is simple to calibrate, scalable, and can be customized for various battery chemistries with minor hardware and firmware changes.

Design and Implementation of Gas Leakage Detector in the Nigerian University: A Case Study of Ajayi Crowther University, Oyo

Authors- Babalola Olutola.O, Ogunkeyede Olabisi.Y, and Adeniran.Adeyemi

Abstract-– Liquefied Petroleum Gas (LPG) is widely used for domestic and industrial applications, particularly in cooking and heating. However, its highly flammable nature poses significant safety and environmental risks in the event of a gas leak. In Nigeria, the increasing frequency of LPG-related incidents underscores the urgent need for more reliable and responsive detection systems. This study presents the design and implementation of a microcontroller-based LPG leakage detection system that incorporates real-time monitoring and adaptive alarm mechanisms. Utilizing an Arduino Nano microcontroller and an MQ-6 gas sensor, the system continuously monitors gas concentration levels and provides dynamic feedback through an LCD display, LED indicators, and a buzzer. Predefined threshold values categorize gas levels into safety zones, enabling timely and appropriate responses to potential hazards. The system is powered by a reliable supply module, ensuring continuous operation. This research demonstrates the practical application of embedded systems in enhancing safety, offering a cost-effective and efficient solution for preventing gas-related accidents and promoting safer living environments.

Identifying and Analyzing Road Accident Hotspots in Maharashtra

Authors- Nilesh Surwase, Soham Zambre, Dr. Mahendra Kondekar

Abstract-– This research analyzes vehicle accident trends to enhance road safety, with a focus on implications for Maharashtra. Using a global Kaggle dataset, the study employs Python, Django, and Bootstrap for data cleaning, trend analysis, and visualization. Key findings show that higher speeds and adverse weather increase accident severity, while larger vehicles and curved roads contribute to frequent incidents. In Maharashtra’s context, these trends align with challenges like congested urban roads and monsoon-related accidents. Visualizations (pie charts, bar graphs, heatmaps) highlight actionable patterns. The study proposes Maharashtra-specific policy interventions, such as improved road signage and monsoon preparedness, laying a foundation for real-time and predictive safety systems.

Virtual Mouse Control Using Hand Gesture by Using Open CV

Authors- Pulluru Saathvika, Chennareddy Monika, Dr. Anumolu Lasmika, Dandu Harsha Vardhan, Veluru Purandhar

Abstract-– This research analyzes vehicle accident trends to enhance road safety, with a focus on implications for Maharashtra. Using a global Kaggle dataset, the study employs Python, Django, and Bootstrap for data cleaning, trend analysis, and visualization. Key findings show that higher speeds and adverse weather increase accident severity, while larger vehicles and curved roads contribute to frequent incidents. In Maharashtra’s context, these trends align with challenges like congested urban roads and monsoon-related accidents. Visualizations (pie charts, bar graphs, heatmaps) highlight actionable patterns. The study proposes Maharashtra-specific policy interventions, such as improved road signage and monsoon preparedness, laying a foundation for real-time and predictive safety systems.

DOI: /10.61463/ijset.vol.13.issue2.336

Speech to Text – Text to Speech Transcription for Indian Languages

Authors- Vishal Jadhav, Siddhesh Khedkar, Siddhesh Auti, Kushmeet Suri, Ms. Sapna Bhuskute, Dr. Nandini C. Nag

Abstract-– The project aims to develop a comprehensive system for Text-to-Speech (TTS) and Speech-to-Text (STT) conversion, providing users with a platform to seamlessly convert between text and speech formats. By leveraging user-friendly interfaces and powerful backend processing, the system offers multiple functionalities, including language translation, text synthesis, and speech recognition. This project is designed to cater to a wide range of user needs, from accessibility solutions for individuals with visual impairments to content creators looking to generate voiceovers or audiobooks. The system architecture integrates several modules, each tailored for specific functionalities. The TextShift module allows users to input text, select a language, and translate it to another language, while the TalkApp module focuses on converting text to speech using a speech synthesis engine. In parallel, the AudioBook module enables users to upload documents, which are then processed and converted into spoken audio. Furthermore, the BlendMaster module supports video processing, where users can input video files, add captions, and merge them with synthesized speech or translated text.

DOI: /10.61463/ijset.vol.13.issue2.337

Smart Voice – Activated Mobility Chair Using ESP32 Microcontroller

Authors- Abdul Hafeeza, B. Vijayb, Ch. N. S Charanc, Ch. Tarund, Assistant Professor K. K. Kishoree

Abstract-– This paper presents the design and development of a smart voice-activated mobility chair aimed at improving the independence of individuals with physical disabilities. The system is built using an ESP32 microcontroller, 12V DC gear motors, L298 motor driver, HC-05 Bluetooth module, and a lithium-ion battery. An Android-based Bluetooth controller app is used to receive voice commands, which are interpreted by the ESP32 to drive the wheelchair in the desired direction. The project integrates low-cost components with reliable speech control to create an efficient, user-friendly assistive mobility solution.

DOI: /10.61463/ijset.vol.13.issue2.338

Thermoelectric Transport Behavior of Gallium Arsenide Nano-Wire

Authors- Dr. M. P. Singh

Abstract-– In the paper, we have study thermoelectric transport behaviour of GaAs nanowire by the formulation based on the Boltzmann relaxation time approach for acoustic phonon scattering. For pure acoustic phonon scattering below 1000 K, the Seebeck coefficient of GaAs nanowires changes from a maximum value to a minimum value in the range of 180-235 µV/K. At low temperature below 100K GaAs nano wire semi metallic to semiconducting behaviour. At high temperature value of ZT at optimum carrier density 0.6 and operating temperature range is longer. The GaAs nanowire are useful for geothermal energy generation.

DOI: /10.61463/ijset.vol.13.issue2.339

Yogfit

Authors- Prasad Muli, Balaji Shardul , Nishant Marathe , Soham Mohite

Abstract-– Today’s web-based technology offers many online services in almost every field. Almost everything can be done online reducing the amount of tasks, cost, and efforts to a greater extent. This paper discusses about the development of the website on ‘Yoga’ named as ‘YOGFIT’. Yoga is a spiritual and physical science that originated in ancient India. Yoga is also one of the Vedic Shad darshans. Before pandemic yoga mostly done offline. But due to this lockdown situation one can’t go outside home. So, this website helps people to do yoga in their home. This website is useful for people in their day today life to remove stress. It is nothing but the virtual guide of yoga at any time. One can book class and perform yoga exercise.

DOI: /10.61463/ijset.vol.13.issue2.340

AI based Abnormalities Detection and Report Generation of Chest Xray: Methods and Applications

Authors- Mrs.Devi.V, Parameshwaran B, Naveen K, Nishanth P, Ramachandiran M.

Abstract-– The rising demand for medical imaging has strained the capacity of available radiologists, resulting in diagnostic delays and increased human error. Artificial Intelligence (AI), especially deep learning and Natural Language Processing (NLP), offers a solution through Automatic Medical Report Generation (AMRG). This paper surveys AMRG methods from 2021 to 2024, highlighting approaches, datasets, evaluation metrics, and performance-enhancing techniques while suggesting future research directions.

Automatic Toll Management System Using RFID Reader

Authors- B. Pavani, E. Susmitha, C. Harsha Vardhan, U. Venkateswarlu

Abstract-– The increasing number of vehicles on roads has raised the need for quicker, more efficient toll collection systems. Manual toll collection methods tend to cause traffic jams, delays, and wastage of fuel. This paper introduces an Automatic Toll Management System based on RFID technology to collect tolls automatically and minimize human effort. In this system, every vehicle is fitted with an RFID tag that contains a unique identification number. When a car comes to the toll booth, an RFID reader picks up the tag, and the system verifies the car’s account balance. In the event that the tag is acceptable and the balance is adequate, the toll figure is automatically subtracted, and the gate swings open via a servo motor. A traffic light module directs the driver via red, yellow, and green colours. The implementation is done by means of an Arduino Uno microcontroller, and the programming is done through Embedded C in the Arduino IDE. It offers contactless, speedier, and more reliable response to current requirements of toll collections. This project enhances traffic flow, decreases operational costs, and reduces the effects of human error. It is also scalable to future upgrades, including integration with SQL databases, mobile applications, and cloud- based monitoring, allowing it to be well-suited for smart city transit infrastructure.

DOI: /10.61463/ijset.vol.13.issue2.341

Wired to Thrive: Positive Emotional Attractors State and Peak Performance

Authors- Ms. Barpat Manisha, Associate Professor Dr. Nazia Sultana

Abstract-– Emotional attractors are the driving force for the individual behavior that is portrayed. The positive emotions arouse certain thoughts that drive a person to perform better and contribute in a good way towards the organization and himself., (Anita Howard,2006). The current research article has focused on the impact of PEA on employee behavior and performance in the software scenario, using the concept of ICT-intentional change theory by Elton Mayo. The findings of the study clearly showed that PEA by supervisor at the workplace instill positivity among the employees and make the work environment happy and optimal. The limitation of the study is that is was limited to city of Hyderabad.

Developing an Intrusion Detection System Using Generative Adversarial Networks (GANs) to Enhance Network Security

Authors- Samir Qaisar Ajmi, Sadiq Sahip Majeed

Abstract-– As the increasing reliance on digital networks and systems across various domains, cyberattacks have been a significant threat to system security and information integrity. Intrusion IDS is a critical tool in enhancing cyber security, but traditional systems have a propensity to struggle to handle new and complex patterns of attacks. This paper aims to develop an intrusion detection system based on generative adversarial networks (GANs) to enhance accuracy and efficiency of network threat detection. GANs is a novel deep learning method that contains two competing models: a generator, which generates imitated attack data mimicking Actual attacks, and a discriminator, which discriminates between actual and imitated data. With Competing interaction between the two models, very accurate attack data is produced and is used to improve the performance of IDSs. The research approach is based on building a GAN modeling with deep network layers to generate attack data that mimics actual attacks, while training the discriminator network to accurately detect this data. Such common data sets include NSL-KDD and CICIDS will be used to assess the effectiveness of the system, with performance evaluated with strict measures of precision, recall, and false alarm rate. The built the system should be able to detect dynamic and complex threats at a faster pace and superior to traditional systems. The research highlights the importance of providing a new solution to the intrusion detection problem leveraging the capabilities of artificial intelligence and deep learning, which aid to enhance network security and protection of them from modern cyber-attacks.

DOI: /10.61463/ijset.vol.13.issue2.342

Advancing Sustainable Development in E Mobility

Authors- Balam Ramprasad, Bandla Harish, Pinnapureddy Saisuresh Reddy, R.Vimalarasi

Abstract-– The project is to develop a comprehensive solution for eco-friendly driving on dynamic wireless charging lanes. This system focuses on optimizing speed control for electric vehicles (EVs) to maximize energy efficiency while maintaining seamless wireless power transfer. Key components include an infrared sensor to accurately position the vehicle within designated charging zones, and an RFID reader and tag system for secure authentication of vehicle access. A transmitter and receiver coil facilitate wireless power transfer, which is initiated via a relay once the vehicle is properly parked and authenticated. The ESP32 microcontroller enables real-time monitoring of vehicle status and power transfer, providing remote oversight and control.

Enhancing Exploratory Data Analysis through Fine-Tuned Generative AI: A Comparative Study with Traditional Machine Learning

Authors- Shiv Hari Tewari

Abstract-– Exploratory Data Analysis (EDA) plays a pivotal role in the early stages of the data science pipeline, providing critical insights into data quality, structure, and potential relationships among features. Traditional Machine Learning (ML) methods such as RandomForest and XGBoost have long been instrumental in EDA for feature selection, anomaly detection, and correlation analysis. However, these approaches often depend heavily on manual intervention, require domain expertise, and involve multiple iterations of preprocessing and parameter tuning. The advent of Generative AI—specifically transformer-based models such as GPT—has opened new avenues for automating and enhancing EDA tasks. These models, when fine-tuned for structured data interpretation, exhibit the ability to extract contextual insights, describe complex patterns in natural language, generate dynamic visualizations, and even propose feature transformations, all with minimal human input. This research presents a comparative analysis between traditional ML techniques and fine-tuned Generative AI models in the context of EDA.

Farming Assistance Using Web Service

Authors- Sowmiya P, Assistant Professor Dr.Krithika.D.R

Abstract-– Farming Assistance using web service” is a web application software in which the farmer can directly sell their product that they harvest with their own price which is affordable to buy for the user. This project has three major modules which includes admin, farmer and user. The admin module will monitor the price that is established by the farmer where the approval is done. The user feedback is monitored by the admin that is given by the user. The registered farmer in which the category of the agriculture product whether it is pulses, grains and cereals are categorized uses the farmer module. Only the farmer in which the approval is given by the admin and the buying procedure is done gives the price. The user module will help the user to view the agriculture product and they can buy if they wish. The user can also give the feedback about the product which is viewed by the admin for the product enhancement. The data are stored and managed only by the admin where the user details and the farmer details are been stored. This application will help the farmer to sell their product directly to the user without any intermediate.

DOI: /10.61463/ijset.vol.13.issue2.343

The Role of AI in HR Analytics and How It Affects Digitalization of Hiring

Authors- Assistant Professor Dr. A. Kamalakannan

Abstract-– Artificial intelligence is transforming human resources by enhancing the digital recruitment process. This paper explores AI’s impact on HR practices and hiring in the IT sector. The findings reveal that AI integration improves experiences for HR professionals and employees alike. By leveraging AI in HR analytics, teams can pinpoint areas where employees require skill development, select suitable candidates, and foster better management practices. The research also highlights challenges HR departments face in digitization, including data management and compliance in hiring. By addressing these issues through proper training and ethical AI implementation, organizations can maximize AI’s potential in HR operations. Clearly, AI will remain pivotal in shaping the future of human resources, improving talent management and creating a more efficient workplace for all.

Automated Car Toll System

Authors- Professor Lakshmi Narayan Gahalod, Aastha Sharma, Akshita Singh Kushwah, Anshika Thakur, Gargi Wadhwani

Abstract-– The research presents an Automated Car Parking System designed to improve parking efficiency by reducing human effort, space wastage, and time delays. Traditional parking systems are inefficient, requiring drivers to search for spaces manually, leading to traffic congestion and fuel wastage. This project automates the entire process using sensors, microcontrollers, and motors, ensuring smoother parking management. The system employs ultrasonic and infrared (IR) sensors to detect vehicles at entry points and check parking slot availability. The Arduino microcontroller processes this data, updating an LCD display to show real- time parking status and controlling servo motors to operate entry and exit gates. The system also integrates LED indicators, where green signals an available slot and red indicates an occupied one. This eliminates human intervention and enhances parking convenience.

Medical Diagnosis Assistant Using Vector Search

Authors- Dr. Uday Pratap Singh, Akshya Jain, Arjunsingh Kuldeepsingh Rana, Divya Meena

Abstract-– The integration of Artificial Intelligence (AI) into healthcare has revolutionized diagnosis systems into smarter, more efficient, and more accurate decision support tools. Among them, vector search approaches have emerged as a useful tool for developing intelligent medical diagnosis assistants. This review discusses the progress in medical diagnosis systems employing vector representations of clinical notes, patient information, and symptoms to present the most suitable diagnostic information. By incorporating medical knowledge and patient feedback into high-dimensional vectors, these systems allow for efficient similarity search, and hence real-time diagnostic recommendations can be supported. The article clarifies the utilization of embedding models such as BioBERT and ClinicalBERT, as well as frameworks such as FAISS for fast scalable vector lookups. Besides, it identifies recent advancements, weighs system performance in terms of accuracy and computational expense, and addresses issues like data bias, interpretability, and patient confidentiality. The review concludes by propounding potential future directions, evincing significance in explainable and light-weight models deployable in real clinical environments.

Algorithmic Capitalism and the Cost of Unsustainable Consumption

Authors- Shaina Bedi, Ramona Sharma

Abstract-– He rapid advancement of technology where consumer behaviour drives corporate profitability, has fundamentally reshaped the market structure in India. The emergence of online platforms has resulted in a growing use of self-learning recommendation algorithms that filter out content based on consumer preferences and suggest optimal purchase options. This Persuasive Technology has replaced deliberate choices with algorithmic recommendations, as outlined in Jean Baudrillard’s Theory of Hyperreality [Simulacra and Simulation, 1981]. This study explores how online applications use recommender systems to customize user experiences, using data such as past purchases and browsing history to deliver personalized product suggestions, targeted ads, and dynamic pricing. It focuses on cognitive shortcuts and biases consumers use, when exposed to these recommendations, drawing on Daniel Kahneman’s behavioural economics concepts such as loss aversion and decision heuristics. This paper will employ a quantitative methodology. An online survey will be conducted to collect data from a minimum of 150 participants aged 17-24, selected via convenience and snowball sampling. Exposure to recommendation algorithms and their subsequent impact on consumption will be measured through self-reported data on time spent engaging with recommended products on platforms such as Amazon and Flipkart, along with the frequency of the subsequent purchases made by consumers. The collected data will be statistically analysed using correlation and linear regression. Using the Fogg Behaviour Model [Fogg, B., 2009], the research will further highlight the shift in decision-making from rational processing to impulsive buying. This study aims to contribute to the limited research on the intersection of persuasive technology and consumer behaviour. It seeks to highlight the implications of recommendation algorithms and influencer trends on compulsive buying and digital consumerism, and its effect on sustainability in the digital era.This research falls particularly relevant in the Indian context due to the rapid digital transformation in the last decade driven by increasing internet access and smartphone penetration. The rise of social media and the influencer economy has made it crucial to understand the algorithm-driven consumerism and its impact on sustainability, economic inequality and mental health

Revolutionizing X-Ray Technology: The Emergence of Liquid Metal Jet Anode Systems

Authors- Mr. Melvin Jacob P, Ms. Binty P Biju, Assistant Professor Mr. Bibin Joseph

Abstract-– Traditional X-ray systems with solid anodes have long served in various domains such as medical imaging, materials analysis, and semiconductor inspection. However, these systems face critical limitations like thermal damage, limited brightness, and reduced lifespan due to target degradation. To address these challenges, a new generation of X-ray sources Liquid Metal Jet anode tubes has emerged, offering a promising solution by utilizing a continuously replenishing liquid metal target.

DOI: /10.61463/ijset.vol.13.issue2.344

MedConnect: Bridging Healthcare Access from PHCs to District Hospitals in Rural Areas

Authors- Vaishnavi Andhale, Dhanshri Chaudhari, Samruddhi Deshmukh, Sarita Khedkar

Abstract-– Med Connect is an Android-based healthcare ap- plication developed using Java and XML, designed to connect rural villages with district-level medical professionals. The system facilitates a seamless communication channel between Village Health Administrators (VHAs) and Doctors. VHAs register patients by entering essential details such as name, disease description, contact number, and by uploading medical reports in PDF format. These reports are securely stored in Firebase Cloud Storage, while all patient data is maintained in Firebase Realtime Database. Role-based access and secure logins are managed through Firebase Authentication, ensuring that VHAs and doctors have access only to relevant data. Once a VHA assigns a patient to a doctor, the doctor reviews the case and either approves or rejects it. Upon approval, the doctor can schedule an online consultation and send an automated SMS to the patient with a Google Meet link, date, and time of the appointment. Doctors can also track appointment history and status of all assigned patients. This system reduces the dependency on physical hospital visits and ensures timely medical intervention for remote communities. Med Connect is a step forward in bridging the healthcare gap in rural India, enabling quality treatment through a smart, cloud-backed mobile platform.

Eco-Friendly Synthesis of Zinc Oxide Nanoparticles Using Evolvulus Alsinoides Plant Extract

Authors- Vijay Raaj M, Dr. Poongodi A

Abstract-– Study introduces a web-based platform for the eco-friendly synthesis of zinc oxide (ZnO) nanoparticles using Evolvulus alsinoides plant extracts, adhering to the principles of green chemistry. The application automates and streamlines the synthesis process, improving efficiency, precision, and environmental sustainability. It incorporates essential functionalities such as resource management, material integration, additive incorporation, and quality assessment, all regulated through advanced algorithms and analytical techniques. Machine learning models, including the Lasso Regressor, are employed to optimize synthesis parameters and evaluate nanoparticle purity, ensuring adherence to high-quality standards. By reducing reliance on hazardous chemicals and utilizing natural, renewable resources, the project minimizes its environmental impact. The platform offers a transparent and systematic workflow, from initial material input to final quality verification, serving as a valuable resource for researchers and industry professionals. The synthesized ZnO nanoparticles, renowned for their photocatalytic and antimicrobial properties, find applications in air and water purification systems, medical devices, antimicrobial coatings, and the cosmetics sector, particularly in sunscreens and skincare products due to their UV-blocking capabilities. This project marks a significant step forward in nanoparticle synthesis, merging sustainable practices with advanced technology to foster innovation and environmental stewardship in scientific research.

DOI: /10.61463/ijset.vol.13.issue2.346

Lung Cancer Detection and Classification using Deep Learning

Authors- Okili Divya, Sudulakunta Rasaghna, Shaik Jameer, Thikkolla Narasimha Raju, Madduri Anil Kumar

Abstract-– This paper presents a deep learning-based approach for lung cancer detection and classification using CT scan images. Our method leverages Convolutional Neural Networks (CNNs) and Transfer Learning techniques to enhance accuracy and efficiency. Experimental results demonstrate that our model achieves a high classification accuracy compared to traditional approaches. This highlights the potential of Enhanced CNN models in transforming lung cancer diagnostics and emphasizes the importance of integrating advanced image processing techniques into clinical practice.

DOI: /10.61463/ijset.vol.13.issue2.347

A Novel Probabilistic Based Broadcasting Scheme for MANETs

Authors- Assistant Professor Dr. Gurpreet Singh

Abstract-– In Mobile Ad-hoc Networks (MANETs), efficient broadcasting is critical due to the dynamic topology and limited bandwidth. Traditional flooding techniques often result in high redundancy, contention, and collision, collectively known as the broadcast storm problem. This paper proposes a novel probabilistic broadcasting scheme that dynamically adjusts the forwarding probability based on local node density and residual energy. The proposed scheme reduces unnecessary retransmissions while maintaining a high reachability rate. Simulation results demonstrate that our method outperforms existing approaches in terms of packet delivery ratio, end-to-end delay, and network overhead, making it a viable solution for scalable and energy-efficient communication in MANETs.

DOI: /10.61463/ijset.vol.13.issue2.348

AI-Powered Career Counselling Platform

Authors- Ms. Ankita Nathe, Prajwal Sondawale, Prajwal Tijare, Sumit Devdhagle, Abhit Popate, Aman Mashkare

Abstract-– A career counselling platform powered by AI provides substantiated and data- driven results for career planning. exercising slice- edge AI algorithms, the platform assesses pupil interests and strengths, performance in academy and council as well as trends in the career request to give tailored recommendations on what course of action to take. scholars learn further about career possibilities in new fields, traditional places, and cross-disciplinary positions through interactive constructive assessments. Each pupil will have their own profile and the platform adapts to it via its algorithm suggesting courses, masteries to develop (soft chops), adulterous, and advanced education choices depending on what suits them stylish. The platform combines real- time data from subject matter experts and job request trends to give scholars with timely guidance that helps them make smart choices. acclimatized for the new generation, the AI- powered result aims to help youths come visionary career explorers and strategic itineraries not only in making education choices that align them towards a fulfilling career but also one which prepares them for the future of work.

DOI: /10.61463/ijset.vol.13.issue2.349

Employee Record System using Apache MySQL

Authors- V.Sivadurga, D.R.Krithika

Abstract-– The Employee Record System (ERS) is a fully web-enabled application created to process and store employee records within an organization. The system provides a structured means of keeping, accessing, and changing employee information such as personal, educational, and professional information. By replacing manual work systems with automated ones, the system reduces errors, guarantees data security, and allows instant access to important information. ERS increases productivity because the system allows administrators to manage employee’s information effectively, and employees can change their profile information and check their own records.

DOI: /10.61463/ijset.vol.13.issue2.360

Design and Implementation of a Solar-Based Wireless Charging Solution

Authors- Vishal Arjun Jadhav, Sachin Dattatray Salunkhe, Somnath Rama Sarvade, Sameer Sanjay Avhad

Abstract-– Access to electricity remains a critical global challenge, as electric power is essential for a wide range of applications—including mobile devices, laptops, cameras, sensors, satellites, offshore platforms, and biomedical implants. Wireless Power Transfer (WPT), also known as electromagnetic power transmission, offers a promising solution by enabling the delivery of electrical energy without physical connectors. This technology eliminates the dependency on copper and aluminium wires, potentially revolutionizing power distribution by enabling electricity to be transmitted wirelessly from generation sources to remote locations. Wireless Charging Systems (WCS) have emerged as a key application of WPT, particularly in electric vehicles (EVs) and other high-power domains. Compared to conventional plug-in systems, wireless charging offers superior convenience, reliability, and user-friendliness. This paper presents a comprehensive review and critical analysis of the principles, techniques, materials, and technologies underpinning wireless EV charging systems. Through an extensive evaluation of current research and technological trends, this study highlights the advancements, challenges, and future potential of wireless energy transfer in the context of sustainable mobility.

DOI: /10.61463/ijset.vol.13.issue2.350

Ultrasonic RADAR Using Arduino Uno

Authors- Aarya Kulkarni, Manasi Kadam, Aditya Agale

Abstract-– This project presents a cost-effective and easy-to-implement radar system utilizing an ultrasonic sensor, an Arduino Uno microcontroller, and a servo motor. It is designed to detect objects and provide real-time distance measurements displayed on a computer screen. The system has potential applications in security, obstacle detection, and proximity monitoring. Future improvements could include enhanced sensors, wireless connectivity, and integration with robotics or drones.

DOI: /10.61463/ijset.vol.13.issue2.351

Behavioual Analysis Using Gamification

Authors- A.Raagul Johnson, Assistant Professor Dr.Lipsa Nayak

Abstract-– This project explores the application of behavioral analysis using gamification principles in PHP. Behavioral analysis involves observing and understanding actions, reactions, and patterns to modify behavior or enhance outcomes, while gamification integrates game-like elements into non-game contexts to engage users and encourage specific behaviors. By leveraging PHP, a versatile programming language, this study employs tools such as data collection, behavioral tracking, and real-time analytics to measure user engagement and behavioral outcomes. Through the design and implementation of gamified systems, such as reward systems, points, leaderboards, and progress tracking, this study demonstrates how behavioral patterns can be influenced by game mechanics. A video that analyzes student stress levels using gamification in behavioral studies typically highlights the potential of using game mechanics and interactive elements to measure, track, and influence student stress. The video begins by defining student stress, which could include academic pressure, social anxiety, or time management issues. It introduces gamification as a concept: applying game-like elements (such as points, rewards, and challenges) to non-game contexts (such as education or behavioral analysis). The goal is to leverage gamification to help understand and mitigate stress through engaging, interactive experiences. The video demonstrates how behavioral analysis in gamification could be applied in classrooms, universities, or online learning environments. It may show case studies or simulations where students’ stress levels are monitored through gamified systems and how these systems promote better learning outcomes and emotional well-being. Benefits include reduced anxiety, enhanced focus, and a more enjoyable learning experience.

DOI: /10.61463/ijset.vol.13.issue2.352

Smart Medi Attender

Authors- Assistant Professor Dr. K. Nandhini, M. Venkata Praveen

Abstract-– In the dynamic healthcare industry of the modern age, it is a top concern in today’s situation to get timely and reliable support for patient care, especially for needy families who require urgent care. This study suggests Smart Medi Attender, an online platform to bridge the gap between seekers of medical attendants and possible caregivers in their respective localities. The system is based on a centralized database and real-time search facility to provide instant attender information based on area and city. HTML, CSS, JavaScript for frontend, Python (Fast-API) for backend, and MSSQL for database management are employed to build the platform. The platform offers a seamless and user-friendly experience. Users can register as attenders by submitting crucial details such as name, age, location, and availability, which are securely stored for future search and access. The platform also includes login and dashboard features for user profile management and presenting attender data effectively. This case study illustrates how Smart Medi Attender bridges the process of caregiver and seeker matching, making the process of short-term medical care more accessible and efficient in local communities.

DOI: /10.61463/ijset.vol.13.issue2.353

Veterinary Hospital Management System Using Ai Integrated Advisory

Authors- Assistant Professor Dr. K. Nandhini, M.Kiran

Abstract-– The Veterinary Hospital Management System (VHMS) is a digital solution that transforms the conventional veterinary clinic management into a streamlined, automated environment. Developed using Python’s Streamlit for the frontend and MySQL for backend data management, the system integrates modules such as pet registration, appointment scheduling, AI-powered diagnosis advisory, prescription handling, billing, and report generation. It is designed to reduce operational inefficiencies, minimize human error, and provide role-based access for security. The system also features an AI Treatment Advisor that suggests diagnosis based on selected symptoms, improving clinical accuracy. This paper elaborates the architecture, module-wise functionality, and future expansion potential of the VHMS.

DOI: /10.61463/ijset.vol.13.issue2.354

Smart Agriculture System in the Modern Era

Authors- M.Udhaya, Dr. K Kumutha

Abstract-– The technology has progressed very rapidly and influenced largely all sectors of life, including that of agriculture. It introduced the Smart Agriculture System, with its implications in much productivity, efficiency in use of resources, and sustainability. Thus, this paper focuses on understanding the arc of these elements-internet of things, artificial intelligence, automation, and data analysis-as related to the world of modern agriculture. The study discovers the advantages in precision farming, real-time monitoring, automated irrigation systems, and smart sensors in optimizing agricultural practice-initially commissioned by looking into the adoption challenges of smart agricultural solutions such as cost, technical know-how, and data security. It will support farmers to make data-driven decisions and reduce wastage of resources and improvement in crop yield with the adoption of these advanced technologies.

DOI: /10.61463/ijset.vol.13.issue2.355

Navigating the Intersection of Intellectual Property and Innovation: Trends, Challenges, and Future Directions

Authors- Assistant Professor Mr J.P Pramod, Annaram Kruthika, Alishety Poojitha

Abstract-– There is an extensive theoretical literature examining best-practice design for patent policy, but a significant shortage of empirical evidence regarding an important parameter in implementing such models: how research spending reacts to the intensity of patent rights. In recent decades, intellectual property rights (IPR) and their macroeconomic impacts have become prominent themes, capturing a significant interest both of policymakers and academics. The findings establish the presence of nonlinear relationship between IPRs and technological innovation that the shape is an inverted Ushaped curve, It’s established that innovations in developing nations grow in its IPRs by seeking the factor of human capital and economic development who have an indirect effect on enhancing the effect of IPR on innovation. I then briefly summarize the findings of two recent papers that have made headway in beginning to surmount these empirical challenges by marrying new data sets capturing biomedical research investments with new sources of variation in the effective intellectual property protection afforded to various inventions. While there exists a large theoretical and empirical literature, evidence on the effects of IPR protection on innovation and economic development is conflicting. In this paper we perform are view and meta-study of the literature on this subject, and determine that IPR overall exert a positive impact on innovation and growth. I outline how the implications of each study to patent policy and propose areas of further research.

DOI: /10.61463/ijset.vol.13.issue2.356

Advancing Electronic Waste Mitigation Using Vitrimer-Based Printed Circuit Board

Authors- Prasanth M, Professor DR.K. Nandhini

Abstract-– This project is about creating and applying a novel printed circuit board (PCB) that will minimize electronic waste. PCBs are not easy to recycle because of their fiberglass substrate, which is composed of woven glass fibers and epoxy resin. Researchers at the University of Washington have come up with a solution through the substitution of epoxy resin with a vitrimer polymer. This material retains the PCB’s shape and insulating properties under use but becomes gelatinous when subjected to a low-boiling-point organic solvent at the end of its life cycle. This facilitates the simple removal of intact glass fibers and electronic components. Surprisingly, 98% of the vitrimer and 91% of the solvent can be recovered and reused. These vPCBs are compatible with current manufacturing processes and are recyclable several times. Research indicates that recycling vPCBs can reduce global warming potential by 48% and carcinogenic emissions by 81% in comparison to traditional PCBs. The project ensures sustainability in electronic manufacturing via a scalable, green method that saves precious components and reduces waste. By embracing this technology, the electronics sector can greatly reduce its ecological impact, promoting additional innovation in electronic recycling and resource preservation globally.

DOI: /10.61463/ijset.vol.13.issue2.357

AI-Powered E-Library System Using React JS

Authors- R Deepak, Assistant Professor Dr.K.Kumutha

Abstract-– The AI-Powered E-Library System is an innovative digital solution designed to enhance book accessibility, management, and user experience through artificial intelligence. This system streamlines the process of browsing, borrowing, and managing digital books, offering personalized recommendations tailored to user reading preferences. It comprises three primary modules: Admin, User, and an AI-driven Recommendation System. The Admin module ensures efficient book inventory control and user activity tracking, while the User module enables registration, book exploration, borrowing, and wishlist management. The AI Recommendation System enhances engagement by analyzing user behavior and suggesting relevant books. Utilizing cloud storage, machine learning, and automation, this platform overcomes the limitations of traditional libraries, cutting operational costs and improving accessibility. Designed for educational institutions, research centers, and avid readers, this system offers a smarter, more user-friendly alternative to conventional library setups.

On Road Vehicle Breakdown Assistance

Authors- S.Yeshwanth, Assistant Professor K.Nandhini

Abstract-– It’s a desktop application fully tailored to assist company employees in efficiently managing customer service orders and handling vehicle issues. It also offers users a mechanic search feature location-wise, so that now they can find mechanics of different areas and use them. Control over user and mechanic details is given to the admin who can approve or block or even monitor mechanics. It makes locating an online mechanic easier, time-saving, and cost-saving. With the On-Road Vehicle Breakdown Assistance (ORVBA) system, individuals who live or travel in remote regions and meet with vehicle trouble may find a good answer. Registered customers can use the ORVBA medium to voice the verified mechanics, which brings with it trust and assurance. Only licensed and authorized mechanics are put into the system and monitored to avoid overcharging. Users can also give their feedback on the services they received in the ORVBA network.

DOI: /10.61463/ijset.vol.13.issue2.358

Text or Audio to Sign – Sign to Text

Authors- Singadi Sravanthi, Chimata vijaya rama raju, G Sirisha, Challa Lakshmipathi, Polepalli Medara Parikshith

Abstract-– The “AUDIO TO SIGN – SIGN TO TEXT” project represents a groundbreaking integration of technology in the field of communication and accessibility. It aims to bridge the gap between audio and sign language, subsequently converting sign language back into text. This bi-directional approach not only promotes inclusivity for the deaf and hard-of-hearing community but also fosters a deeper understanding and integration across different modes of communication. The project leverages advanced technologies like natural language processing (NLP), computer vision, and machine learning to interpret and translate languages. frameworks, enabling communication with hearing-impaired users. To close the communication loop, the system captures hand signs from users using MediaPipe’s real-time hand tracking. The gestures are classified and converted back into text, and then into spoken audio via the TTS module. The proposed system enhances accessibility, especially in public services, classrooms, and healthcare environments, by creating an AI- driven communication bridge between diverse users. It addresses key challenges such as real-time gesture recognition, contextual language understanding, and speech synthesis, all within a scalable and user-friendly framework. Keywords: Accessibility Technology, Sign Language Interpretation, Audio Processing, Speech-to-Text Conversion, Natural Language Processing (NLP), Computer Vision Deep Learning, Artificial Intelligence.

DOI: /10.61463/ijset.vol.13.issue2.359

Look-in-room: A Digital Platform for Home Furnishing

Authors- Tanzeela Zeba, Shrutika Adhau, Radha Mathane, Prathamesh Agarkar

Abstract-– Augmented reality (AR) is increasingly being adopted in online retail to enhance user experience and improve purchasing decisions. This paper presents the development of a web-based application that uses AR technology to help users visualize home furniture in their own spaces using 2D images. Unlike traditional methods that rely on static catalogs, our approach allows users to overlay PNG-format furniture images onto their real-world environment using a mobile or desktop camera view. The system is built using Python for backend processing and HTML/CSS for the frontend interface. This AR tool simplifies digital interaction by bridging the gap between online exploration and real space understanding through a user-friendly interface. A 2D AR application offers substantial value to furniture retailers through interactive experiences, tailored product recommendations, and increased customer confidence in their purchasing decisions.

DOI: /10.61463/ijset.vol.13.issue2.361

Detection Of Neurocognitive Impairment In The Elderly Using Deep Learning Algorithms

Authors- K. Purushothaman, Assistant Professor Dr.Lipsa Nayak

Abstract-– With the aging global population, neurocognitive disorders such as Alzheimer’s and various forms of dementia are becoming increasingly prevalent. These conditions not only affect patients’ quality of life but also pose significant challenges to healthcare systems. Early detection plays a crucial role in improving patient outcomes, yet traditional diagnostic methods often fail to identify the earliest signs of cognitive decline. This study proposes a deep learning-based approach that leverages both structured clinical data and MRI imaging to enhance diagnostic accuracy. We utilize the TabTransformer model to analyze structured clinical data, including demographic, lifestyle, and cognitive assessments, following an in-depth Exploratory Data Analysis (EDA) to better understand key features. In parallel, the ConvNeXt model processes MRI images to detect structural brain abnormalities associated with neurocognitive impairment. A multimodal learning strategy integrates insights from both models, allowing for a more comprehensive assessment of cognitive health. This approach improves predictive performance and provides a practical, AI- driven solution for early detection. Experimental results demonstrate that combining clinical and imaging data enhances diagnostic reliability, supporting healthcare professionals in making more informed decisions.

DOI: /10.61463/ijset.vol.13.issue2.362

Flood Detection and Warning for Water Bodies

Authors- Associate Professor Mr. Sanjay Balamwar, Assistant Professor Mr. Purushottam Chawke, Assistant Professor Ms. Prajakta Upadhye, Mr. Ayush Dongre, Mr. Pranay Wardhekar, Mr. Sahil Wankar, Mr. Saurabh Wankhede, Mr. Yash Tidke

Abstract-– Floods pose a significant threat to human life, infrastructure, and ecosystems, particularly in regions surrounding major water bodies such as rivers, lakes, and reservoirs. This study focuses on the development and implementation of a flood detection and early warning system tailored for water bodies. The system integrates real-time data collection using sensors (e.g., Water Level, Rainfall and Flow Rate Sensors), Satellite Imagery and weather forecasts with machine learning models to predict potential flood events. By leveraging IoT and cloud-based platforms, the system ensures timely data processing and dissemination of alerts to relevant authorities and at-risk communities. The proposed approach aims to minimize the impact of floods by enabling proactive response strategies, improving disaster preparedness, and enhancing resilience in vulnerable regions.

DOI: /10.61463/ijset.vol.13.issue2.363

Exploitation of Recommendation Framework for Inadequate Approach

Authors- B.Kishore, Professor R.Priya

Abstract-– Recommender Systems (RS) are systems through which users can find projects out of a large number of possibilities and retrieve projects. They assist organizations in utilizing resources optimally in overcoming cold-start and sparsity problems. To address the problems mentioned above, we employ a gradient classifier algorithm, which refers to a predictive regression method where a loss function is minimized by iteratively selecting the function reflecting against the negative gradient. This method always identifies some problem and provides a timely solution associated with project completion.

DOI: /10.61463/ijset.vol.13.issue2.364

Dog Teeth Protection Design: Improved Dog Bite Function And Prevent Rabies Infection

Authors- Souhridya Bhattacharjee, Satyajit Chakrabarti, Saprativ Malakar, Dr. Prabir Kumar Das, Rituparna Saha

Abstract-– Rabies migration through dog stitching poses considerable risk to humans and other animals. In this article, we consider a study that will help new dog dental protection developments to help dogs with essential biting and purchasing capabilities, while also preventing teeth penetration from human skin. Protectors also aim to reduce the risk of rabies infection by chewing on exposure to infected saliva. We discuss the design capabilities, materials, practical considerations of such devices, and the impact on public health, particularly in areas with rabies. The study highlights how important it is to integrate health security with animal functioning and ultimately provide precautions to combat rabies.

The Influence of Artificial Intelligence on Algorithmic Trading and Its Impact on Predicting Financial Market Trends

Authors- Janhavi Bhattad

Abstract-– This paper explores the transformative role of Artificial Intelligence in algorithmic trading, focusing on its impact on financial market predictions. It investigates how AI-driven models, particularly those employing machine learning techniques such as deep learning and reinforcement learning, have demonstrated remarkable capabilities in forecasting market trends and executing trades with speed and precision surpassing human capabilities. The integration of AI in finance enhances market efficiency by swiftly analyzing vast datasets to predict price movements, identify fraudulent activities, and manage risks, thereby reshaping investment strategies and risk management. However, the deployment of AI in algorithmic trading also introduces critical challenges, including concerns about data quality, model interpretability, and potential systemic risks. The paper also addresses the ethical considerations surrounding AI’s autonomy in trading decisions, emphasizing the necessity for robust regulatory frameworks and transparency to ensure fairness and stability in financial markets. By examining both the advancements and challenges, this study aims to provide a comprehensive overview of AI’s current and future role in financial market predictions, contributing to a deeper understanding of its implications for investors, regulators, and the overall financial ecosystem.

Face Recognition Based Attendance System

Authors- Mr Gaurav Rane, Dr. Jasbir Kaur, Assistant Professor Ms.Sandhya Thakkar

Abstract-– This paper presents the development and evaluation of a Face Recognition Based Attendance System leveraging deep learning and real-time computer vision. This system automates student attendance using facial biometrics, offering a contactless, efficient, and reliable alternative to traditional methods. Built using OpenCV, the face recognition library, and Firebase integration, the system achieves high recognition accuracy and real- time data synchronization. This paper details the methodology, implementation, evaluation, and identifies future directions for enhancing functionality and scalability.

DOI: /10.61463/ijset.vol.13.issue2.365

Respiratory Disease Detection Using Machine Learning for Sound Classification

Authors- M Divya, Dr. C. Meenakshi

Abstract--Respiratory sounds play a meaningful part in surveying of aspiratory wellbeing and recognizing respiratory disarranges at an early arrange. The emergence of artificial intelligence (AI) has encouraged the utilization of machine learning (ML) methods so as to analyze respiratory irregularities, inclusive of conditions like asthma, pneumonia, bronchiolitis, and also chronic obstructive pulmonary disease (COPD). Conventional auscultation strategies remain important, but subjectivity, variations in clinical translation, and irregularities in sound quality often limit their adequacy. Subsequent advances in computational techniques have upgraded demonstrative precision. These advances empower mechanized discovery of various unusual lung sounds, for example, wheezes and crackles. This investigation examines if choice tree-based classification models can diagnose respiratory illnesses, reaching an impressive accuracy of 90%. The aforementioned highlights underscore the possibility of AI-powered symptomatic tools in yielding durable results and advancing respiratory medicine.

DOI: /10.61463/ijset.vol.13.issue2.366

Career Guidance App for Students – AI Assisted

Authors- R Shiyam Ganesh, Dr. Kumutha

Abstract--At present social network sites are part of the life for most of the people and want to free-frank with new friends. Every day several people are creating their profiles on the social network platforms and they are interacting with others independent of the user’s location and time, Technology is associated with online social networks which has become a part in every one’s life in making new friends and keeping friends, and share a personal information with other users their interests are known easier. Traditional methods cannot differentiate between real and fake accounts efficiently. To analyze, who are encouraging threats in social networking sites with user profiles. There are numerous cases where produced accounts have been effectively distinguished utilizing machine adapting techniques characters made by people.

DOI: /10.61463/ijset.vol.13.issue2.367

Tracking of Biochar Production

Authors- Madhar Umar I, Professor Dr.V. Sumalatha

Abstract--At present social network sites are part of the life for most of the people and want to free-frank with new friends. Every day several people are creating their profiles on the social network platforms and they are interacting with others independent of the user’s location and time, Technology is associated with online social networks which has become a part in every one’s life in making new friends and keeping friends, and share a personal information with other users their interests are known easier. Traditional methods cannot differentiate between real and fake accounts efficiently. To analyze, who are encouraging threats in social networking sites with user profiles. There are numerous cases where produced accounts have been effectively distinguished utilizing machine adapting techniques characters made by people.

DOI: /10.61463/ijset.vol.13.issue2.392

Improving Code Quality and Workflow: Integrating Static Analysis Tools with Version Control Systems

Authors- Aditya Kumar Singh, Deepanshu Tomar and Prof. A. Viswanathan

Abstract--In software development, it’s critical to minimise errors and efficiently maintain code quality. With an emphasis on how they might be combined with version control systems like Git’s pre-commit hooks, this study compares many well-known static code analysis tools, including FindBugs, SpotBugs, and PMD. The research assesses the performance, Among the additional advantages and disadvantages demonstrated by the varied strategies used for various static code analysis tools during this process were accuracy, integration, and ease of use. Additionally, we offer comprehensive instructions for incorporating these analysers into Git hooks along with recommended ways to make it more efficient. The findings show that maintaining code quality and improving development workflow efficiency need automating code analysis.The results indicate that automating the analysis of codes is essential for preserving their quality as well as enhancing development workflow efficiency.

DOI: /10.61463/ijset.vol.13.issue2.368

Experimental Investigation of the One-Way Speed of Light Using Entangled Quantum Clocks

Authors- Shamit Bipin Bulunge

Abstract--Determining the one-way speed of light without relying on prior assumptionsabout clock synchronization remains a fundamental challenge in physics. Thismanuscript proposes an experiment to investigate this problem by measuring thetime-of-flight of a light pulse over a known distance using entangled quantumclocks as synchronized time references. This approach aims to potentiallycircumvent the conventionality of clock synchronization inherent in classicalmethods. The experimental design emphasizes the generation and distribution ofhigh-fidelity entanglement, continuous monitoring of the quantum clocks, andrigorous verification of their temporal correlation through cross-correlationanalysis. While acknowledging the significant technological hurdles and thephilosophical complexities of defining absolute time, this study outlines anovel pathway towards a potentially assumption-lean measurement of the one-wayspeed of light.

AI-Enabled Real-Time Health Monitoring for Elderly Care: A Smart Solutions Approach

Authors- Prabhu Prasad

Abstract--As the global population ages, the demand for effective healthcare solutions for the elderly continues to rise. Traditional healthcare systems often struggle to provide personalized, real-time care for aging individuals, leading to increased healthcare costs and burdens on caregivers. Artificial Intelligence (AI), particularly in the form of real-time health monitoring systems, offers a promising solution to address these challenges. By leveraging AI technologies such as machine learning, wearables, and Internet of Things (IoT) devices, healthcare providers can monitor the health of elderly patients in real-time, detect health issues early, and provide personalized care recommendations. This paper explores the role of AI-enabled real-time health monitoring systems in elderly care, focusing on their applications in chronic disease management, fall detection, medication adherence, and overall health management. Furthermore, it discusses the benefits, challenges, and future directions of AI in elderly care, emphasizing the importance of collaboration between healthcare providers, technology developers, and policymakers.

DOI: /10.61463/ijset.vol.13.issue2.369

AI-Driven Optimization of Supply Chain Processes: Enhancing Efficiency and Reducing Costs

Authors- Noushad Pasha

Abstract--As the global population ages, the demand for effective healthcare solutions for the elderly continues to rise. Traditional healthcare systems often struggle to provide personalized, real-time care for aging individuals, leading to increased healthcare costs and burdens on caregivers. Artificial Intelligence (AI), particularly in the form of real-time health monitoring systems, offers a promising solution to address these challenges. By leveraging AI technologies such as machine learning, wearables, and Internet of Things (IoT) devices, healthcare providers can monitor the health of elderly patients in real-time, detect health issues early, and provide personalized care recommendations. This paper explores the role of AI-enabled real-time health monitoring systems in elderly care, focusing on their applications in chronic disease management, fall detection, medication adherence, and overall health management. Furthermore, it discusses the benefits, challenges, and future directions of AI in elderly care, emphasizing the importance of collaboration between healthcare providers, technology developers, and policymakers.

DOI: /10.61463/ijset.vol.13.issue2.370

Machine Learning for Sustainable Agriculture: Enhancing Crop Yield Predictions and Resource Management

Authors- Lakshmi. N

Abstract--Sustainable agriculture is essential to ensure the efficient use of resources and meet the growing global demand for food. With the increasing challenges posed by climate change, population growth, and environmental degradation, there is an urgent need for innovative solutions that improve crop yield and resource management. Machine learning (ML), a subset of artificial intelligence (AI), offers significant potential to transform agriculture by providing data-driven insights into crop performance, soil health, weather patterns, and resource allocation. This paper examines the role of machine learning in sustainable agriculture, with a focus on its applications in crop yield prediction, pest and disease management, soil quality monitoring, and water usage optimization. Additionally, it explores the benefits of ML in enhancing precision agriculture and reducing the environmental impact of farming practices. Despite its potential, the adoption of machine learning in agriculture faces challenges related to data quality, infrastructure, and farmer education. The paper concludes with a discussion of the future of ML in agriculture, highlighting the need for continued research and collaboration between technology providers, farmers, and policymakers.

DOI: /10.61463/ijset.vol.13.issue2.371

Deep Learning Approaches for Natural Disaster Prediction and Response Planning

Authors- Fasal Ahmed

Abstract--Natural disasters, such as hurricanes, earthquakes, floods, and wildfires, pose significant risks to human life, property, and the environment. Predicting and mitigating the impact of these events is crucial for disaster management and response planning. Traditional disaster forecasting and response strategies often face limitations due to the complexity, unpredictability, and scale of these events. Deep learning, a subset of artificial intelligence (AI), offers promising solutions by leveraging vast amounts of data and advanced algorithms to enhance prediction accuracy, early warning systems, and response strategies. This paper explores the application of deep learning techniques, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), and long short-term memory (LSTM) networks, in predicting natural disasters. It discusses how these models can process diverse data sources such as satellite imagery, sensor networks, weather patterns, seismic activity, and social media to generate actionable insights. Furthermore, the paper examines the role of deep learning in improving disaster response and recovery efforts through automation, resource allocation, and real-time decision-making. Despite the potential, challenges such as data quality, model interpretability, and computational resources are discussed, along with future research directions in integrating deep learning with other technologies to create more resilient disaster management systems.

DOI: /10.61463/ijset.vol.13.issue2.372

AI-Powered Personalization in E-Commerce: Transforming Consumer Experience Through Data Insights

Authors- Swamy.M

Abstract--The e-commerce industry has experienced exponential growth over the last decade, driven by increasing internet penetration, technological advancements, and a shift in consumer behavior. One of the key factors contributing to this growth is the ability of e-commerce platforms to deliver personalized experiences, which can enhance consumer satisfaction and drive sales. Artificial Intelligence (AI) plays a crucial role in enabling personalized shopping experiences by analyzing vast amounts of data, such as user behavior, demographics, purchase history, and preferences. This paper explores the various AI-driven techniques used in e-commerce to enhance personalization, including recommendation systems, natural language processing (NLP), and predictive analytics. Additionally, it examines the impact of AI on customer engagement, conversion rates, and customer loyalty. While AI has the potential to revolutionize e-commerce personalization, challenges such as data privacy concerns, algorithmic biases, and the need for transparency are also discussed. The paper concludes with a look at the future of AI-powered personalization in e-commerce, emphasizing the importance of balancing personalization with consumer trust and privacy.

DOI: /10.61463/ijset.vol.13.issue2.373

Cross-Modal Ai Architectures For Audio-Visual Emotion Recognition

Authors- David E

Abstract--Emotion recognition technology has become a critical area in affective computing, aiming to enable machines to understand human affect through various signals. While unimodal systems based on either audio or visual data provide partial insights into emotions, they fall short in capturing the multimodal nature of human emotional expression. Cross-modal Artificial Intelligence (AI) architectures integrate both audio and visual inputs to provide a more robust and holistic emotion recognition capability. This paper explores the foundational technologies of cross-modal emotion recognition, including deep learning, feature fusion, and multimodal representation learning. Practical use cases in virtual assistants, mental health diagnostics, education, and customer service are discussed, alongside real-world applications. Ethical concerns regarding privacy, surveillance, and emotional manipulation are examined, as well as limitations such as model interpretability, cultural variability, and data alignment challenges. Finally, the paper highlights future prospects including adaptive emotion-aware systems, neurosymbolic AI, and the growing importance of context in emotional modeling. Cross-modal AI stands as a powerful tool for creating emotionally intelligent systems capable of understanding and interacting with humans in deeper, more meaningful ways.

DOI: /10.61463/ijset.vol.13.issue2.373

Vision-Based Ai Models For Wildlife Conservation And Species Tracking

Authors- Gokul Nandan A

Abstract--The accelerating loss of biodiversity due to habitat destruction, climate change, and poaching necessitates the development of advanced conservation tools. Traditional wildlife monitoring methods—manual tracking, field surveys, and camera traps—are often labor-intensive, costly, and geographically limited. Vision-based Artificial Intelligence (AI) models have emerged as transformative technologies capable of automating and scaling wildlife observation through image and video data. Powered by deep learning and computer vision, these models can perform species recognition, behavioral analysis, population estimation, and illegal activity detection with unprecedented speed and accuracy. This paper explores the foundations of vision-based AI in wildlife conservation, including key algorithms and data sources. It details various use cases such as species identification, habitat monitoring, and poacher detection. Real-world applications from regions including Africa, South America, and Southeast Asia demonstrate how these models are revolutionizing conservation strategies. The paper also discusses ethical and regulatory considerations related to data usage, community involvement, and potential surveillance misuse. Technical challenges such as data quality, model bias, and deployment in remote areas are evaluated. Finally, future innovations such as edge AI, explainable models, and citizen science integration are explored. Vision-based AI holds immense potential to support biodiversity preservation and empower conservationists with precise, scalable, and non-invasive monitoring tools.

DOI: /10.61463/ijset.vol.13.issue2.376

Centralized Monitoring System for Faulty Street Light Detection and Location Tracking

Authors- Ankush Bokade, Pratik Gabhane, Sanchit Samudre, Shreya Himane, Professor Y.M. Babnekar

Abstract--The main objective of this project is to create an intelligent and automated street light fault monitoring system. The system proposed here has been conceptualized to counter the traditional methods of monitoring street lights, which are largely dependent on human visits or citizen complaints. With the use of contemporary technologies like sensors, microcontrollers, and IoT platforms, the system can monitor continuously the working condition of street lights and detect at once any fault. This real-time street light fault detection system will issue automatic notifications to a centralized platform whenever a streetlight goes dark. This helps in quicker response times, optimizes maintenance schedules, and guarantees that faulty street lights can be repaired promptly. The system not only increases road safety and nighttime visibility for citizens but also contributes to lowering maintenance costs overall as well as unwarranted inspections.

DOI: 

Daily Expenses Tracker System

Authors- Ramya. S, Assistant Professor DR. Lipsa Nayak

Abstract--In today’s fast-paced world, managing personal and organizational finances has become a critical aspect of maintaining financial health. A major challenge faced by individuals and businesses alike is the inability to effectively track and manage their expenses and income. Without proper oversight, people often struggle to stay within their budget, leading to overspending and financial instability. This problem is exacerbated by the lack of intuitive, easy-to-use systems that enable real-time tracking and analysis of financial data. To address these challenges, we propose the development of an Intelligent Expense Tracker web application, designed to help individuals and organizations efficiently monitor their monthly expenses and receivables. The primary objective of the system is to provide users with a simple yet powerful tool to track their daily spending and ensure that they stay within their set budget for the entire month. By offering users the ability to record and categorize their expenses, this application empowers them to make informed decisions about their financial habits./10.61463/ijset.vol.13.issue2.378

Interactive Portal for Secure Book Trading

Authors- Assistant Professor Dr. K.Kumutha, T.Narane

Abstract--Secure Book Trading Platform is an innovative e solution to transform the process of online book buying, selling, and trading. Designed with a safe and easy-to-use web interface, the platform features user authentication, book listing, improved search and filtering, transaction processing, and real-time notifications. Focusing on data security along with secure payment processing, the system uses role-based access control and encrypted data handling to secure the user data. The platform also facilitates peer-to-peer trading while ensuring a transparent and reliable environment. This paper presents the system architecture, major features, and its future scope for growth in the domain of digital book commerce./10.61463/ijset.vol.13.issue2.379

Autonomous Drones and Artificial Intelligence: A New Era of Surveillance and Security Applications

Authors- Okpala Charles Chikwendu, Udu Chukwudi Emeka

Abstract--The integration of autonomous drones with Artificial Intelligence (AI) is revolutionizing surveillance and security, enhancing monitoring, threat detection, and rapid response capabilities. This study examines the role of AI-driven drones in modern security systems, highlighting their potential to improve situational awareness, reduce human intervention, and optimize operational efficiency. With machine learning algorithms, computer vision, and real-time data analytics, autonomous drones can autonomously detect anomalies, track suspicious activities, and respond to security threats with precision. These advancements are particularly valuable for border security, law enforcement, critical infrastructure monitoring, and disaster response.Despite their benefits, AI-powered drones face challenges such as ethical concerns, privacy issues, regulatory constraints, and cybersecurity risks. This research explores the legal and ethical implications of autonomous surveillance, reviewing current policies and governance to ensure responsible use. It also addresses technical limitations, including power constraints, environmental adaptability, and AI biases in threat assessment, while suggesting solutions to improve reliability and security.Through case studies and analysis of emerging trends, this study provides an evaluation of the evolving role of autonomous drones in security operations. The findings contribute to discussions on responsible AI use, regulatory policies, and future innovations in autonomous surveillance. Ultimately, this research emphasizes the need for a balanced approach that maximizes the benefits of AI-driven drones while addressing their ethical, legal, and technical challenges.

Artificial Intelligence- Fake Currency Detection System

Authors- M. Varalakshmi, Bushra Tahseen , S. Bhargavi, M. Navya, N. Bhanu Keerthana, S. Sravani

Abstract--In recent years a lot of fake currency note is being printed which have caused great loss and damage towards society. So, it has become a necessity to develop a tool to detect fake currency. Our proposed system will follow an approach that will detect fake currency note being circulated in our country by using their image. Our project will provide required mobility and compatibility to most peoples as well as credible accuracy for the fake currency detection. We are using image processing to make this application efficient. This project will find some important features in notes using image processing process which will determine the originality of the currency note. By using this tool fake notes can easily identify and minimize the count of fake notes in the market.

Detecting The Movement of Objects with Webcam and Alert Using Machine Learning

Authors- P. Swathi, Shaik Haseena, M. Lahari, V. Swetha, R. Sujitha

Abstract--The Movement Detection Alert System is a software application developed using Python 3, OpenCV, and pyttsx3 that enables the detection of moving objects in a video stream or webcam feed. The system captures images of the detected objects and provides real-time alerts to notify the user about the presence of movement. The system’s main objective is to enhance security and surveillance by providing an automated mechanism to detect and monitor any unauthorized movement or activity in a specific area. It can be used in various scenarios, such as home security, office monitoring, or public space surveillance. To achieve movement detection, the system utilizes the computer vision library OpenCV, which provides a robust set of functions for image and video processing. By continuously analyzing consecutive frames of a video stream, the system detects changes in pixel intensities and identifies regions where motion has occurred. Upon detecting movement, the system captures images of the moving object using the connected webcam or video source. These images can serve as evidence or references for further analysis. Additionally, the system generates an alert to inform the user about the detected movement. For this purpose, the pyttsx3 library is used to convert text to speech, allowing the system to audibly notify the user. The Movement Detection Alert System offers flexibility in configuration. Users can set parameters such as sensitivity levels, the region of interest, and the notification mechanism according to their specific requirements. The system can be run on various platforms and operating systems that support Python and the necessary dependencies. Overall, the Movement Detection Alert System provides a reliable and customizable solution for detecting and capturing moving objects while delivering real-time alerts. Its application can significantly contribute to improving security measures and reducing the response time to potential threats or suspicious activities.

Deep Fake Detection a Systematic Literature

Authors- P. Poojitha, M Veeresh, M. Devasree, M. Supriya, P. Rakshitha

Abstract--Over the last few decades, rapid progress in AI, machine learning, and deep learning has resulted in new techniques and various tools for manipulating multimedia. Though the technology has been mostly used in legitimate applications such as for entertainment and education, etc., malicious users have also exploited them for unlawful or nefarious purposes. For example, high-quality and realistic fake videos, images, or audios have been created to spread misinformation and propaganda, foment political discord and hate, or
even harass and blackmail people. The manipulated, high-quality and realistic videos have become known recently as Deepfake. Various approaches have since been described in the literature to deal with the problems raised by Deepfake. To provide an updated overview of the research works in Deepfake detection, we conduct a systematic literature review (SLR) in this paper, summarizing 112 relevant articles from 2018 to 2020 that presented a variety of methodologies. We analyze them by grouping them into four different categories: deep learning-based techniques, classical machine learning-based methods, statistical techniques, and blockchain-based techniques. We also evaluate the performance of the detection capability of the various methods with respect to different datasets and conclude that the deep learning-based methods outperform.

Short Term Arrival Delay Time Prediction in Freight Rail
Operations using Data-Driven Models

Authors- P. Bharath Kumar Reddy, C Mohammed Gulzar, S. Abuzar, K. Eranna, P. Vijeyudu, G. Pavan Kumar

Abstract--Short-term arrival delay prediction is a critical aspect of optimizing freight rail operations, enabling operators to mitigate disruptions and improve operational efficiency. This study investigates a set of data-driven models to predict short-term arrival delay times in freight rail operations. Using historical train operation data from the National Railway Company of Luxembourg (CFL) and multiple European freight terminals, we develop and compare machine learning models, including linear regression, k-nearest neighbors, random forest, and gradient boosting techniques. Among these, the Light Gradient Boosting Machine (Light GBM) model demonstrated superior predictive performance, achieving the highest accuracy in forecasting arrival delays. The study identifies key influential features such as departure delay time, trip distance, and train composition, which significantly impact the delay prediction. Furthermore, the Shapley Additive Explanations (SHAP) method is employed to interpret model results and analyze the relative importance of input features. The proposed predictive model can serve as a ShortTerm Decision Support System (STDSS) for railway operators, offering real-time delay estimates to optimize scheduling and minimize operational disruptions. The findings contribute to enhancing the reliability and efficiency of freight rail transport by providing a robust framework for delay prediction and mitigation strategies.

Personal Voice Assistant

Authors- M. Raviteja, Samson Paul, P. Dinesh Kumar, S. Abdulla, B. Subhash, Z. Abhilash Surya

Abstract--Today there is huge Advancement in the Technical field which is increasing day by day. In early days there were only computer systems where we were able to perform only few tasks, but today new technologies like machine learning, artificial intelligence, deep learning, and few some others have made computer systems so advance that we can perform any type of task with them. In recent years, Artificial Intelligence (AI) have done remarkable progress and its Capability is increasing day by day. One of the application Area of AI is Natural Language Processing (NLP). Natural Language Processing (NLP) helps Humans to communicate with the computer system in their own Language. For example, Voice Assistant. Various voice assistants were developed and they are still being improved more for better performance to overcome struggling of humans to interact with their machine. we are trying to develop a voice assistant using python which will help user to perform any type of task without interaction with keyboard. The aim of this paper is to study how voice assistants behaves smartly and can be used to get everyday work done and also be used for educational purpose also.

AI based Internet of Things (IoT) Platform for Structural Health Monitoring (SHM)

Authors- Dr. G.U. Kharat, Prof. R. S. Bansode, Poonam D. Sable

Abstract--Structural Health Monitoring (SHM) is a crucial component in ensuring the safety and reliability of infrastructure. With the increasing risks of environmental impacts on structures, such as bridges, buildings, and industrial assets, an automated system for monitoring and detecting structural issues is imperative. This project focuses on developing a cost-effective SHM system using a Raspberry Pi microcontroller and environmental sensors for temperature and humidity measurement. The system continuously monitors these environmental conditions and employs a beta regression model to analyze the impact of these factors on structural integrity. The proposed SHM system aims to address the limitations of existing monitoring techniques, which often rely on infrequent manual inspections or costly specialized equipment. This automated approach offers a scalable, real-time, and adaptive solution for various types of infrastructures. The system also enables timely alerts and notifications, thus allowing stakeholders to conduct proactive maintenance and avoid potentially catastrophic failures. By bridging the gap between environmental monitoring and predictive maintenance, the project contributes to enhanced safety, reduced costs, and increased operational efficiency in infrastructure management.

Skill Development Nexus: Digital Learning Platform

Authors- M Vignesh, Dr.K.Kumutha

Abstract--The Skill Development Nexus is an innovative digital learning platform designed to enhance skill acquisition, knowledge management, and personalized learning experiences through advanced technology. This system streamlines the process of course discovery, enrollment, and progress tracking, offering structured learning paths based on user career goals. It comprises three core modules: Administrator, Learner, and a Recommendation System. The Administrator module ensures efficient course catalog management and learner activity monitoring, while the Learner module enables registration, course exploration, enrollment, and learning path creation. The Recommendation System enhances engagement by categorizing courses based on relevance and popularity. Utilizing cloud infrastructure, data analytics, and automated learning workflows, this platform addresses limitations of traditional learning management systems, reducing administrative overhead and improving accessibility. Designed for educational institutions, corporate training centers, and independent learners, this system offers a smarter, more adaptive alternative to conventional learning platforms.

/10.61463/ijset.vol.13.issue2.380

Defense Strategies for Epidemic Cyber Security Threads: Modelling and Analysis by Using a Machine Learning Approach

Authors- N Perumal, Nagasundaram S Professor

Abstract--To assure cyber security of an enterprise, typically SIEM(Security Information and Event Management) system is in place to normalize security events from different preventive technology and flag alert. Analysis in the security operation centre(SOC) investigation the alerts to decide if it is truly malicious or not. However, generally the number of alerts is overwhelming with majority of them being false positive and exceeding the SOC’s capacity to handle all alerts. Because of this, potential and exceeding the SOC’s capacity to handle all alerts. Because of this potential malicious attacks and compromised hosts may be missed. Machine learning is a viable approach to reduce the false positive rate and improve the productivity of SOC analysis. In this paper, we develop a user – centric machine learning framework for the cyber security operation centre in real enterprise environment. we discuss the typical data sources in SOC, their work flow and how to leverage and process these data sets to built and effective machine these data sets to built and effective machine learning and process these data sets to built an effective machine learning system. The paper is targeted toward two groups of readers. The first group is data scientists or machine learning systems for security operation centre.

DOI: /10.61463/ijset.vol.13.issue2.381

Data Visualization for Agriculture in Tableau

Authors- Boya Ravi Teja, Muktha Vamshi, Lekkala Abhilash Reddy, Professor Dr Diana Moses

Abstract--To assure cyber security of an enterprise, typically SIEM(Security Information and Event Management) system is in place to normalize security events from different preventive technology and flag alert. Analysis in the security operation centre(SOC) investigation the alerts to decide if it is truly malicious or not. However, generally the number of alerts is overwhelming with majority of them being false positive and exceeding the SOC’s capacity to handle all alerts. Because of this, potential and exceeding the SOC’s capacity to handle all alerts. Because of this potential malicious attacks and compromised hosts may be missed. Machine learning is a viable approach to reduce the false positive rate and improve the productivity of SOC analysis. In this paper, we develop a user – centric machine learning framework for the cyber security operation centre in real enterprise environment. we discuss the typical data sources in SOC, their work flow and how to leverage and process these data sets to built and effective machine these data sets to built and effective machine learning and process these data sets to built an effective machine learning system. The paper is targeted toward two groups of readers. The first group is data scientists or machine learning systems for security operation centre.

DOI: 

AI Content Generation in Digital World

Authors- P Harsha Veni, Dr. K Nandini

Abstract--With the overwhelming prevalence of digital communication, the demand for high-quality, precise, and content crafted at great speed is urgent across all industries, including education, marketing, entertainment, and journalism. Manual content creation has so far withstood the benefits of automation, which is why our AI Content Generator project seeks of revolution and done through sophisticated Natural Language Processing (NLP) and Machine Learning (ML) algorithms. Through AI, the system provides prompts or topics as starting points and creates relevant and coherent text, ranging from emails, blogs, marketing content and social media posts, to product descriptions and academic essays. In addition, the application provides basic functions of embedding SEO keywords, real-time grammar correction, plagiarism checking, tone adjustment, and translation. The time, cost, and effort required by the users for content generation is drastically reduced, leading to a paradigm shift in technology for innovative content solutions for the enterprises and individuals. Beyond just transforming writing into an effortless task, our aim is to incorporate the highest levels of optimization, making writing sophisticated content possible at the touch of a button.

DOI: 

Sync Vortex – B2b Project Management Platform

Authors- Jishnuraj K, Dr. V. Sumalatha Professor

Abstract--Sync Vortax is a comprehensive B2B project management solution designed to enhance team collaboration, streamline workflows, and optimize project execution. The platform integrates role- based access control, real-time analytics, and task management to improve team efficiency. Sync Vortax enables organizations to create multiple workspaces, track project progress, and analyze team performance through an intuitive interface. By leveraging modern web technologies, it ensures seamless project coordination, robust security, and scalable performance.

DOI: /10.61463/ijset.vol.13.issue2.396

SMS Spam Detection Using Machine Learning

Authors- B.Venkata Siva, B Purushotham, G.Yaswanth, G.Loknath, D.Kalyan

Abstract-– In the modern digital landscape, SMS spam has become a significant issue, leading to wasted time, security concerns, and potential financial fraud. Traditional spam filtering methods, such as keyword-based filters and rule-based detection, have proven inadequate due to their inability to adapt to new spam patterns. This project aims to build a robust SMS spam detection system using Machine Learning techniques, focusing on Naïve Bayes, Logistic Regression, and Support Vector Machine (SVM) classifiers. The system undergoes extensive training on labelled SMS datasets and is evaluated using metrics such as accuracy, precision, recall, and F1score. By utilizing feature extraction techniques like Term Frequency-Inverse Document Frequency (TF-IDF) and Count Vectorizer, the model efficiently differentiates between spam and ham messages. The ultimate goal is to integrate this system into messaging applications for real-time spam filtering, significantly reducing the risk of fraud and inconvenience for users. Additionally, this project highlights the importance of adaptive learning, allowing the model to improve over time by incorporating new spam trends. Future enhancements include integrating deep learning models like LSTM or BERT for higher accuracy and extending detection to multilingual spam messages. This system presents an automated, scalable, and intelligent solution to combat SMS spam effectively.

Hospital Information Management System

Authors- D. Vidya Sree, Shaik Haseena, B. Lakshmi, A. Swetcha, K. Amulya

Abstract-– – India Meteorological Department has implemented state level medium range rainfall forecast system applying multi model ensemble technique, making use of model outputs of state-of-the-art global models from the five leading global NWP centers. The pre-assigned grid point weights on the basis of anomaly correlation coefficients (CC) between the observed values and forecast values are determined for each constituent model utilizing two season datasets and the multi model ensemble forecasts are generated at the same resolution on a real-time basis. The ensemble forecast fields are then used to prepare forecasts for each state, taking the average value of all grid points falling in a particular district. In this paper, we describe the development strategy of the technique and performance skill of the system during 15 years of rain fall at different states in india. The study demonstrates the potential of the system for predicting future rainfall forecasts for upcoming years and scale over Indian region. District wise performance of the ensemble rainfall forecast reveals that the technique, in general, is capable of providing reasonably good forecast skill over most states of the country, particularly over the states where the monsoon systems are more dominant.

Real Time Stock Price Prediction and Market Analysis Using Machine Learning

Authors- B.Sirinath, Samson paul, K.Ramesh, K.Anjaneya Prasad

Abstract-– The utilization of Long Short-Term Memory (LSTM) algorithms for real-time stock price prediction and market analysis represents a cutting-edge approach in financial technology. LSTMs, a specialized form of Recurrent Neural Networks (RNNs), are particularly adept at capturing temporal dependencies in time-series data, making them ideally suited for the volatile and sequential nature of stock market data. This abstract delves into the development and implementation of an LSTM-based model that analyzes historical stock prices, trading volumes, and potentially other financial indicators to forecast future market trends. By effectively learning from the long-term dependencies and avoiding the vanishing gradient problem common in traditional RNNs, LSTM models can provide more accurate and reliable predictions. This capability is invaluable for investors, traders, and financial analysts, offering enhanced insights for decision-making processes. The real-time aspect of the analysis ensures that the model continuously updates and refines its predictions based on the latest market data, thereby maintaining relevance and accuracy in a rapidly changing financial landscape. This research not only underscores the potential of LSTM in financial forecasting but also paves the way for more sophisticated and adaptive market analysis tools in the future. In the rapidly evolving landscape of financial markets, the ability to accurately predict stock prices holds immense value for investors, traders, and policymakers. This study introduces a novel framework that leverages advanced machine learning techniques for real-time stock price prediction and comprehensive market analysis. By integrating diverse datasets, including historical stock prices, market sentiment from news articles and social media, economic indicators, and company fundamentals, our model provides a holistic view of the market dynamics.

Food Demand Forecasting Using Machine Learning and Statistical Analysis

Authors- M. Sirisha, CH.Sri Lakshmi Prassana, N.Ashwini, M. Prathyusha, S. Monisha

Abstract-– Accurate food demand forecasting is essential for minimizing waste, optimizing inventory, and improving supply chain efficiency in the food industry. Traditional forecasting methods often rely solely on historical data and fail to capture complex patterns influenced by dynamic factors such as holidays, weather, promotions, and consumer behavior. This project proposes an intelligent system that combines machine learning algorithms with statistical analysis to predict future food demand more effectively. The system utilizes historical sales data, time-series analysis, and external variables to train models such as Linear Regression, Random Forest, and Long Short-Term Memory (LSTM) networks. These models are evaluated based on their predictive accuracy and ability to generalize across different scenarios. Statistical methods are also applied to identify trends, seasonality, and anomalies in the data. By integrating both machine learning and statistical techniques, the proposed system offers a robust and adaptive solution for demand forecasting. It helps businesses make data-driven decisions, reduce overproduction and underproduction, and improve customer satisfaction through timely and adequate food availability.

ChatBoT Application Using Machine Learning

Authors- G. Prathiba, Bushra Thasin , C. Hindu, B. Kavyanjali, K. Nohita

Abstract-– The advancement of artificial intelligence and machine learning has led to significant innovations in human-computer interaction, with chatbots emerging as one of the most impactful applications. This project focuses on developing a chatbot application using machine learning techniques to simulate intelligent human-like conversations. The system is designed to understand user queries, process natural language input, and provide relevant and context-aware responses. It incorporates Natural Language Processing (NLP) to interpret the user’s intent and machine learning algorithms to improve its responses over time through continuous learning from interactions. The chatbot can be customized for various domains, including education, healthcare, customer service, and e-commerce. By automating repetitive tasks and providing instant support, the chatbot enhances user experience, reduces human workload, and demonstrates the practical utility of machine learning in real-world applications. The system is trained on diverse datasets to handle different types of queries and is evaluated for accuracy, efficiency, and user satisfaction.

Nutri Flow: An Integrated Approach To Pregnancy Nutrition, Diabetes Diet, and Water Tracking

Authors- K.S. Rakeshvar, Kumutha

Abstract-– Proper nutrition and hydration are the very foundations of a healthy lifestyle, especially concerning pregnant individuals and those suffering from diabetes. This article introduces Nutri Flow, a fully-fledged mobile application that combines pregnancy diet planning with diabetes-friendly meal monitoring and intelligent water intake monitoring. Bringing together digital health technologies, artificial intelligence-based recommendations, and real-time tracking, Nutri Flow is positioned to offer a hassle-free, customized approach to healthy living. The article recounts the need for an integrated system and elaborates on the impact this might have on user health outcomes..

DOI: /10.61463/ijset.vol.13.issue2.385

Solid-State Batteries: Revolutionizing the Future of Energy Storage and Electric Vehicles

Authors- Nitesh Vasave

Abstract--The worldwide effort to obtain sustainable energy recipes and the fast development of electric cars has created a stronger need for protective energy storage systems with higher efficiency and larger storage capacity. Lithium-ion (Li-ion) batteries have maintained market dominance since three decades ago but scientists are now developing next-generation alternatives because of their internal pitfalls involving thermal runaway threats and flammable liquid electrolytes together with their restricted energy storage properties and aging shortcomings. Solid- state batteries (SSBs) represent a transformative power solution because they deliver raised energy density alongside boosted safety performance and extended battery life along with adaptable forms and small sizes. The document assesses the technological advancement of industry giants and startup firms Toyota as well as Quantum Scape alongside Solid Power through evaluations of their specific advancements together with their strategic blueprints. The paper conducts assessments regarding SSB performance metrics versus conventional Li-ion storage metrics by evaluating their energy density (Wh/kg), charging capabilities, thermal behavior and cycle stability. SSBs are analyzed regarding their integration with both advanced EV structures along with battery management systems (BMS) and their regulatory compliance and environmental repercussions.

Metal Detector Using Arduino Technology

Authors- Sanskruti Vadiya, Shruti Pateriya, Saksham Namdeo, Samarth Saxena, Prof. Lakshmi Narayan Gahalod

Abstract--This research presents the design and development of a cost-effective metal detector using an Arduino Uno microcontroller. Metal detectors are essential tools in various fields such as security screening, archaeological surveys, industrial inspections, and hobbyist activities like treasure hunting. However, commercial detectors often present a barrier due to their high cost and complexity. This project addresses these challenges by creating a simple, affordable, and efficient metal detector that can be easily built by students, hobbyists, and professionals alike. The proposed system employs a copper-wire inductive search coil to generate an alternating magnetic field. When a metallic object enters this field, it disrupts the magnetic flux, causing variations in the signal, which are captured through the Arduino’s analog-to-digital converter (ADC). An embedded algorithm processes these changes and activates an alert system—comprising a buzzer and LED indicators—to notify users of metal detection events. The design prioritizes simplicity, using minimal components, while ensuring reliable detection performance. Experimental validation demonstrated the system’s capability to accurately detect various metallic objects such as coins, keys, and aluminum samples, while effectively ignoring non-metallic materials. The metal detector achieved a reasonable detection range of up to 10–15 cm, depending on the size and conductivity of the target object. Careful calibration of the detection threshold minimized false positives and improved stability against environmental noise. The project showcases the potential of Arduino- based embedded systems for practical, low-cost applications, highlighting key learning aspects in electromagnetism, sensor interfacing, and signal processing. Future enhancements could include metal type differentiation, wireless data transmission using IoT technologies, and improved sensitivity through advanced coil designs. Overall, the work demonstrates a scalable, customizable, and educational solution for low-budget metal detection needs, opening avenues for further research and development in portable detection systems.

A Hybrid Approach Adopted for Credit Card Fraud Detection Based on Deep Neural Networks and Attention Mechanism

Authors- S.Padmesh, Dr.S.Nagasundaram

Abstract--This study explores the implementation of machine learning techniques to address the significant challenge of detecting credit card fraud, focusing on data preprocessing, sampling strategies, and the evaluation of different classifiers. The dataset used comprises 284,807 transactions, where only 0.172% are fraudulent, presenting a severe class imbalance. Initially, we analyze and preprocess the data by scaling features and splitting the dataset into training and testing sets to ensure effective model evaluation. To address the class imbalance, we apply both undersampling, through Random Undersampling, and oversampling, via Synthetic Minority Over-sampling Technique (SMOTE), while also exploring anomaly detection methods such as Isolation Forests. Dimensionality reduction using t-SNE helps visualize the clustering of transactions, aiding in feature analysis. Several classifiers, including Logistic Regression, Random Forest, and Neural Networks, are employed to detect fraud, with particular emphasis on evaluating performance across different sampling techniques. A deeper analysis of Logistic Regression is conducted to understand its interpretability and performance, while SMOTE is leveraged to balance the dataset. The classifiers are tested, and their performance is evaluated using precision, recall, and F1-score, given the highly imbalanced nature of the data. Logistic Regression provides a strong baseline, but the Random Forest classifier demonstrates superior performance on the SMOTE-balanced dataset. Neural Networks, while effective, require greater computational resources and time. The study concludes that sampling methods significantly improve detection accuracy and that SMOTE coupled with Random Forest yields the best results. Finally, a comparative analysis between undersampling and oversampling techniques is conducted, particularly in testing Neural Networks, to assess their impact on model performance, offering insights into optimizing fraud detection systems for real-world applications.

Challenges and Possibilities of Implementing Iot-5G System: Methodological Approach and Limitations

Authors- Priya, Dr. Mukesh Singla

Abstract--Combining Internet of Things (IoT) devices with robust 5G networks creates enormous opportunities as our world grows more interconnected, ranging from faster, more responsive technology to smarter cities. But it’s not easy to make this happen. The possibilities and difficulties of combining IoT and 5G are examined in this paper. We examine the technical challenge such as ensuring that various devices are compatible, maintaining system security, handling massive volumes of data, and ensuring that everything functions effectively. We also examine useful tools and concepts that can improve the system’s intelligence and dependability. This study intends to assist engineers and researchers in creating better, more integrated systems by dissecting existing methods and highlighting areas that require further effort.

Evasion Attacks and Defense Mechanisms for Machine Learning-Based Web Phishing Classifiers

Authors- Sanjay Varthan T B S, Assistant Professor Dr.Lipsa Nayak

Abstract--Abstract Phishing attacks are a major threat to cybersecurity because they occur when malicious actors impersonate reliable companies in an effort to fool individuals into divulging personal information. This paper presents a phishing detection system designed to identify and prevent phishing attempts in email correspondence. The system uses Python and Django to create the Support Vector Machine (SVM) algorithm, which identifies emails as either legitimate or phishing. The three key components of the system are data preparation, model training, and feature extraction. The findings show that the SVM-based phishing detection system, along with Python and Django, offers a dependable and effective way to spot phishing attempts in emails.

DOI: /10.61463/ijset.vol.13.issue2.386

Impact of Home Delivery Services on Customer Purchase Decision and Satisfaction

Authors- Dr. S. Mohana Priya, A.Nandhini

Abstract--Abstract The rise of digital commerce and shifting consumer preferences have fueled the demand for efficient home delivery services. This study investigates the influence of home delivery services on customer purchase decisions and overall satisfaction. It examines critical factors such as delivery speed, cost-effectiveness, service quality, convenience, and order accuracy, which play a crucial role in shaping consumer behavior. Findings indicate that timely and reliable deliveries significantly enhance customer trust, satisfaction, and brand loyalty, while delays, high shipping costs, and poor service negatively impact purchasing decisions. Additionally, the study explores the role of emerging technologies such as real-time tracking, automated logistics, and contactless delivery in improving customer experiences. The insights from this research provide valuable recommendations for businesses seeking to optimize their delivery strategies and strengthen customer relationships in an increasingly competitive marketplace.

Anti-Malware Software

Authors- Velmanian B, Assistant Professor Dr.K.Nandhini

Abstract--In an era dominated by digital connectivity, safeguarding data and systems against evolving cyber threats has become paramount. This project introduces a Python Full Stack Antivirus Software designed to provide efficient, lightweight, and user-friendly protection against malware. Leveraging Python for core logic and scanning, Flet for the cross-platform frontend interface, and SQLite3 for secure and structured data storage, the software delivers real-time protection, comprehensive scanning capabilities, and intuitive user experience. The antivirus solution includes essential functionalities such as malware detection through signature and heuristic analysis, threat quarantine and removal, and accessible scan history management. Emphasis is placed on minimal system resource usage, ensuring smooth performance even on low-end devices. The architecture integrates a responsive UI with backend scanning logic, ensuring seamless operation and user interaction. Extensive unit, integration, system, performance, and security testing validate the robustness of the system, while future enhancements such as cloud-based threat intelligence, AI-based detection, and multi-platform support are proposed to expand its effectiveness and scope. This antivirus software exemplifies a scalable and secure solution tailored for both individual and small enterprise use, bridging the gap between powerful cybersecurity and simplicity.

DOI: /10.61463/ijset.vol.13.issue2.387

The Role of Human Resources in Library: The Context of Management and Function

Authors- Manasmita Maharana

Abstract--The library is regarded as an integral part of an institute. A quality education is impossible without a quality library. . The objective of the library fulfilled by the librarian and good staffs with the good administration of the library, conversion of the library into an intellectual work shop, selection of the books, organization of the books, daily needs of the students and academic community. To improve the quality and infrastructure of academic libraries in India national organizations like Govt. of India. MHRD, UGC, NAAC, NKC and various educational commissions provide important guidelines for the academic libraries in India to measure the quality of higher education. We cannot expect quality education without a good academic library .According to The Kothari commission (1964-66), human resources play an important role in efficient and management of the library.In libraries Human resource Management (HRM) plays a crucial role in ensuring the effective functioning of library services by recruiting and retaining skilled professionals. Every library should have a set of personnel policies for decision-making. The selection process includes: application forms, applicant testing, and personal interview, verification of past performance and background and finally selection.

Social Media Analysis in Criminal Investigation

Authors- Anish Chauhan, Aman Kumar, Anushka Thakur, Dipa Rao , Mr.Manish Kumar Goyal

Abstract--The emergence of social media platforms has revolutionized communication, creating vast repositories of data that can be leveraged for criminal investigations. This paper delves into the integration of social media analysis in law enforcement practices, exploring its methodologies, advantages, and ethical considerations. Social media analysis enables investigators to extract valuable evidence, identify behavioural patterns, map relationships, and predict potential threats. By employing advanced techniques like natural language processing (NLP), machine learning algorithms, and network analysis, law enforcement can process and interpret vast quantities of data with precision. Several case studies, including the Boston Marathon bombing investigation, underscore the efficacy of these tools in locating suspects and solving high-profile crimes. However, the deployment of social media analysis poses challenges, such as privacy concerns, data accuracy, and regulatory limitations. Ethical practices, transparency, and compliance with legal standards are essential to mitigate these risks. Future advancements in artificial intelligence and predictive analytics promise more efficient, real-time capabilities, enhancing crime prevention and resolution. Social media analysis represents a transformative approach to modern criminal investigations, offering innovative solutions while emphasizing ethical responsibilities. Its potential continues to grow, demanding collaborative efforts among technologists, policymakers, and law enforcement agencies to strike a balance between security and individual rights.

DOI: /10.61463/ijset.vol.13.issue2.388

Block chain Based Access Control And Data Sharing Model in Decentralized Storage System

Authors- Shardul T Rushesary, d Kunjan V Sharma, Rutvik M Sonare, Prapti P Ugale, Mr V. D. Khandar

Abstract-– The integration of autonomous drones with Artificial Intelligence (AI) is revolutionizing surveillance and security, enhancing monitoring, threat detection, and rapid response capabilities. This study examines the role of AI-driven drones in modern security systems, highlighting their potential to improve situational awareness, reduce human intervention, and optimize operational efficiency. With machine learning algorithms, computer vision, and real-time data analytics, autonomous drones can autonomously detect anomalies, track suspicious activities, and respond to security threats with precision. These advancements are particularly valuable for border security, law enforcement, critical infrastructure monitoring, and disaster response.Despite their benefits, AI-powered drones face challenges such as ethical concerns, privacy issues, regulatory constraints, and cybersecurity risks. This research explores the legal and ethical implications of autonomous surveillance, reviewing current policies and governance to ensure responsible use. It also addresses technical limitations, including power constraints, environmental adaptability, and AI biases in threat assessment, while suggesting solutions to improve reliability and security.Through case studies and analysis of emerging trends, this study provides an evaluation of the evolving role of autonomous drones in security operations. The findings contribute to discussions on responsible AI use, regulatory policies, and future innovations in autonomous surveillance. Ultimately, this research emphasizes the need for a balanced approach that maximizes the benefits of AI-driven drones while addressing their ethical, legal, and technical challenges.

Navigating The Maze Exploring Blockchain Privacy And its Information retrieval

Authors- Vinoth V, Poongodi A

Abstract-– This research work critically analyzes the complex relationship among blockchain technology, privacy concerns, and information retrieval systems. While blockchain technology’s decentralized and immutable nature has increased its adoption rate, it conversely also introduces new challenges of preserving the privacy of the user while ensuring effective data retrieval. While blockchain offers immense potential as a secure medium for data storage, the fundamental transparency and open nature of blockchain have the propensity to inadvertently exposesensitive information and hence createprivacy risks. In this study, we explore several privacy-preserving methods, including zero-knowledge proofs, encryption, and privacy-focusedconsensus mechanisms, and analyze their implications in information retrieval. We propose an architecture that ensures privacy without infringing on information retrieval processes’ integrity and effectiveness in blockchain-based systems and thus responds to the need for harmony between transparency and confidentiality the an indecentralized networks.

DOI: /10.61463/ijset.vol.13.issue2.389

Fake account identification on social network sites using machine learning

Authors- MCA Student R.Jaiganesh, Prof. V.Sumalatha

Abstract-– At present social network sites are part of the life for most of the people and want to free-frank with new friends. Every day several people are creating their profiles on the social network platforms and they are interacting with others independent of the user’s location and time, Technology is associated with online social networks which has become a part in every one’s life in making new friends and keeping friends, and share a personal information with other users their interests are known easier. Traditional methods cannot differentiate between real and fake accounts efficiently. To analyze, who are encouraging threats in social networking sites with user profiles. There are numerous cases where produced accounts have been effectively distinguished utilizing machine adapting techniques characters made by people.

DOI: /10.61463/ijset.vol.13.issue2.390

Examining the Effectiveness of Online Learning Platforms in Commerce Education

Authors- Ankita Kamlesh Mishra, Assistant Professor Dr Pritama Devi Tyagi

Abstract-– This study inquiries into the impact of online learning platforms in give commerce education. With the quick evolution of innovation, e-learning consumes converted a fundamental of learning facilities. This study examines several online stages, there educational methods, and there influence on pupil assignation, holding, and knowledge results in trade sequences. Information stayed composed done reviews, examination, and conferences of abstract presentation, see-through that although connected stages proposal gives and convenience, tests such as lack of individual communication and self-discipline amongst pupils continue. The results advise a essential for cross replicas that syndicate the strong point of traditional and online teaching systems. Through the several returns of reachable, reasonable operational teaching, groups and children can increase overhead the deficiency line close by and internationally. Online schooling can put up the time restraints and monetary means of current-day pupils. In spite of these assistances, a doubt of the efficiency in skill continues, as numerous qualms about behind the influences complete in a old-style teaching space. However, this uneasiness about online education must not depress possible pupils after chasing a excellence teaching over the means of the Internet. In directive for apprentices to connect the individualization, flexible judgement, and low-slung expenses that online instruction proposals, there necessity be extra elevation in the mass media of the exclusive chances that online knowledge delivers.

A Comparative Forecast Of The City Of Meerut’s Air Quality Index Using Artificial Neural Networks (Ann)

Authors- Research Scholar Lokesh Kumar, Prof. Gaurav Kumar

Abstract-– One of the main air pollutants that causes air pollution is PM10. This study analyzed data gathered between 2019 and 2023 to evaluate the impact of this pollutant on people’s health and the surroundings using ANN, a popular learning technique. The air monitoring center of the UPPCB (Pollution Control Board of Uttar Pradesh) collected data on the industrial center in Meerut and used SPSS programming to finish the simulation and optimization processes. The estimated air quality findings underwent a multilayer perceptron analysis prior to comparison with the real air quality data. Furthermore, there have been instances in which the Meerut province’s air quality index (AQI) readings have exceeded the allowable limit, especially during times of peak output.

DOI: /10.61463/ijset.vol.13.issue2.391

Gold Price Prediction using Machine Learning

Authors- V Tamilarasu, Assistant Professor Dr. A. Poongodi

Abstract-– We predict future gold rates supported twenty two market variables victimization machine learning technique. One machine learning algorithm, random forest regression were used in analyzing these knowledge[1]. Historically, gold was used for supporting trade transactions around the world besides alternative modes of payment. Various states maintained and increased their gold reserves and were recognized as rich and progressive states. In present times, precious metals like gold area unit control with central banks of all countries to make sure re-payment of foreign debts, and conjointly to control inflation. Moreover, it conjointly reflects the Imoney strength of the country[2]. Besides government agencies, varied transnational firms and people have conjointly invested with in gold reserves. In ancient events of Asian countries, gold is in addition presented as gifts/souvenirs and in marriages, gold ornaments are conferred as gift in Republic of India

DOI: /10.61463/ijset.vol.13.issue2.392

AI-Powered Sentiment Analysis of Social Media: Trends, Challenges, and Insights

Authors- Koniki Ganesh, Y Sai Pallavi, Dr.Swapna M, P Srinivas, B Saikumar

Abstract-– The proliferation of social media sites has generated an unprecedented amount of user-created content, rendering sentiment analysis a vital tool to gauge public opinion, brand reputation, and social trends. Methods of sentiment analysis like supervised and unsupervised learning, as well as lexicon-based models, and hybrid models have been proposed to categorize social media posts as positive, negative, or neutral. Sophisticated AI techniques like Random Forest, Decision Tree, and XGBoost quite highly boost sentiment classification accuracy. In spite of all these developments, difficulty exists in situations like sarcasm detection, multilingual text analysis, and contextual ambiguity, making sentiment analysis in real-time difficult with changing slang and noisy data. The ubiquitous use of sentiment analysis in marketing, politics, healthcare, finance, and crisis management underscores its increasing significance. Companies utilize it for learning consumer sentiments, governments make use of it to analyze policy issues, and academics use it for research purposes. The ongoing research in the area of deep learning, explainable AI, and cross-lingual analysis will further enrich techniques for sentiment analysis so that it can monitor and analyze social media even better. Real-time sentiment monitoring and more sophisticated analytical techniques will improve decision-making, customer interaction, and trend identification, enabling organizations to lead in a more digital age. With the development of sentiment analysis, it will prove to be an integral tool for companies, researchers, and policymakers alike, offering meaningful insights into public opinion and allowing data-driven approaches to achieve improved results in various fields.

DOI: /10.61463/ijset.vol.13.issue2.393

Respiratory Disease Detection Using Machine Learning For Sound Classification

Authors- M Divya, Assistant Professor Dr. C. Meenakshi

Abstract-– Respiratory sounds play a meaningful part in surveying of aspiratory wellbeing and recognizing respiratory disarranges at an early arrange. The emergence of artificial intelligence (AI) has encouraged the utilization of machine learning (ML) methods so as to analyze respiratory irregularities, inclusive of conditions like asthma, pneumonia, bronchiolitis, and also chronic obstructive pulmonary disease (COPD). Conventional auscultation strategies remain important, but subjectivity, variations in clinical translation, and irregularities in sound quality often limit their adequacy. Subsequent advances in computational techniques have upgraded demonstrative precision. These advances empower mechanized discovery of various unusual lung sounds, for example, wheezes and crackles. This investigation examines if choice tree-based classification models can diagnose respiratory illnesses, reaching an impressive accuracy of 90%. The aforementioned highlights underscore the possibility of AI-powered symptomatic tools in yielding durable results and advancing respiratory medicine.

DOI: /10.61463/ijset.vol.13.issue2.394

Artificial Intelligence Health Care Chatbot System

Authors- G.Hariharan, Professor Priya.R

Abstract-– Through chatbots one can communicate with text or voice interface and get reply through artificial intelligence. Typically, a chat bot will communicate with a real person. Chat bots are used in applications such as ecommerce customer service, call centers and Internet gaming. Chatbots are programs built to automatically engage with received voice messages. Chatbots can be programmed to respond the same way each time, to respond differently to voice messages containing certain keywords and even to use machine learning to adapt their responses to fit the situation. Many developing number of hospitals, nursing homes, and even private centers, presently utilize online Chatbots voice message for human services on their sites. These bots connect with potential patients visiting the site, helping them discover specialists, booking their appointments, and getting them access to the correct treatment. In any case, the utilization of artificial intelligence in an industry where individuals’ lives could be in question, still starts misgivings in individuals. It brings up issues about whether the task mentioned above ought to be assigned to human staff. This healthcare chatbot system will help hospitals to provide healthcare support online 24 x 7 live data. It also helps to generate leads and automatically delivers the information of leads to sales.

DOI: /10.61463/ijset.vol.13.issue2.395

Disease-Specific Fruit Recommendations Using AI and ML

Authors- Sarthak V. Bante, Khushi S. Bhaisare, Pratik S. Dudhe, Gayatri P. Bhakare, Professor Prajwal Shete

Abstract-– This paper presents a personalized fruit recommendation system leveraging Artificial Intelligence (AI) and Machine Learning (ML) to support dietary management for users with specific health conditions. Addressing the critical role of nutrition in managing chronic diseases like diabetes, hypertension, and cardiovascular disorders, this system provides tailored fruit recommendations based on the user’s symptoms, disease profiles, and seasonal fruit availability. Implemented with a Web technologies user interface, the system combines ease of use with a powerful recommendation engine. The AI algorithm assesses user data to suggest optimal fruits, promoting balanced nutrient intake and mitigating disease symptoms. Over time, the system refines its recommendations based on user interactions, enhancing relevance and accuracy. By providing actionable, individualized dietary guidance, this approach aims to improve health outcomes and assist in the proactive management of chronic conditions through informed nutrition choices. The study contributes a scalable, interactive tool to health informatics, offering a novel application of AI in personalized nutrition and preventive healthcare.

A Novel Prediction Method for Glaucoma Detection Using Retinographies

Authors- A. Ramakrishna Prasad, B. Gayatri, A. Lakshmi Sree, G. Jaswanth Kumar

Abstract-– The primary objective of the proposed Glaucoma Detection system is to facilitate the early identification of glaucoma through the analysis of retinal fundus images using advanced medical imaging techniques. Glaucoma is a progressive and irreversible eye disease that, if not diagnosed and treated promptly, can lead to permanent vision loss. Early detection and timely intervention are critical in mitigating the risk of severe visual impairment and improving the quality of life for affected individuals. The proposed system is designed to assist in pre-diagnosing glaucoma by employing a combination of image processing and machine learning approaches. In particular, the system leverages the Support Vector Machine (SVM) algorithm to classify retinal images, using the publicly available ORIGA dataset as a benchmark for evaluation. A comprehensive set of pre-processing steps including image segmentation, enhancement, binarization, and thresholding is applied to isolate key anatomical structures, specifically the optic cup and optic disc, from the retinal images. Accurate segmentation of the optic disc and optic cup is crucial for calculating the Cup-to-Disc Ratio (CDR), a widely used clinical indicator in glaucoma assessment. However, recognizing that the CDR alone may not provide sufficient discriminatory power between glaucomatous and non-glaucomatous conditions, the system also incorporates the Rim-to-Disc Ratio (RDR) as an additional diagnostic feature. This dual-feature approach enhances the robustness of the detection process.

Internet of Things: Smart Voting System Using Through Fingerprint

Authors- Sanket Gawande, Jagruti Jeughale, Pooja Gawande, Sakshi Deshmukh

Abstract-– The main thing our design is to introduce a new electoral system that uses fingerprints to check choosers. Our proposed system can be used to run choices at colorful situations, similar as Congress, Panchayats, etc. on the same day. This reduces the cost of running choices with a variety of data. This is progressing compared to current system because it does not bear mortal performance. And choosers do not need the visit the polling station. The operation will follow applicable authentication measures to avoid fraud by using the systems. This technology helps druggie’s handover votes without visiting the polling station.

DOI: /10.61463/ijset.vol.13.issue2.397

Dynamic Cryptographic Key Generation Using Real-Time Environmental Data

Authors- Neel Kesariya, Princy Shah, Mohit Zinzuwadia, Jiya Doshi, Darsh Badhiya, Asst. Prof. Ms. Twinkle Patel

Abstract-– research examines a new method for dynamic cryptographic key creation that uses real-world environmental data that changes quickly to generate keys. Every few seconds, a random city from around the world is chosen, parameters like temperature, humidity, wind speed, and seismic readings are retrieved from public APIs (like OpenWeatherMap and USGS), combined (for example, by concatenation or summation), and hashed (for example, using SHA-256) to generate a new encryption key. This strategy attempts to make keys context-dependent and more difficult to guess by altering the key on a regular basis. We review relevant work on environmental and context-based keys and examine the system design, methodical approach, and entropy considerations . Implementation concerns (API constraints, timing), practical data sources (weather, seismic, and geographical) and their APIs, and benefits and drawbacks are all covered. The article describes future developments (new sensors, more robust hashes) and possible real-world uses (IoT devices, mobile agents, secure triggers). In order to provide a thorough and rigorous analysis of this new idea, we reference pertinent literature and standards at every stage.

Online Medicine Donation System for NGO

Authors- V. Gopala Krishnan, Professor R. Priya

Abstract-– Internet technology subsequently with its wide network can offer potential solutions to facilitate growing demand for better health centers and services. This proposed Medicine Donation System will have an online platform where individuals donate unused medicines to NGOs. Individuals using this website can register and enter details of the medicines to be donated, and NGOs can manage their stock of medicines with ease. The system has integrated screening to detect and eliminate expired and inappropriate medicines such that only safe donations will be accepted. The programmer will thus help avoid drug wastage and improve access among poor ill individuals who require medication.

DOI: /10.61463/ijset.vol.13.issue2.396

Using Existing CCTV Network for Crowd Management and Crime Prevention Using AIML

Authors- Deborah Dorothy S, Professor R.Priya

Abstract-– With the increasing prevalence of Closed Circuit Television (CCTV) networks in urban environments and workplaces, there is a significant opportunity to leverage existing infrastructure for enhanced crowd management, crime prevention, and work monitoring. This paper explores the multifaceted applications of CCTV systems, focusing on their potential to contribute to public safety and organizational efficiency. The integration of advanced analytics, machine learning algorithms, and real-time monitoring capabilities enhances the capabilities of traditional CCTV setups. The first aspect addressed is crowd management, where the paper discusses the utilization of CCTV networks to monitor and analyze crowd behavior in public.

DOI: /10.61463/ijset.vol.13.issue2.398

Utilizing CNN Architectures for Herbal Leaf Health and Pathogen Detection in Deep Learning

Authors- Sripriya M S, Assistant Professor Dr. Nandhini. K

Abstract-– This research delves into the transformative application of Convolution Neural Networks (CNNs) in the realm of herbal plant identification and categorization. CNNs, a subset of artificial neural networks, have emerged as a powerful tool for processing visual data, particularly in the context of image-based tasks such as recognizing diseases in herbal plants. The architecture of CNNs, characterized by convolution layers for hierarchical feature extraction, proves instrumental in capturing intricate patterns within images. Through meticulous data preprocessing and augmentation techniques, including normalization and data augmentation, the model is optimized for robust performance. Training CNNs involves exposing the model to labeled datasets, with transfer learning from pre-trained models facilitating accelerated knowledge transfer. The application of CNNs in herbal plant identification showcases notable advancements in accuracy and speed compared to traditional methods. This research envisions the continued refinement of CNN-based models, addressing challenges such as limited annotated datasets and interpretability concerns. As we look forward, the integration of CNNs into agricultural practices holds the promise of revolutionizing disease detection in herbal plants, contributing significantly to global food security by enabling early and accurate identification of plants.

DOI: /10.61463/ijset.vol.13.issue2.399

Skill Development Nexus: Digital Learning Platform

Authors- M Vignesh, Assistant Professor Dr. H. Jayamangala

Abstract-– The Skill Development Nexus is an innovative digital learning platform designed to enhance skill acquisition, knowledge management, and personalized learning experiences through advanced technology. This system streamlines the process of course discovery, enrollment, and progress tracking, offering structured learning paths based on user career goals. It comprises three core modules: Administrator, Learner, and a Recommendation System. The Administrator module ensures efficient course catalog management and learner activity monitoring, while the Learner module enables registration, course exploration, enrollment, and learning path creation. The Recommendation System enhances engagement by categorizing courses based on relevance and popularity. Utilizing cloud infrastructure, data analytics, and automated learning workflows, this platform addresses limitations of traditional learning management systems, reducing administrative overhead and improving accessibility. Designed for educational institutions, corporate training centers, and independent learners, this system offers a smarter, more adaptive alternative to conventional learning platforms.

DOI: /10.61463/ijset.vol.13.issue2.400

A Novel High-Performance Face Anti-Spoofing Detection Method

Authors- Srinath S, Assistant Professor Dr. Krithika. D. R

Abstract-– Existing face anti-spoofing models using deep learning for multi-modality data suffer from low generalization when using a variety of presentation attacks such as 2D printing and high-precision 3D face masks. One of the main reasons is that the non-linearity of multi-spectral information used to preserve the intrinsic attributes between a real and a fake face is not well extracted. To address this issue, we propose a multi-ability data-based two-stage cascade framework for face anti-spoofing. The proposed framework has two advantages. Firstly, we design a two-stage cascade architecture to selectively fuse low-level and high-level features from different modalities to improve feature representation. Secondly, we use multi-modality data to construct a distance-free spectral on RGB and infrared (IR) to augment the non-linearity of data. The presented data fusion strategy is different from popular fusion approaches, since it can strengthen the discrimination ability of network models on physical attribute features rather than identity structure features under certain constraints. In addition, a multi-scale patch-based weighted fine-tuning strategy is designed to learn each specific local face region. Experimental results show that. The proposed framework achieves better performance than other state-of-the-art methods on both benchmark datasets and self-established datasets, especially on multi-material mask spoofing.

DOI: /10.61463/ijset.vol.13.issue2.401

Examining the Effectiveness of Online Learning Platforms in Commerce Education

Authors- Ankita Kamlesh Mishra, Assistant Professor Dr Pritama Devi Tyagi

Abstract-– This study inquiries into the impact of online learning platforms in give commerce education. With the quick evolution of innovation, e-learning consumes converted a fundamental of learning facilities. This study examines several online stages, there educational methods, and there influence on pupil assignation, holding, and knowledge results in trade sequences. Information stayed composed done reviews, examination, and conferences of abstract presentation, see- through that although connected stages proposal gives and convenience, tests such as lack of individual communication and self-discipline amongst pupils continue. The results advise a essential for cross replicas that syndicate the strong point of traditional and online teaching systems. Through the several returns of reachable, reasonable operational teaching, groups and children can increase overhead the deficiency line close by and internationally. Online schooling can put up the time restraints and monetary means of current-day pupils. In spite of these assistances, a doubt of the efficiency in skill continues, as numerous qualms about behind the influences complete in a old-style teaching space. However, this uneasiness about online education must not depress possible pupils after chasing a excellence teaching over the means of the Internet. In directive for apprentices to connect the individualization, flexible judgement, and low-slung expenses that online instruction proposals, there necessity be extra elevation in the mass media of the exclusive chances that online knowledge delivers.

Canteen Food Ordering System and Management

Authors- G D Bala Hariharan, Assistant Professor Dr. A Poongodi

Abstract-– The Canteen Automation System allows the user to register for online ordering, enabling them to view the e-menu and order food from the web application. When an order is placed, it directly reflects on the screen in the kitchen, thus easing the process of cooking. The system saves considerable staff effort while highly improving efficiency. The major issue in universities canteens is huge crowd density during the peak hours of lunch time. In order to solve this problem, the Smart Canteen, which includes an online food ordering system and a virtual queue management system, was implemented for optimized operation and enhanced user experience.

DOI: /10.61463/ijset.vol.13.issue2.402

Embedded System for Food Contamination Detection Using Gas and Color Sensors

Authors- T. Cowsalya, Jeni Narayanan L A

Abstract-– A compact embedded system is developed to detect food contamination using gas and color sensors. MQ-series sensors identify spoilage-related gases, while TCS34725 sensors monitor visual changes in food items. A microcontroller processes the collected data and uses lightweight machine learning algorithms to assess contamination levels. Alerts are triggered when unsafe conditions are detected, and real-time data is displayed or transmitted to a mobile interface. Designed for storage, transportation, and retail environments, the system provides a low-cost, non-invasive solution to improve food safety, minimize spoilage, and support effective quality control throughout the supply chain.

Skin Cancer Detection Using Machine Learning

Authors- Helan Sharmila M, Assistance Professor Dr.V.Sumalatha

Abstract-– Skin cancer is one of the most common types of cancer globally, and early detection plays a crucial role in improving patient outcomes. Traditional methods for diagnosing skin cancer involve visual examination by dermatologists, followed by biopsies for confirmation, which can be time-consuming and subject to human error. This study proposes a machine learning-based approach for the automatic detection of skin cancer from dermoscopic images. Using various machine learning algorithms, the model is trained on a large dataset of skin lesion images to distinguish between benign and malignant tumors. The proposed system utilizes feature extraction and classification techniques to detect melanoma and other skin cancers with high accuracy. This automated solution aims to assist dermatologists by providing a reliable and efficient tool for early skin cancer detection, reducing diagnostic errors and enabling timely intervention.

DOI: /10.61463/ijset.vol.13.issue2.403

E-Farming: Revolutionizing Agriculture Through Digital Innovation

Authors- Annamalai. A, Dr. S. Prasanna

Abstract-– The fast pace of digital transformation influences the changes taking place in farming traditions. Hence, agriculture becomes more efficient, sustainable, and above all, profit-oriented. It is an upcoming paradigm shift in agriculture-is the utilization of digital platforms, Internet of things enabled smart farming, Artificial Intelligence and blockchain technologies in maximizing productivity, managing resources better, and minimizing adverse environmental impact. Therefore, this paper we will look at digital innovation in e-farming through precision farming, automated monitoring systems, and data-driven decision-making to revolutionize agriculture. It will also study what challenges the practice faces and its prospects for e-farming, highlighting the needs arising from policy decisions, infrastructure in technology, and adaptability of farmers. Through the unification of conventional and modern agricultural practices, e-farming offers a means toward food security, yield efficiency, sustainable farming internationally.

Breast Cancer Detection Using Efficientv2b0

Authors- Assistant Professor Ms.S.Jayachitra, Sneha K, Srinithi G, Sujitha M, Tharani S

Abstract-– Breast cancer remains one of the leading causes of mortality among women worldwide, emphasizing the critical need for early and accurate detection methods. Recent advancements in deep learning have shown significant promise in enhancing diagnostic accuracy through automated image analysis. In this work, we propose an approach for breast cancer detection using the EfficientNetV2B0 architecture, a state-of-the-art convolutional neural network known for its efficiency and superior performance. Leveraging transfer learning techniques, the pre-trained EfficientNetV2B0 model is fine-tuned on a curated dataset of histopathological images to classify benign and malignant cases. Comprehensive preprocessing methods, including data augmentation and normalization, are applied to improve model robustness. Experimental results demonstrate that our approach achieves high accuracy, precision, and F1-score, highlighting the model’s potential to assist healthcare professionals in reliable and early breast cancer diagnosis. This work discusses the effectiveness of EfficientNetV2B0 in medical image classification tasks and its applicability in real-world clinical environments.

GPS-Enable Helmet for Accident Alert

Authors- Professor Prasad Bhosle, Shreejit Karpe, Sayali Kasab, Ashwini Janrao, Anushka Katkade

Abstract-– The integration of shrewd innovation with security gear presents an inventive approach to upgrading rider security. This venture centers on the advancement of a GPS-enabled protective cap outlined to supply real-time notice alarms for riders. By inserting GPS innovation inside the protective cap, the framework can track the rider’s area and transfer critical data, such as mishap alarms, activity overhauls, and course notices. These cautions offer assistance minimize diversions whereas permitting the rider to stay mindful of basic data. The protective cap moreover incorporates an crisis alarm framework that consequently sends notifications to pre-designated contacts within the occasion of an mischance, in this way improving emergency reaction times. This inventive arrangement points to make a more secure riding involvement by consolidating innovation with conventional defensive adapt, tending to both security and comfort for riders in a single, effective gadget.

Integrated Smart Home Management and Automation Control System

Authors- Dhanush D, Assistance Professor Jayamangala Hariharan

Abstract-– This project presents the development of an Integrated Smart Home Management Web Portal Application, a centralized digital platform designed to streamline the process of connecting interior designers with potential clients. The web portal offers a user-friendly interface where users can browse designer portfolios, view sample projects, check availability, and book consultation or design services directly. Designers can create profiles, showcase their work, manage schedules, and respond to client inquiries. The platform incorporates features such as real- time booking, secure payment integration, user reviews, project tracking, and a messaging system to facilitate seamless communication. Built using modern web technologies, the portal aims to enhance accessibility, reduce friction in the design hiring process, and improve overall customer satisfaction. This application bridges the gap between design professionals and clients, creating a more efficient and organized service experience. The platform will provide users with access to a wide range of design professionals, customizable packages, and an intuitive interface for selecting design styles, preferences, both clients and designers. By digitizing and automating key aspects of the design hiring process, the platform aims to enhance customer satisfaction, expand market reach for designers, and elevate the overall efficiency of the interior design ecosystem.

DOI: /10.61463/ijset.vol.13.issue2.404

Electronic Product Search Based Chatbot System

Authors- Amarnath K, Dr.S.Prasanna

Abstract-– E-commerce has the potential to be a huge success in today’s commercial world. E- Commerce platform is used for purchasing of the human required goods and selling of items through it as well as the payments of the purchased goods made through online moreover the different payment platforms over the internet. E-commerce may represent a paradigm change that affects both marketer sand customers. Rather, e- commerce is merely a new way to spice up existing business methods. It’s causing a complete shift in the traditional company model. A web application, which is a piece of software that runs on a website. It is a computer-based software programme that is saved locally on the system’s operating system. It is used by users or administrators over the Internet. It can be used to create an application for a variety of platforms. However, we believe that using PHP is the safest option and can be employed in a variety of situations. It also plays a significant role in our web apps. Hackers won’t be able to use SQL Injectionas rapidly if the web application is written in PHP. This project is intended to create a chatbot to be used by customers to get their queries responded easily from the e-commerce website. This chatbot has the capacity to make friendly conversations and respond to the queries. Moreover, it provides information with product details, payment method and many more. The chatbot allows users to look up electronics, compare specs, and check prices, availability, and promotions in a conversational way that resembles humans. The user experience, accuracy of responses, and accessibility is greatly enhanced, and customer service workload is reduced. Technologies used include: PHP for the frontend, MySQL for the backend, and Microsoft Azure cognitive services for sentiment analysis and QnA.

DOI: /10.61463/ijset.vol.13.issue2.405

Authors- Priyanka Kelkar, Pranjal Aanjna, Preeti Anjana, Asst. Prof. Garima S. Tukra, Asst. Prof. Kailash Kumar Baraskar

Abstract-– Antimicrobial resistance (AMR) is a critical global health challenge, resulting in over 1.27 million deaths annually due to ineffective antibiotic treatments. Traditional AMR detection methods, including culture tests and PCR-based assays, are time-consuming and resource-intensive. In this study, we propose an automated AMR detection framework using Vision Transformers (ViTs) to classify blood culture images into resistant (R) and susceptible (S) categories. The ViT architecture effectively processes high-resolution blood culture images by leveraging multi-head attention mechanisms, enabling the model to capture fine-grained morphological differences associated with resistance patterns. The proposed model was evaluated on the PathAI AMR Image Dataset, containing 20,000 labeled blood culture images, achieving an accuracy of 94.2%, surpassing state-of-the-art CNN-based models by 6.7%. The model also achieved a precision of 92.8%, a recall of 95.4%, and an F1-score of 94.1%. Furthermore, Grad-CAM-based attention heatmaps provided interpretability, highlighting bacterial regions contributing to the resistance classification. The results demonstrate that ViT-based AMR detection offers superior accuracy, faster inference, and greater interpretability compared to traditional methods. This automated framework has the potential to significantly reduce diagnostic time in clinical settings, enabling faster and more precise antibiotic administration, thereby improving patient outcomes.

DOI: /10.61463/ijset.vol.13.issue2.406

Online Product Comparison Ecommerce Website

Authors- P.Elavarasan, R.Priya

Abstract-– With the rise of e-commerce, consumers frequently compare prices across various online stores before making a purchase. However, manually visiting multiple websites for price comparison is time consuming and inefficient. This study addresses this challenge by developing an automated price comparison website that gathers product prices from different e-commerce platforms and presents them in a structured format.

A Critical Discourse Analysis of John F. Kennedy’s Inaugural Address (1961) Using Fairclough’s Model

Authors- Radwan Alkarrash

Abstract-– This research applies Norman Fairclough’s Critical Discourse Analysis (CDA) to analyze John F. Kennedy’s inaugural address delivered on January 20, 1961. The speech undergoes CDA analysis to study how language builds ideological frameworks, displays power relationships, and creates political identities. The research investigates how Kennedy employed lexical choices, rhetorical devices, pronouns, metaphors and appeals to establish American global leadership while promoting domestic unity against both domestic civil rights tensions and foreign cold war pressures. The research investigates the relationship between the speech’s linguistic elements and its historical and socio-political setting.

DOI: /10.61463/ijset.vol.13.issue2.407

Hybrid Approach for Multi Domain Aspect Extraction Using Specific Closed Sequential Pattern and Extended Random Set Technique

Authors- Sani Abdullahi

Abstract-– Our studies have been focusing on extracting better aspects that carry more weight from a textual dataset. One solution is to use data mining techniques, such as low minimum support for extracting sequential, relevant, specific and informative features .Such methods adopted the concept of extracting sequential patterns and assign high weight to top-k features based on the Extended Random Set (ERS) probability weight by pruning non-closed pattern from the representation by removing noisy features.

DOI: /10.61463/ijset.vol.13.issue2.408

Online Product Comparison Ecommerce Website

Authors- P.Elavarasan, Professor R.Priya

Abstract-– With the rise of e-commerce, consumers frequently compare prices across various online stores before making a purchase. However, manually visiting multiple websites for price comparison is time consuming and inefficient. This study addresses this challenge by developing an automated price comparison website that gathers product prices from different e-commerce platforms and presents them in a structured format.

DOI: /10.61463/ijset.vol.13.issue2.409

Fraud Detection in Bank Sector Using Machine Learning

Authors- Mohanraj G, Associate Professor DR. C. Meenakshi

Abstract-– This project explores a comprehensive machine learning approach to detect anomalous financial transactions, specifically focusing on credit card fraud. Utilizing a simulated dataset of customer transactions, the system applies key steps such as data cleaning, exploratory data analysis, feature engineering, and model building. Techniques like SMOTE are used to handle class imbalance, which is common in fraud detection scenarios. Multiple models—including Logistic Regression, Decision Tree, and Random Forest—are developed and evaluated using accuracy, precision, recall, and F1-score. Among them, Random Forest showed superior performance in identifying fraudulent patterns. The project demonstrates how combining data preprocessing, visualization, and robust machine learning algorithms can lead to an effective, scalable, and adaptable fraud detection solution, paving the way for real-time applications and future enhancements using deep learning.

DOI: /10.61463/ijset.vol.13.issue2.410

AI-Powered Document Summarizer Using Generative AI

Authors- Harish Kesavan, Assistant Professor Dr. Kavitha

Abstract-– The AI-Powered Document Summarizer is an advanced system designed to extract text from PDF and DOCX files and generate concise summaries using Google’s Generative AI model. The system is implemented in Python, utilizing Streamlit for the user interface and pdfplumber/docx for text extraction. The summarization process is optimized for large document inputs, providing quick and meaningful insights. The solution enhances productivity by condensing lengthy texts into brief, informative outputs. With an interactive and user-friendly interface, it supports research, academic reviews, and report analysis.It aims to reduce manual effort and improve the efficiency of document analysis, making it highly effective for students, researchers, and professionals.

DOI: /10.61463/ijset.vol.13.issue2.411

Lightweight Cryptography – Cryptographic Algorithm for IoT and Constrained Devices

Authors- Krishnakumar Shukla, Twinkle Patel, Chintan Patel, Ayush Patel

Abstract-– The rapid expansion of the Internet of Things (IoT) has led to a surge in smart, interconnected devices deployed across diverse fields such as healthcare, agriculture, smart homes, and industrial automation. These devices, often referred to as constrained devices, operate under strict limitations in terms of processing power, memory, energy, and bandwidth. Securing communication and data transmission in such environments presents a significant challenge, as traditional cryptographic methods are too resource-intensive for these devices. This is where lightweight cryptography comes into play—offering tailored cryptographic algorithms that provide adequate security while consuming minimal computational resources. Algorithms such as PRESENT, Speck, Simon, and HEIGHT are specifically designed to meet the unique requirements of constrained IoT devices. This paper explores the foundational concepts of IoT and constrained devices, introduces the need and principles of lightweight cryptography, and examines key algorithms in this domain. It also highlights the interdependency between these technologies and emphasizes the importance of lightweight cryptographic solutions in ensuring secure and efficient IoT deployments.

Hydrogel Coating Formulations for Optimized Heat Transfer and Efficiency

Authors- Boopathi A, Dr.S. Prasanna

Abstract-– This project introduces a novel hydrogel-based coating system engineered to enhance surface protection and operational efficiency across diverse industrial applications. Leveraging the unique properties of hydrogels-such as high-water retention, thermal stability, and environmental responsiveness-the proposed coating addresses persistent industrial challenges including limescale accumulation, inefficient heat transfer, and elevated carbon emissions. By reducing maintenance frequency and chemical usage, the system offers a sustainable alternative to conventional coatings. The architecture comprises five core modules, with the Coating Estimation Module calculating optimal application parameters based on surface area and environmental conditions. Experimental data is iteratively fed into subsequent processes for refinement. Centralized control is managed by the Admin Module, which oversees system operations and securely distributes cryptographic keys. To safeguard sensitive data, Blowfish encryption is integrated for secure handling of coating parameters and test results. The use of Blowfish-an efficient symmetric-key block cipher—ensures robust data protection without compromising system performance.

DOI: /10.61463/ijset.vol.13.issue2.412

Career Resume Builder

Authors- T. Sneha Akshaya, Assistant Professor Dr. Krithika.D. R

Abstract--In the fiercely competitive job landscape of today, having an integrated and properly crafted resume is of utmost need for securing job opportunities. This study proposes a novel automated resume building system aimed at improving a user’s chances of getting employed by utilizing Artificial Intelligence (AI) and data driven approaches. The proposed system applies natural language processing (NLP) technology to interpret job postings, recommend appropriate skills, and create resumes optimized for applicant tracking systems (ATS). Moreover, machine learning models are used to give users tailored suggestions that meet the requirements of a given industry and employer’s preferences. In addition, the system contains real-time feedback capabilities and customizable templates to enhance the overall aesthetics and readability of resumes. This paper seeks to improve the connection between job applicants and employers by automating critical elements of resume formulation and editing. This change, in turn, will increase hiring rates. In this case study, the system developed in this research is tested on its ability to generate professional resumes ATS system with keyword in ease of access and upgraded formatting consistency, as well as relevance and consistency, which has proven successful.

DOI: /10.61463/ijset.vol.13.issue2.413

Ev Smart Charging with Advance Reservation

Authors- R.Mohammed Shafi, Dr.Lipsa Nayak

Abstract--The infrastructure for charging electric vehicles is a crucial component of the mobility sector. It displays the functional groups for electric vehicles, with smart chargers being the most widely used and offering numerous advantages over offline ones. Additionally, charger kinds are shown based on connector type, charger wattage, and application. It looks into communication methods while making reference to relevant standards for electric vehicles. Various charging modes enable the charging of electric vehicles, where a number of intricate parameters are required to guarantee high-power AC and DC charging. The structure of the ISO 15118 standard is examined together with the description and analysis of the DC communication mechanism. It also explains the OCCP protocol, which guarantees communication between the charger and the operator. Different charging modes allows to charge an electric vehicle, where a series of complex conditions are desirable.

Intelligent Brain Tumor Identification in MRI Using Fuzzy Clustering and Machine Learning

Authors- Subhasis Misra

Abstract--MRI is the main tool for detecting brain tumors. This paper employs data mining methods to categorize MRI scans. A novel hybrid approach combining support vector machine (SVM) and fuzzy c-means is introduced for classifying brain tumors. The suggested algorithm merges support vector machine (SVM) with fuzzy c-means, creating a hybrid technique for predicting brain tumors. In this method, image enhancement is performed through techniques like contrast enhancement and mid-range stretching. Skull stripping is achieved using double thresholding and morphological operations. Fuzzy c-means (FCM) clustering is used to segment an image and identify suspicious areas in a brain MRI. The grey level run length matrix (GLRLM) is utilized for feature extraction from the brain image, followed by the application of the SVM technique to classify brain MRI images, yielding precise and highly effective results for the classification process.

Online Public Distributive Service Administration

Authors- A. Gilbert Paul Ricky, Assistant Professor H.Jayamangala

Abstract--The Public Distribution System (PDS) plays a vital role in providing essential commodities to the economically weaker sections of society. However, the existing manual ration distribution process is plagued with issues such as corruption, inefficiency, and lack of transparency. This paper proposes an advanced web-based solution, Online Public Distributive Service Administration, aimed at modernizing the ration distribution process. The system enables digital registration, authentication, online tracking of commodities, and digital payment mechanisms to improve the accuracy and transparency of the distribution. With role-based modules for administrators and customers, it facilitates real-time stock updates, complaint management, and the issuance of new ration cards. The system categorizes ration entitlements based on card types and automates billing and reporting processes. By transitioning to an online platform, this solution aims to eliminate common flaws in the conventional system and ensure equitable and transparent service delivery to the public.

DOI: /10.61463/ijset.vol.13.issue2.414

GSM-Based Home Automation Using Arduino with Android App

Authors- Mr. Atharva Chakote, Mr. Shivam Dhavle, Dr. Jasbir Kaur, Assistant Professor Mr. Suraj Kanal

Abstract--This paper presents the design and implementation of a cost-effective GSM-based home automation system utilizing an Arduino microcontroller interfaced with an Android application. The system enables remote control of home appliances through SMS commands sent from the Android app, leveraging the global reach of GSM networks. This solution is particularly appropriate for remote or rural locations because it doesn’t require internet access, unlike Wi-Fi or Bluetooth-based systems. Real-time appliance control, an easy-to-use mobile interface, low power consumption, and scalability are some of its key features. According to earlier research on GSM-based automation and embedded control systems, the prototype exhibits excellent responsiveness and dependability, providing a workable solution for contemporary smart homes.

Library Management System Using Android Studio

Authors- Mr. Bhushan Bhurke, Mr. Sumit Jadhav, Dr. Jasbir Kaur, Assistant Professor Ms. Ifrah Kampoo

Abstract--The Rapid learning of mobile technology has revolutionized conventional workflows in various industries, including library management. This research introduces the design, implementation, and testing of a Library Management System (LMS) developed with Android Studio, Google’s leading IDE for Android app development. The system seeks to automate and simplify library processes like book cataloguing, user registration, borrowing/returning, and real-time stock monitoring. Capitalizing on XML for easy-to-use interfaces, Kotlin for solid application logic, and Firebase for cloud-based data storage, the LMS provides a scalable, user-friendly platform. Test results point towards marked improvements in operational efficiency and end-user satisfaction, making the system a viable option against traditional desktop or manual library management strategies.

Online Vehicle Registration Rto System

Authors- Pragadeesh R. K, Professor K. Kumutha

Abstract--The Regional Transport Office (RTO) is a government organisation responsible for issuing driving licenses, maintaining databases of vehicles and sells personalized registration of vehicles. It has been observed since years that the RTO is not able to deliver quality public services to the citizen without delay. That is, it has been a difficult job for citizens to get a driving license and to register their vehicles. Hence, this project is aimed at developing a computerized system for the functioning of RTO This system will reduce the manpower required in the RTO and make the existing system fast and efficient. The aim is to build a user-friendly webpage where the citizens can apply for learner’s license, driving license and vehicle registration. The webpage also provides provision for citizens to submit their complaints he projects intends to provide quality services to the citizens of the state. It does so by reducing the delay in services provided by the RTO through computerization of the system. Imagine if the RTO system was offline based, then the citizens must go to any one of the RTO offices just to apply for learner’s license, driving license and vehicle registration. This is not an easy job for citizens and even for the RTO officials too where they must maintain huge amount of offline records. That’s the primary reason why the s0ystem computerized. This also reduces the burden on manpower working in the RTO. Since the project is web based, the changes or modifications required by the system over a long term of period can be done very easily. This helps in easy system maintenance and be up to date with the user requirements. That means updating the system as per the user requirements will be an effortless job to the system maintenance group.PHP provides support to all major servers like Apache and databases like MySQL. Since it uses its own memory, the loading time is decreased and processing time is increased. Next, we have used HTML,JavaScript, CSS and Bootstrap for front-end implementation. They provide a front-end development framework to create fully responsive web pages and define proper styles and presentation of the document. Lastly, MySQL is used as the back-end database since it is one of the most popular open source databases, and it provides fast data access, easy installation and simplicity.

DOI: /10.61463/ijset.vol.13.issue2.415

Deep Learning for Cloud Removal in Satellite Images

Authors- Dr.Pankaj Malik, Govind Prajapati, Akshad Men, Aradhya Gupta, Yashika Nikhoriya, Aryan Gangrade

Abstract--Cloud coverage in satellite imagery poses a significant challenge for remote sensing applications, obstructing surface features and degrading the quality of environmental monitoring, land use analysis, and change detection tasks. This study proposes a deep learning-based framework for effective cloud removal using a U-Net architecture enhanced with attention mechanisms and residual connections. The model is trained on a dataset comprising paired cloudy and cloud-free satellite images from the Landsat-8 and Sentinel-2 missions. Data augmentation and spectral band fusion techniques are utilized to enhance model generalizability across different geographies and seasons. Quantitative evaluations using Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), and Mean Absolute Error (MAE) demonstrate the effectiveness of the proposed model. The framework achieves a PSNR of 32.8 dB, SSIM of 0.92, and MAE of 0.035, outperforming traditional cloud removal methods and baseline deep learning models by a significant margin. Visual inspection also reveals high-fidelity reconstruction of obscured land features and minimized spectral distortion. The results suggest that the proposed approach is a promising solution for enhancing the usability of satellite data in the presence of cloud cover, and it can be integrated into preprocessing pipelines for downstream applications such as crop monitoring, urban development analysis, and disaster management.

Data Security and Privacy Protection in Cloud Storage

Authors- S. Harshith, Assistant Professor Dr. H. Jayamangala

Abstract--The rapid rise of IoT, smart cities, digital transformation, and the global digital economy has driven exponential data growth, placing immense pressure on storage systems. Cloud storage has become essential for managing this surge, with governments, enterprises, and individuals migrating data to the cloud. However, this brings significant risks, including unauthorized access and data leakage. This study reviews existing literature on data security and privacy in cloud storage, focusing on encryption technologies and protection strategies to address these emerging challenges.

Cyber Threat Detection Using Machine Learning

Authors- S. Vetrivel, Assistant Professor Dr. A. Poongodi

Abstract--This research suggests a method employing machine learning techniques for identifying and countering cyber-crimes on social networking platforms. An online application is employed using the Django and Flask frameworks for classifying and visualizing potentially offensive messages, such as instances of cyberstalking and bullying. Supervised machine learning algorithms, i.e., Support Vector Machine (SVM) and Naive Bayes classifiers, are employed for sentiment analysis and threat identification on text data. An interactive dashboard is provided for presenting analysis insights, such as time-series trends of sentiment, topic modeling results, and different affective aspects. The site is made compatible with the likes of Facebook and Twitter and enables proactive identification and moderation of offensive content.

Heart Disease Prediction Using Random Forest Algorithm and Machine Learning Techniques

Authors- S. Hari Krishnan, Assistant Professor Dr. V. Sumalatha

Abstract--This paper introduces a machine learning-based cardiovascular disease prediction system with focus on Random Forest and Decision Tree classifiers. Based on clinical data in the UCI repository, the model goes through intensive data preprocessing, feature selection, and classification modeling to enhance the prediction of cardiovascular diseases. The model has a good accuracy of 92%, outperforming traditional algorithms such as SVM and Naive Bayes. The solution employs Python for front-end analytics and PostgreSQL as the backend, showcasing a cost-effective and efficient diagnostic tool in the healthcare industry. The system not only improves early diagnosis but also minimizes the clinical workload, enhances decision-making, and provides real-time prediction through an easy-to-use web interface.

Exploring the Effects of Blockchain Based Security Systems on Cyber Security

Authors- R. Thilak Raaj, Assistant Professor Dr. C. Meenakshi

Abstract--The unbridled proliferation of cyber assets and data exchange on the internet has led to an alarming increase in cyber threats, necessitating innovative cybersecurity solutions. This study recommends the implementation of Blockchain-Based Cybersecurity Systems (BCSSs) to increase the integrity, transparency, and robustness of secure information management in modern organizations. Leaning on the decentralized nature of blockchain, the system ensures tamper-proof storage, auditability, and end-to-end traceability— effectively mitigating risks due to unauthorized access, insider threats, and system intrusions.

Airline Price Prediction Using Machine Learning

Authors- S. Madhan, Assistant Professor Dr. Lipsa Nayak

Abstract--Airline ticket prices are highly volatile and subject to numerous variables such as demand, travel time, airline, route, and booking time. Volatility of this kind will be inconvenient for travellers, especially when booking trips at a low price. The objective of this project is to design a machine learning model that can accurately predict airline ticket prices from historical flight data and a variety of features that come with flights. We downloaded and cleaned a dataset of flight data such as airline name, source and destination cities, departure and arrival time, flight duration, stops, and ticket price for the project.

Multi-Task Learning for Predicting Wind Turbine Failures and Remaining Useful Life (RUL)

Authors- Harsh Gupta, Harsh Malviya, Harsh Prajapat, Dr. Pankaj Malik, Prof. Sonal Modh

Abstract--Wind turbines are critical for sustainable energy production, but unexpected failures and inefficient maintenance strategies can lead to significant operational costs and energy losses. Traditional machine learning approaches often treat failure prediction and Remaining Useful Life (RUL) estimation as separate tasks, limiting their ability to leverage shared knowledge. This paper proposes a Multi-Task Learning (MTL) framework that simultaneously predicts wind turbine failures and estimates RUL by utilizing shared feature representations. The proposed deep learning model integrates Long Short-Term Memory (LSTM) networks with task-specific output layers, enabling the extraction of temporal dependencies from Supervisory Control and Data Acquisition (SCADA) sensor data. The model is trained and tested on a real-world dataset containing historical wind turbine sensor readings, including temperature, vibration, and pressure data. Experimental results demonstrate that the MTL approach achieves 12.5% higher accuracy in failure prediction and 18.3% lower RMSE in RUL estimation compared to traditional single-task models. Additionally, the model provides early failure warnings up to 30% sooner, allowing for improved maintenance planning. These findings highlight the effectiveness of MTL in enhancing predictive maintenance strategies, ultimately reducing downtime and optimizing wind farm operations.

Attacks on RSA Cryptosystem

Authors- Nidhi Patel, Priyanshu Prajapati, Ekta Gajjar, Nandini Prajapati , Milan Panchal, Twinkle Patel

Abstract--The RSA cryptosystem is a foundational component of modern public-key cryptography, widely utilized for securing digital communication. Despite its robust theoretical underpinnings and extensive adoption, RSA is vulnerable to several attacks if improperly implemented. This paper explores various types of attacks on the RSA algorithm, including mathematical, protocol-level, and side- channel attacks. It provides detailed explanations of Wiener’s attack, Bleichenbacher’s padding oracle attack, and Coppersmith’s method, highlighting their mechanisms and implications. The paper also discusses essential protective measures, such as optimal key sizes and secure padding schemes like OAEP. Emphasis is placed on the importance of best practices and evolving standards to maintain RSA’s effectiveness in the face of emerging cryptographic threats.

Revocable Encryption for Secure Data Sharing and Malicious Attack Prevention

Authors- Magalakshmi.M, Professor, Dr.V.Sumalatha

Abstract--Distributed computing enables flexible and efficient information exchange, offering considerable advantages to both individuals and organizations. However, due to the sensitive nature of the data often shared in such environments, robust security measures are essential to prevent misuse and unauthorized access. A major concern in cloud-based data sharing is the inability to revoke access effectively once it has been granted. Traditional encryption schemes lack the flexibility to enforce dynamic access control, especially when user roles or trust levels change. This limitation leads to potential data leakage or continued access by malicious users. To address this, cryptographically enhanced access control mechanisms, such as attribute-based encryption (ABE) and proxy re-encryption (PRE), provide promising solutions. These methods support fine-grained permission management, allowing data owners to revoke specific user access without re-encrypting the entire dataset. When a user’s permission is revoked, the system ensures complete removal of access from the cloud environment, thereby preventing any further retrieval of the data. The proposed revocable encryption model emphasizes reliability, usability, and effectiveness in securing shared data while incorporating mechanisms to detect and prevent malicious activities. This intelligent and practical framework not only of a secure and adaptable data-sharing architecture.

DOI: /10.61463/ijset.vol.13.issue2.416

Detecting E-banking Phishing Using Associative Classification

Authors- Kavin K T, Assistant Proffesor Dr.K.Krithika

Abstract--There are number of users who purchase product online and make payment through e- banking. There are e- banking websites who ask user to provide sensitive data such as username, password or credit card parts etc. often for malicious reason. This type of e-banking website is known as phishing website. In instruction to detect and predict e-banking phishing website, we planned an smart, flexible and effective method that is based on by means of classification Data mining algorithm. We implemented classification algorithm and systems to extract the phishing data sets criteria to classify their legitimacy. The e-banking phishing website can be recognized built on some significant characteristics like URL and Domain Identity, and safety and encryption criteria in the last phishing detection rate. Once user makes deal through online when he makes payment through e-banking website our system will use data mining algorithm to detect whether the e-banking website is phishing website or not.

DOI: /10.61463/ijset.vol.13.issue2.417

Crop Recommendation System Using Machine Learning

Authors- Jeevanantham R, Dr. V. Sumalatha

Abstract--Approximately 17% of India’s GDP derives from the agricultural sector, which even employs more than 60% of the country’s workforce. This field has seen some changes with new technologies like vertical farming and so on. However, many Indian farmers still follow traditional ways and beliefs to use their land. For example, they wait for the weather to match their farming practices, rather than adjusting to the weather changes. Our research goal is to help farmers pick the best crops for their situation and environment by predicting which crops fit well with the factors that influence crop growth, such as soil nutrients, soil pH, humidity and rainfall. We use different machine learning models, such as Decision Tree (DT), Support Vector Machine (SVM), Logistic Regression (LR), and Gaussian Naïve Bayes (GNB).

AI Powered Online Recruiting App

Authors- Subiksha Devi J, Assistant Professor Dr.Krithika.D.R

Abstract--The AI-Based Recruiting Application is a web platform designed to automate and optimize the hiring process using artificial intelligence. It streamlines candidate sourcing, screening, and matching by leveraging machine learning and natural language processing (NLP) to analyse resumes and job descriptions. Employers can post job openings, and AI algorithms rank candidates based on skills, experience, and qualifications, ensuring an efficient and data-driven approach. The platform also automates interview scheduling, reducing administrative tasks for recruiters and improving the candidate experience. A key feature is its ability to minimize hiring bias by focusing on objective criteria while eliminating subjective influences. Additionally, the system provides insightful analytics to track recruitment metrics such as time-to-hire and candidate success rates. Built using Django, the platform ensures scalability, security, and effective data management, making recruitment faster, fairer, and more reliable.

DOI: /10.61463/ijset.vol.13.issue2.418

Micro-teaching Practices in EFL Teacher Education: A Systematic Review of Methods, Outcomes, and Challenges

Authors- Shema Bashir Ali Zrike

Abstract--Microteaching has emerged as a pivotal component in the professional preparation of English as a Foreign Language (EFL) teachers, offering a structured environment for the development of instructional skills, self-reflection, and confidence. This systematic review explores how microteaching is implemented in EFL teacher education programs across diverse contexts, focusing on the instructional methods employed, the pedagogical outcomes achieved, and the recurring challenges encountered. Findings reveal that microteaching promotes a range of benefits, including enhanced lesson planning, improved classroom management, increased teaching confidence, and deeper reflective practice. However, challenges such as artificial classroom settings, technological limitations, and learner anxiety persist, often hindering the full potential of microteaching. The review highlights the importance of supportive feedback structures, integration with digital tools, and culturally responsive practices to optimize its impact. By synthesizing insights from existing literature, this paper offers practical implications for enhancing the design and delivery of microteaching in EFL teacher education globally.

E-Commerce Recommendation System

Authors- Mahitharini S, Assistant Professor Dr. Jayamangala Hariharan

Abstract--EcoMart is an innovative e-commerce platform developed using Django web framework that focuses on environmentally conscious electronic products. With the growing awareness of environmental impacts from electronic devices, there is an increasing demand for eco-friendly alternatives. EcoMart addresses this market gap by featuring an eco-score rating system that allows consumers to make environmentally informed purchasing decisions. The platform incorporates a recommendation engine based on cosine similarity to suggest products with similar environmental credentials. The solution enhances consumer awareness by providing transparent sustainability metrics integrated directly into the shopping experience. With a responsive design and comprehensive user authentication system, it supports research into sustainable consumer behavior and promotes environmentally responsible purchasing patterns in the electronics sector.

DOI: /10.61463/ijset.vol.13.issue2.419

AI-Driven Discovery of Nanomaterial Synergies for Next-Generation Antibiotic Alternatives

Authors- Manasa

Abstract-– The rise of antimicrobial resistance (AMR) is an urgent global health concern, necessitating the development of alternative strategies to combat resistant pathogens. Nanomaterials, with their unique physicochemical properties, offer significant promise as next-generation antimicrobial agents. However, the identification of effective nanomaterial combinations remains a complex challenge due to the vast array of possible properties and interactions. Artificial intelligence (AI), particularly machine learning (ML), has emerged as a powerful tool to navigate this complexity. By analyzing high-dimensional datasets, AI models can predict synergistic combinations of nanomaterials with enhanced antimicrobial efficacy while minimizing toxicity. Moreover, generative AI models, such as variational autoencoders and generative adversarial networks, facilitate the de novo design of novel nanomaterial structures, enabling the exploration of new therapeutic candidates. AI’s integration with omics data also provides mechanistic insights into nanomaterial-microbe interactions, which is crucial for designing materials that are both effective and resistant to resistance development. Despite the promise, challenges remain, including the need for high-quality, standardized datasets and the interpretability of AI models. Future research should focus on overcoming these challenges by developing transparent AI systems and fostering collaboration between computational scientists, microbiologists, and clinicians. This approach holds the potential to accelerate the development of nanomaterial-based antibiotics, offering a new frontier in the fight against AMR.

DOI: /10.61463/ijset.vol.13.issue2.425

Graph-Based AI for Tracking Horizontal Gene Transfer in Microbial Communities

Authors- MOHAN K

Abstract-– Horizontal gene transfer (HGT) is a pivotal mechanism in microbial evolution, enabling the exchange of genetic material between organisms and contributing to the spread of important traits such as antibiotic resistance. Understanding and tracking HGT events within microbial communities is crucial for addressing challenges in public health, environmental management, and industrial biotechnology. However, due to the complexity of microbial interactions and the diverse mechanisms through which HGT occurs, traditional methods often fail to capture the full scope of these processes. Graph-based artificial intelligence (AI) offers a promising approach to modeling and tracking HGT events, providing a powerful framework for understanding microbial gene exchange in ways that conventional techniques cannot. This article explores the application of graph-based AI methods to study HGT dynamics and their potential in tracking and predicting gene transfer events within microbial communities.

DOI: /10.61463/ijset.vol.13.issue2.426

Machine Learning Approaches to Engineer Nanoantibiotics for Treating Infections in Immunocompromised Patients

Authors- PAVAN T.K

Abstract-– The increasing prevalence of drug-resistant infections among immunocompromised individuals presents a critical challenge in modern medicine, as these patients are particularly vulnerable to infections that often fail to respond to traditional antibiotics. Nanoantibiotics, which include nanoscale materials with inherent antimicrobial properties or those serving as delivery systems for antibiotics, offer a promising therapeutic solution. However, the design of nanoantibiotics requires careful optimization of multiple parameters to ensure their efficacy and safety. Machine learning (ML) has emerged as a transformative tool in the development of nanoantibiotics, enabling the prediction of complex biological and physicochemical interactions. By utilizing large datasets from experimental and clinical studies, ML models can predict antimicrobial potency, toxicity, drug release profiles, and stability, thus reducing the need for extensive trial-and-error experimentation. Moreover, ML facilitates the identification of non-linear relationships between nanoparticle features and therapeutic outcomes, providing deeper insights into nanoparticle design. The integration of machine learning with experimental synthesis platforms can expedite the development of optimized nanoantibiotics, while also enabling the personalization of therapies tailored to individual patient needs. This paper explores the applications, challenges, and future potential of machine learning in the rational design of nanoantibiotics for immunocompromised patients.

DOI: /10.61463/ijset.vol.13.issue2.427

Federated AI for Privacy-Preserving Analysis of Wearable Nanobiosensor Data

Authors- Shruthi

Abstract--The emergence of wearable nanobiosensors has opened new frontiers in real-time, continuous health monitoring by enabling detection of physiological and biochemical changes at the molecular level. These sensors generate highly personalized data streams that are crucial for early diagnosis, chronic disease management, and adaptive therapeutic interventions. However, the sensitive nature of such data introduces substantial privacy risks, especially in centralized data processing systems. Federated artificial intelligence (AI) has recently gained traction as a solution for decentralized machine learning that protects user data. This article investigates the integration of federated AI with wearable nanobiosensor networks, emphasizing its ability to facilitate large-scale, privacy-respecting biomedical analytics without compromising the utility or integrity of health data.

DOI: /10.61463/ijset.vol.13.issue2.428

AI Models for Multiplexed Detection of Microbial Toxins Using Nanoplasmonics

Authors- Simreena

Abstract--The precise and rapid identification of microbial toxins is vital in managing infectious diseases, ensuring food and water safety, and protecting public health. Nanoplasmonic biosensors have emerged as promising tools due to their sensitivity, specificity, and label-free operation. However, as these platforms are scaled for multiplexed detection—simultaneously targeting multiple toxins—the complexity of signal interpretation grows significantly. Artificial intelligence (AI), particularly through machine learning and deep learning techniques, provides a solution for deciphering the intricate spectral patterns produced during multiplexed assays. This article explores the synergy between AI and nanoplasmonic sensing technologies to enable efficient, real-time, and high-throughput detection of microbial toxins.

DOI: /10.61463/ijset.vol.13.issue2.429

Prediction of Microbe-Host Interaction Networks Using Graph Neural Networks

Authors- VINAY. K

Abstract-– Mymensingh is the capital of Mymensingh Division in central Bangladesh. Ensuring water sanitation and hygiene in Mymensingh is vital for community health, requiring effective measures and collaboration between authorities, organizations, and residents for sustainable implementation. The objectives of this study were to investigate the water supply and sanitation status of Mymensingh City. Data were collected primarily based on a reconnaissance survey with the help of a structured questionnaire. A cross-sectional survey design was employed to collect data on variables related to water, sanitation, and hygiene in the area. Many homes rely on submersible pumps and deep tube wells for drinking water, while access to piped water is limited. Inadequate water supply and limited access to clean water contribute to waterborne illnesses and negatively impact public health. Sanitation infrastructure in Mymensingh City Corporation varies, with reliance on septic tanks and pit latrines, while limited sewage systems and waste management exist. Inconsistent hygiene practices contribute to waterborne illnesses, highlighting the need for improved infrastructure and behavior change interventions. Improving the drainage system, implementing effective measures for waste management, and promoting hygiene education programs are essential for minimizing waterborne diseases and enhancing residents’ quality of life.

DOI: /10.61463/ijset.vol.13.issue2.430

AI-Based CRISPR Guide RNA Design for Viral Genome Editing

Authors- Yogesh K

Abstract-– The development of CRISPR-Cas systems has opened new frontiers in genome editing, with applications extending into virology for targeting and disabling viral genomes. A critical component of CRISPR technology is the guide RNA (gRNA), which directs the Cas enzyme to specific genomic loci. The precision and efficiency of viral genome editing depend heavily on the accurate design of these gRNAs. Artificial intelligence (AI) has emerged as a transformative tool in this domain, providing predictive models that enhance the selection of effective gRNA sequences while minimizing off-target risks. This research article explores how AI-based approaches are being integrated into gRNA design pipelines to optimize viral genome editing, particularly in therapeutic contexts.

DOI: /10.61463/ijset.vol.13.issue2.431

ML-Based Functional Annotation of Hypothetical Proteins in Bacterial Genomes

Authors- VAISHNAVI

Abstract-– Bacterial genomes contain numerous genes that encode hypothetical proteins—proteins whose functions remain unknown due to lack of experimental validation or sequence-based annotation. The presence of these uncharacterized proteins represents a significant gap in our understanding of microbial biology and their potential roles in various physiological processes, including virulence, antimicrobial resistance, and metabolism. Traditional methods of protein annotation rely heavily on sequence homology, which may not always be applicable to novel or poorly characterized proteins. Machine learning (ML) techniques have emerged as powerful tools for overcoming these limitations. This article explores the application of ML-based approaches to functionally annotate hypothetical proteins in bacterial genomes, highlighting the potential of these methods to predict protein functions, identify novel protein families, and contribute to the advancement of microbiological research.

DOI: /10.61463/ijset.vol.13.issue2.432

Deep Generative Models for Engineering Synthetic Microbes for Bioremediation

Authors- Vinod

Abstract--The increasing burden of environmental pollutants from industrial waste, plastic accumulation, and toxic chemical spills necessitates innovative and sustainable remediation strategies. Synthetic biology offers a powerful means to address this issue by enabling the design of microbes capable of degrading and neutralizing harmful substances. However, the design of such microbes is complicated by the vast combinatorial space of genetic elements and the complex behavior of biological systems. Deep generative models, a class of artificial intelligence techniques capable of producing novel biological sequences and pathways, provide a promising approach to accelerate and optimize the engineering of synthetic microbes for bioremediation. This research discusses how these models can support the generation of novel enzymes, regulatory elements, and whole-cell designs for improved environmental detoxification.

DOI: /10.61463/ijset.vol.13.issue2.433

Reinforcement Learning for Adaptive Traffic Signal Timing in Range-Based TMS (Traffic Management System)

Authors- Associate Professor Dr .M Purna Kishore, P. V. Sai Bharadwaj, T. N. V. Satya Sandeep, Sk. Nazeer, P. Siva Krishna, Sk. Abdul Rasheed

Abstract--Conventional traffic signal control systems frequently fail to adapt to fluctuating traffic patterns, resulting in suboptimal traffic flow and heightened congestion levels. This project introduces an innovative solution that harnesses the capabilities of Reinforcement Learning (RL) to enhance traffic signal timing within a Range-Based Traffic Management System (RTMS). The RL agent is designed to optimize signal timing decisions. It receives feedback from the system, and there by reductions in congestion and enhancements in traffic flow. From a process of iterative learning, the RL agent refines its decision- making capabilities, resulting in progressively more efficient traffic signal management. Through this paper we tried to present a progress in the existing manual traffic control system.

DOI: /10.61463/ijset.vol.13.issue2.430

Secure Communication System Using Quantum Cryptography

Authors- S. Bala Chandran, Dr..S. Prasanna

Abstract--Quantum cryptography offers an attack-proof model for communication that is resistant to quantum computing advances. This study proposes a hybrid model that unifies Quantum Key Distribution (QKD) with Post-Quantum Cryptography (PQC) to ensure security in diverse networks. Efficient key distribution, life-cycle management, and performance monitoring are facilitated through a collaborative Quantum Key Management System (QKMS) and Q-controller. By having QKD in the center and using PQC in the user terminals, this model overcomes the limitations based on the range and expense of QKD. Scalable, fault-tolerant, and end-to-end secure communications are made possible by the proposed approach.

Data Leakage Detection in Cloud Computing Environment

Authors- S.Dhanish, Dr.R. Priya Anand

Abstract--Internet technology have emerged as the foundation of any company organization in recent years. By moving data between locations, many firms utilize this capability to increase their efficiency. Transferring important corporate data, however, carries a lot of risks because any negligent employee could make the information public. Data leakage is the term for this issue. We are proposing a model for the data leaking problem in this paper. Our goal in this approach is to find the person responsible for the leak of important organizational data. Because more and more people are using cloud computing services, data leak detection is a crucial area of study. Finding any data leaks or unauthorized access is essential to preventing data breaches as more and more data is processed and stored on the cloud. This research paper’s goal is to suggest a method for detecting data leaks in cloud computing settings. To identify data leakage, the suggested system employs a hybrid strategy that blends static and dynamic analysis methods. While dynamic analysis is done during runtime to find any real data leaks, static analysis is done on the source code to find possible data leakage sites.

AI-Enhanced Agricultural Robotics: From Precision Farming to Autonomous Harvesting

Authors- Arjun. P

Abstract--The integration of artificial intelligence (AI) in agricultural robotics has led to significant advancements in farming practices, from precision farming techniques to autonomous harvesting. AI-enhanced agricultural robots are designed to optimize crop production, improve yield, reduce resource consumption, and automate labor-intensive tasks. By using advanced algorithms and machine learning, these robots can monitor crop health, detect pests, perform weeding, and assist in the harvesting process, all while minimizing environmental impact. This paper explores the role of AI in agricultural robotics, examining its applications, current trends, and challenges. It also discusses the ethical and regulatory considerations surrounding the deployment of such technologies and the future of AI-enhanced agriculture in addressing global food security challenges.

DOI: /10.61463/ijset.vol.13.issue2.431

Analysis of Cardiovascular Diseases with Causes and Future

Authors- Vishav Mehra

Abstract--A complicated clinical disease known as heart failure manifests as a variety of symptoms and indicators. Even with improvements in our knowledge of its pathogenesis and the availability of several treatment options, it continues to be a major source of morbidity and death. Precise risk assessment is necessary to direct managerial choices and support patient-focused treatment. While recognized risk variables have historically been the focus of epidemiological studies to predict mortality in heart failure, newer machine learning algorithms provide a unique method for uncovering other important predictors. In this review, we examine the role of supervised learning algorithms in predicting mortality in heart failure, with a view to facilitating the development of evidence-based treatment guidelines and health care policies.

DOI: /10.61463/ijset.vol.13.issue2.433

House Price Prediction Using Machine Learning Teachniques

Authors- P. Ajay, Assistant Professor Dr. K. Nandhini

Abstract--Accurate house price prediction is a critical component of real estate analytics, impacting buyers, sellers, investors, and policymakers. This paper presents a comprehensive approach to forecasting house prices using advanced machine learning algorithms and statistical methods. It explores the influence of key factors such as property size, location, amenities, market trends, and economic conditions on house prices. The findings highlight the potential for machine learning to transform real estate decision-making, offering valuable insights into pricing strategies and market assessment. Future research directions include integrating geospatial data and economic indicators for even more precise forecasting.

DOI: /10.61463/ijset.vol.13.issue2.434

Deepfer: Camera Vision Based Automatic Facial Expression Analysis System Using Deep Learning Techniques

Authors- E. Ragul, Assistant Professor Dr. Lipsa Nayak

Abstract--Facial expressions are a powerful form of non-verbal communication and play a crucial role in assessing human emotions. This paper presents DeepFER, a camera vision-based automatic facial expression recognition system that leverages deep learning to analyze emotions in real time. The system uses a Deep Convolutional Neural Network (DCNN) to classify seven basic facial expressions—angry, disgust, fear, happy, sad, surprise, and neutral—captured through live or offline video feeds. A custom dataset of annotated facial expressions is used to train the DCNN model, ensuring high accuracy across varied lighting and environmental conditions. With its accurate classification capabilities and scalable architecture, DeepFER presents a significant step toward intelligent emotion-aware systems.

DOI: /10.61463/ijset.vol.13.issue2.436

Cybercecurity Awareness Platform Using Gamafication

Authors- K. Abhiraj Mohan, Professor Dr. S. Prasanna

Abstract--Cybersecurity threats continue to rise, posing significant risks to organizations by exploiting human vulnerabilities, particularly among employees. Traditional security awareness programs, relying on slideshows, classroom training, and videos, often fail to sufficiently engage or educate users. This research introduces a gamified cybersecurity awareness program known as the Zero-Day Awareness Program (ZDAP), designed to enhance employees’ knowledge and response to cyber threats through interactive learning. This study reinforces the importance of experiential learning in cybersecurity education and encourages organizations to adopt gamified methods for security training.

DOI: /10.61463/ijset.vol.13.issue2.437+

Lung Cancer Risk Prediction with Machine Learning

Authors- Rafael Jernaldin Raj , Assistant Professor Dr. Lipsa Nayak

Abstract--Lung cancer remains a major contributor to cancer-related mortality worldwide, highlighting the critical need for early detection and accurate risk prediction. This project introduces a machine learning-based lung cancer risk prediction system that analyzes clinical, demographic, and imaging data to deliver real-time, accurate assessments. Using supervised learning techniques like Random Forest and XGBoost for structured data, and convolutional neural networks (CNNs) for medical imaging, the system offers a comprehensive evaluation of a patient’s cancer risk. With automated data processing, explainability tools like SHAP and LIME, and integration with electronic health records (EHRs), the system enhances clinical decision-making and promotes timely interventions.

DOI: /10.61463/ijset.vol.13.issue2.438+