Volume - 12 Issue - 5

Area of research: Computer Application

The Smart Mobile Application Is An Advanced Digital Platform Developed To Streamline Communication And Service Management Between Patients And Doctors Through A Secure And User Friendly Mobile Environment. The System Is Designed With Two Primary Modules: Patient Module And Doctor Module, Enabling Efficient Healthcare Interactions And Appointment Management. The Application Is Developed Using Flutter, Which Provides A High- Performance Cross-platform User Interface With Responsive Design, While Firebase Is Integrated As The Backend Solution For Authentication, Real-time Database Management, And Secure Cloud Storage.In The Patient Module, Users Can Register, Log In Securely, And Access A Personalized Dashboard To Manage Their Healthcare Activities. Patients Are Able To Upload And Maintain Medical Records, Review Previous Health History, Search For Doctors Based On Specialization And Availability, And Schedule Appointments Conveniently. Once Appointments Are Approved, Patients Canconsult Doctors And Receive Digital Prescriptions Through The Application.In The Doctor Module, Medical Professionals Can Create And Manage Profiles, Update Consultation Schedules, Monitor Incoming Appointment Requests, And Review Patient Medical Information Before Confirmation. After Consultation, Doctors Can Generate Prescriptions, Recommend Medications, And Provide Treatment Guidancedigitally.The Proposed System Minimizes Traditional Appointment Delays, Improves Patient Record Management, Enhances Doctor- Patient Communication Through Real-time Digital Services. This Project Demonstrates The Practical Implementation Of Mobile Application Development, Cloud Computing, Healthcare Automation, And Real-time Scheduling Using Modern Technologies. . It Supports Real-time Communication Between Doctors And Patients, Making Healthcare Services More Efficient And Organized. The Application Can Be Expanded In The Future By Adding Features Such As Online Payment System The Application Is Developed Using Flutter, Which Provides A High-performance Cross- Platform User Interface With Responsive Design, While Firebase Is Integrated As The Backend Solution For Authentication, Real- Time Database Management, And Secure Cloud Storage.In The Patient Module, Users Can Register, Log In Securely

Author: Abdul Shariff K S | Ms. Dharani A

Area of research: Computer Applications

The Rapid Advancement Of Automobile Manufacturing Technologies Has Increased The Need For Secure, Efficient, And Transparent Production Workflows. Traditional Manufacturing Systems Rely Heavily On Fragmented Data Handling And Manual Verification Processes, Which Often Lead To Data Inconsistencies, Operational Delays, And Reduced Production Efficiency. This Paper Presents A Blockchain-Based Automobile Production Data Security And Workflow Optimization Framework Designed To Provide Secure Multi-stage Workflow Management Across Automobile Production Environments. The Proposed System Integrates Blockchain Technology, AES Encryption, Automated Validation Mechanisms, And Rolebased Access Control To Ensure Secure Data Management Throughout Design Review, Component Validation, Quality Control, Testing, And Report Generation Stages. The Framework Enables Real-time Validation Of Seating Configurations, Fuel Or Battery Capacities, And Component Specifications While Maintaining Transparency And Traceability. The System Is Implemented Using Java, Servlets, JSP, MySQL, And Blockchain-supported Secure Workflow Mechanisms. Experimental Evaluation Demonstrates Improved Workflow Efficiency, Enhanced Data Integrity, Secure Access Management, And Reduced Operational Errors. The Modular Architecture Supports Future Extensions Including AI-driven Analytics, Cloud Deployment, And Smart Manufacturing Integration.

Author: PRASHANTH KUMAR E | Ms. Dharani

Area of research: NA

The Rapid Digitalization Of Recruitment Has Created The Need For Automated, Accurate, And Unbiased Resume Screening Systems. This Research Presents An AI-powered Resume Analyzer That Utilizes Natural Language Processing (NLP), Machine Learning (ML), And Semantic Matching Techniques To Evaluate Resumes Efficiently. The System Extracts Key Information Such As Skills, Experience, Education, And Achievements Using Text-processing Algorithms And Transforms Them Into Structured Data. A Machine-learning–based Relevance Model Then Compares Candidate Profiles With Job Descriptions To Generate A Match Score, Highlight Missing Skills, And Provide Improvement Suggestions. The Proposed System Reduces Manual Screening Time, Enhances Decision-making Accuracy, And Minimizes Human Bias. Experimental Results Demonstrate That The AI Resume Analyzer Improves Candidate–job Matching Efficiency And Delivers Consistent, Objective Evaluations, Making It A Valuable Tool For Modern Recruitment Workflows.

Author: Kamale D.S. | Kamale D.S. | Nagargoje D.G. | Jagadale M.S. | Shelke M.R.

Area of research: AWS

The Rapid Growth Of Cloud Computing And Distributed Systems Has Increased The Complexity Of Managing Large-scale Enterprise Infrastructures. Modern Platforms Such As Microsoft Azure And Microservice-based Environments Require Efficient Monitoring And Failure Detection Mechanisms To Ensure High Availability, Reliability, And Performance. The Proposed Project, “Microsoft Failure Detection System,” Is Designed To Identify, Monitor, And Predict System Failures Occurring In Cloud Servers, Applications, And Network Services In Real Time.The System Uses Intelligent Monitoring Techniques To Continuously Analyze System Parameters Such As CPU Usage, Memory Utilization, Server Response Time, Network Traffic, And Application Health Status. By Collecting And Processing These Metrics, The Proposed Model Can Detect Abnormal Behavior And Generate Early Failure Alerts Before Critical System Breakdowns Occur. The System Aims To Reduce Downtime, Improve Fault Tolerance, And Enhance The Overall Efficiency Of Enterprise Cloud Infrastructures.

Author: Samarth Dipak Dedge | Samarth Dipak Dedge | Aniket Ghogare | Suyash Kaspate | Ranjit Bhange

Area of research: Mechanical Engineering – Machine Design & Manufacturing Engineering

Swivel Joints Are Important Rotating Mechanical Components Used In Industrial, Agricultural, Marine, Chemical, And Hydraulic Piping Systems Where Rotational Movement Is Required Without Interrupting Fluid Transfer. This Paper Presents The Design, Fabrication, And Performance Analysis Of A Swivel Joint Capable Of Continuous 360-degree Rotational Movement. The Developed System Minimizes Hose Twisting, Reduces Leakage, Improves Flexibility, And Increases Operational Efficiency. The Work Includes Material Selection, CAD-based Conceptual Design, Fabrication Methodology, Assembly Process, Testing Procedures, And Result Analysis. Stainless Steel Swivel Joints Demonstrated Excellent Corrosion Resistance, Higher Pressure Handling Capability, And Long Service Life. The Study Also Discusses Future Developments Such As IoT-enabled Monitoring Systems, Predictive Maintenance, And Smart Sealing Technologies.

Author: Yash Satyanarayan Myana | Prof. D. A. Jhaveri | Gauri Kailas Lanke | Neha Rajendra Jadhav

Area of research: Computer Science Engineering

The AI Voice Email Assistant Is A Web-based Application That Enables Users To Interact With Email Services Entirely Through Voice Commands. The System Uses Speech Recognition And Text-to-speech And Speech-to-text Technologies To Allow Users To Compose, Send, Read, And Manage Emails Without Requiring Physical Interaction With A Keyboard Or Mouse. This Solution Is Particularly Beneficial For People With Disabilities, Elderly Users, And Individuals Who Prefer Hands-free Communication. By Combining Artificial Intelligence (AI) With Voice-based Interaction, The Application Improves Accessibility, Convenience, And Productivity In Email Communication.

Author: Vaishnavi Sadanand Pegada | Rohit Shivsharan Dhange | Vishwadeep Prashant Devane | Aditya Sanjay Gire | Prof. Kambale S.A.

Area of research: Computer Applications

Fetal Health Assessment Is A Critical Component Of Prenatal Care, Directly Influencing The Safety And Outcomes Of Both Mother And Fetus During Pregnancy. Conventional Diagnostic Approaches, Which Rely Heavily On Manual Interpretation Of Cardiotocography (CTG) Readings And Clinical Observations, Are Often Susceptible To Subjectivity, Latency, And Diagnostic Inaccuracies, Particularly In Resource-constrained Healthcare Environments. This Study Proposes An AI-driven Predictive Modeling Framework That Employs Advanced Machine Learning Techniques To Classify Fetal Health Conditions Into Three Categories: Normal, Suspect, And Pathological. The System Analyzes Structured Physiological And Clinical Datasets Including Fetal Heart Rate Variability, Uterine Contraction Signals, And Maternal Health Indicators. A Comparative Evaluation Of Multiple Classifiers—including Random Forest, Support Vector Machine, Gradient Boosting, And Decision Tree—was Conducted To Identify The Most Effective Model. Feature Selection And Data Preprocessing Techniques Were Applied To Improve Model Accuracy And Computational Efficiency. Experimental Results Demonstrate That The Proposed Framework Achieves A Classification Accuracy Of 97.8%, With High Sensitivity And Specificity. This System Provides Obstetricians And Gynecologists With A Reliable, Automated Decision-support Tool, Reducing Diagnostic Delays And Enhancing Prenatal Care Quality, Particularly In Rural And Under-resourced Clinical Settings.

Author: C.Jesi | Dr. S. Senthamariselvi

Area of research: Civil Engineering

The Construction Industry Is Using Cloud-based Management Systems More And More To Improve The Way People Talk To Each Other Work Together And Manage Projects. The Old Ways Of Talking To Each Other In Construction Often Have Problems Like Delays People Not Having The Right Information And People Not Working Together. This Research Looks At How Cloud-based Toolsre Changing The Way People Communicate In Construction By Studying Digital Management Systems Like Autodesk BIM 360 Procore And Cloud-enabled Common Data Environments. The Research Checks How These Systems Help People Share Information In Time Make Decisions And Work More Efficiently And How They Help People Work Together Better. The Research Used A Mix Of Methods Including Looking At What Other People Have Written, Studying Specific Cases And Checking How Well Communication Is Working. The Results Show That Cloud-based Systems Really Help Reduce Delays In Communication Make Things More Transparent Reduce The Need To Do Things Over And Make Projects More Productive. However There Are Still Problems Like Concerns About Cybersecurity The Need For Training And People Not Wanting To Change To Systems. The Research Says That Cloud-based Tools Are Very Important For Communicating In Construction Projects Today And Suggests That Companies Should Have A Plan For Implementing These Tools To Get The Most Benefit, From Them.

Author: Mamta Sanjay Dawale

Area of research: Structural Engineering

In The Present Time The World Wide Cement Production Is About 1.6 Billion Tons. This Huge Amount Of Production Leads To Consumption Of Natural Resources And It Is Also Harmful For Environment. The Main Components Of The Gases Emitted From Cement Industries Are CO2, N2, O2, SO2, Water Vapors And Micro Components I.e. CO And NOx. Large Quantity Of Waste By Products Are Produced From The Manufacturing Industries Such As Mineral Slag, Fly Ash, Silica Fumes Etc. The Construction Industry Is Increasingly Seeking Sustainable Alternatives To Reduce The Carbon Footprint Of Concrete And Manage Agricultural Waste. This Research Investigates The Synergistic Effect Of Incorporating Arecanut Fibers Is A Natural Fibers Obtained From The Areca Palm Tree. It Acts As A Light Weight Composite Material And Silica Fume Industrial By Product Found To Be An Attractive Cementations Material Which Is Byproduct Of Smelting Process In The Silicon And Ferrosilicon Industry.. The Primary Objective Is To Evaluate The Mechanical Properties Of A Binary Blended Concrete Mix. In This Experimental Study, Silica Fume Was Used As A Partial Replacement For Cement At A Constant Rate Of 10% (by Weight), While Arecanut Fibers Were Added In Varying Volume Fractions Of 0%, 1%, 2%, 3% And 4%. A Control Mix Of M30 Grade Concrete Was Used For Comparative Analysis. The Performance Assessment Involved Testing For Fresh Properties (slump Test), Mechanical Strength (compressive, Flexural, And Split Tensile Strength) Of Concrete..

Author: A Kranthi Kumar | K Urmila Devi

Area of research: Human Resource Management

The зtudy Focuзeз On Compenзation And Welfare Meaзureз And Their Impact On Employee зatiзfaction At Sri Sai Tubeз And Pipeз. Employee зatiзfaction Playз A Vital Role In Improving Productivity, Organizational Commitment, And Overall Performance. The Reзearch Aimз To Analyze Employeeз’ Opinionз Regarding зalary, Incentiveз, Welfare Facilitieз, зafety Meaзureз, Medical Benefitз, And Working Conditionз Provided By The Company. Primary Data Were Collected From Employeeз Through A зtructured Queзtionnaire, And зtatiзtical Toolз зuch Aз Chi-зquare Teзt, Correlation, And ANOVA Were Uзed For Analyзiз. The findingз Reveal That Effective Compenзation And Welfare Meaзureз Poзitively Influence Employee зatiзfaction And Motivation. The зtudy Concludeз With зuggeзtionз To Improve Welfare Policieз, Incentiveз, And Employee Engagement Practiceз To Enhance Organizational Effectiveneзз And Employee Morale.

Author: S.Naveena | Dr.S.Senthilkumar | Dr.S.Prakash

Area of research: Human Resource

This Study Explores Employee Job Satisfaction At Kovai Maruthi Papers And Boards Private Limited, Namakkal, Examining How It Affects Employee Performance, Productivity, Motivation, And Overall Organizational Growth. The Research Aims To Evaluate Employees' Satisfaction Levels Concerning Salary, Working Environment, Workload, Working Hours, Job Security, Supervisor Support, Promotion Opportunities, And Training Programs. Primary Data Were Gathered From A Sample Size Of 100 Employees Using Structured Questionnaires Via A Simple Random Sampling Technique. Secondary Data Were Sourced From Academic Books, Journals, Corporate Records, And Websites. Statistical Tools, Including Percentage Analysis, Chi-Square Analysis, And Correlation Analysis, Were Executed For Data Interpretation And Hypothesis Testing.

Author: C. Pavithra | Dr. S. Suganya | Dr. R. Florence Bharathi

Area of research: Blockchain & AI

GreenLand Is A Secure And Intelligent Land Registration System Designed To Improve Transparency, Security, And Efficiency In Agricultural Land Management Using Blockchain And Artificial Intelligence Technologies. Traditional Land Registration Systems Rely On Centralized Databases And Manual Verification Processes, Which Are Vulnerable To Document Forgery, Unauthorized Ownership Transfers, Duplicate Registrations, And Data Tampering. The Proposed System Integrates Blockchain Technology, Smart Contracts, Artificial Intelligence Models, And Decentralized Document Storage To Provide A Reliable And Tamper-proof Land Registration Framework. AI Algorithms Including Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), Extreme Gradient Boosting (XGB), And Light Gradient Boosting Machine (LGBM) Classify Land Records As Genuine Or Fraudulent Before Storage. Verified Records Are Stored Using The InterPlanetary File System (IPFS), While Only Hash Values Are Maintained On-chain To Improve Storage Efficiency And Data Integrity. Smart Contracts Automate Verification, Approval, And Ownership Transfer, Minimizing Human Intervention. Experimental Results Confirm The System’s Effectiveness In Fraud Detection, Transaction Security, And Scalability.

Author: Kaushik H V | Ms. Dharani A

Area of research: Human Resource Management

Recruitment And Selection Are Essential Functions Of Human Resource Management That Help Organizations Attract And Appoint Qualified Employees For Achieving Organizational Objectives. This Study Focuses On The Effectiveness Of The Recruitment And Selection Process Practiced By Believe HRD Consultancy Services. The Study Is Based On Primary Data Collected From 120 Respondents Through A Structured Questionnaire Using The Survey Method. Various Statistical Tools Such As Chi-square Test, Correlation Analysis, ANOVA, And Regression Analysis Were Used For Data Analysis And Interpretation. The Study Identifies The Strengths And Limitations Of The Existing Recruitment And Selection Practices And Offers Suggestions To Improve The Efficiency Of The Hiring Process And Employee Satisfaction.

Author: B.Kaaviya | R. Florance Bharathi | A. Aravinth

Area of research: MBA

Occupational Health And Safety Practices Play A Vital Role In Ensuring Employee Safety, Improving Productivity, And Maintaining A Healthy Work Environment In Industrial Organizations. This Study Focuses On Analyzing The Occupational Health And Safety Practices Followed At Sri Sai Tubes And Pipes. The Main Objective Of The Study Is To Examine Employee Awareness Regarding Workplace Safety Measures And Evaluate The Effectiveness Of The Safety Practices Implemented By The Organization. The Study Is Based On Both Primary And Secondary Data. Primary Data Was Collected Through Structured Questionnaires From Employees, While Secondary Data Was Collected From Books, Journals, Websites, And Company Records. Statistical Tools Such As Chi-Square Test, Correlation Analysis, And ANOVA Were Used For Analyzing The Collected Data. The Findings Of The Study Reveal That Effective Safety Measures Improve Employee Satisfaction, Reduce Workplace Accidents, And Enhance Employee Productivity. The Study Also Suggests That Regular Safety Training Programs And Proper Monitoring Systems Can Further Strengthen Workplace Safety Practices Within The Organization.

Author: M.Meena | S. Senthilkumar | S. Prakash

Area of research: MBA(HR)

This Study Examines Employee Stress Management At Current Electro Mech Pvt. Ltd., Erode, With The Objective Of Identifying The Factors Causing Workplace Stress And Evaluating The Effectiveness Of Stress Management Practices Followed By The Organization. The Study Adopted A Descriptive Research Design And Quantitative Research Approach. Primary Data Was Collected From 91 Employees Using A Structured Questionnaire, While Secondary Data Was Collected Through Company Records, Journals, And Research Articles. Statistical Tools Such As Percentage Analysis, Independent Sample T-test, One-way ANOVA, And Linear Regression Were Used For Data Analysis. The Findings Reveal That Gender Does Not Significantly Influence Most Stress-related Variables, While Work Experience Has A Partial Impact On Employee Perceptions. The Regression Analysis Indicates That Working Hours Do Not Significantly Predict Employee Stress Levels. The Study Concludes That Organizational Support, Workload Management, And Work-life Balance Are Essential Factors In Reducing Employee Stress And Improving Employee Well-being.

Author: S R Tharanekhaa | Mr. A. Aravinth | Mrs. R. Kokila

Area of research: Human Resource

Human Resource Consultancies Play An Important Role In Connecting Job Seekers With Employment Opportunities. In Recent Years, HR Consultancy Firms Have Faced Several Challenges In Attracting And Retaining Quality Candidates Due To Changing Candidate Expectations, Trust Issues, Communication Gaps, And Fake Job Concerns. The Present Study Titled “A Study On Challenges Faced By HR Consultancies Based On Candidate Perception With Reference To Covaijobs HR Service, Coimbatore” Was Conducted To Identify The Major Challenges Faced By HR Consultancy Firms And To Understand Candidate Perception Towards Consultancy Services. The Study Is Descriptive In Nature And Primary Data Were Collected From 50 Respondents Through A Structured Questionnaire. Secondary Data Were Collected From Books, Journals, And Websites Related To Human Resource Management And Recruitment Practices. Statistical Tools Such As Percentage Analysis, Chi-Square Analysis, And Correlation Analysis Were Used Through SPSS Software For Data Interpretation. The Findings Of The Study Reveal That Lack Of Trust, Fake Job Advertisements, Poor Communication, Inadequate Follow-up, And Lack Of Transparency Are The Major Issues Affecting Candidate Perception Towards HR Consultancy Firms. The Chi-Square Analysis Indicates A Significant Relationship Between Consultancy Experience And Candidate Trust. The Correlation Analysis Shows That Communication Satisfaction Has A Strong Positive Relationship With Overall Candidate Satisfaction. The Study Concludes That HR Consultancy Firms Should Focus On Ethical Recruitment Practices, Effective Communication, Transparency, Timely Follow-up, And Suitable Job Matching To Improve Candidate Satisfaction And Strengthen Trust Among Job Seekers.

Author: P. Thashmithaa | Dr. S. Prakash

Area of research: Internet Of Things

This Paper Presents An IoT And Deep Learning-based Smart Fire Prevention And Suppression System Designed To Improve Fire Safety And Emergency Response Efficiency. Traditional Fire Detection Systems Rely Mainly On Smoke Detectors And Manual Monitoring, Which Often Result In Delayed Detection And Increased Damage. The Proposed System Integrates IoT Sensors, Deep Learning Algorithms, And Cloud Connectivity For Real-time Fire Monitoring And Automated Suppression. Sensors Such As Smoke, Temperature, And Gas Sensors Continuously Monitor Environmental Conditions And Transmit Data To A Cloud Platform Through A NodeMCU Microcontroller. Deep Learning Models Analyze Sensor And Image Data To Accurately Identify Fire Incidents And Reduce False Alarms. When Abnormal Conditions Are Detected, Alerts Are Sent To Users And Suppression Mechanisms Are Activated Automatically. The System Improves Response Time, Enhances Safety, Minimizes Property Damage, And Supports Smart Building And Smart City Applications.

Author: Muhamed.J | M SANTHOSH KUMAR | B. AJAY VIGNESH | H. MOHAMED SHAFIQ

Area of research: HUMAN RESOURCES

Workplace Stress Has Become One Of The Most Significant Challenges Affecting Employee Well-being And Organizational Performance In Today's Competitive Business Environment. Employees Are Increasingly Exposed To Demanding Workloads, Strict Deadlines, Extended Working Hours, Communication Barriers, And Work-life Balance Challenges. These Factors Contribute To Occupational Stress, Which Can Adversely Influence Employee Productivity, Motivation, Efficiency, And Overall Job Performance. The Present Study Examines The Impact Of Workplace Stress On Employee Performance At EEE Infra Equipments Pvt. Ltd. The Research Adopted A Descriptive Research Design And Collected Data From 100 Employees Through A Structured Questionnaire. Statistical Tools Such As Percentage Analysis, Mean Analysis, Chi-Square Analysis, Correlation Analysis, And ANOVA Were Used To Analyze Employee Responses. The Findings Indicate That Workload Pressure, Workplace Conflicts, Communication Issues, Tight Deadlines, And Long Working Hours Are Major Contributors To Employee Stress. The Study Further Reveals That Increased Stress Levels Negatively Affect Employee Motivation, Concentration, Work Quality, And Productivity. The Research Highlights The Importance Of Stress Management Programs, Employee Wellness Initiatives, Supportive Leadership Practices, And Effective Communication Systems In Enhancing Organizational Performance. The Study Concludes That Organizations Must Prioritize Employee Well-being And Implement Proactive Stress Management Strategies To Achieve Sustainable Productivity And Long-term Organizational Success.

Author: Ms. U.Rithika | Dr. S.Suganya | Dr. R.Florancebharathi

Area of research: Computer Science And Business Systems

The Chemical Tracker System Is A Web-based Application Developed To Monitor And Track Precursor Chemicals From Production To End Use. The System Integrates Blockchain, Internet Of Things (IoT), And Modern Web Technologies To Ensure Transparency, Security, And Accountability In The Chemical Supply Chain. The Application Enables Authorities Such As The Narcotics Control Bureau (NCB) To Monitor Chemical Movement In Real Time, Detect Tampering During Transportation, And Prevent Illegal Diversion Of Chemicals. GPS Tracking And Tamper Detection Sensors Help Provide Continuous Monitoring Of Shipments. The System Also Includes Role-based Dashboards For Administrators, Zonal Officers, Manufacturers, Transporters, And End Users. Features Such As Secure Login, Order Creation, Shipment Tracking, Alert Generation, And Report Analysis Improve Operational Efficiency And Reduce Manual Effort. The Proposed Solution Provides A Scalable And Secure Platform For Modern Chemical Supply Chain Management.

Author: Boopathy Eswaran B | Santhosh Saravanan G | Yogesh S | Santhosh Kumar M

Area of research: Web Development

In The Current Digital Era, College Students Frequently Miss Internships, Hackathons, Job Openings, And Coding Contests Because Relevant Information Is Fragmented Across Dozens Of Platforms, Social Media Groups, And Institutional Portals. This Paper Presents The Design And Development Of An Event Aggregator For College Students — A Centralized, Rolebased Web Application That Aggregates, Verifies, And Delivers Academic And Professional Opportunities Through Three Dedicated Portals: The Student Portal, The Company Portal, And The Admin Portal. The Student Portal Enables Profile Creation, Skillbased Filtering, Application Tracking, And Personalized Deadline Reminders. The Company Portal Allows Verified Organizations To Post And Manage Opportunities And Evaluate Applicants. The Admin Portal Ensures Platform Integrity Through Opportunity Moderation And Company Verification. The System Is Built Using React.js For The Frontend, Java Spring Boot Or Node.js For The Backend, And MySQL/MongoDB As The Database. The Agile SDLC Model Guides Iterative Development Across Sprints. This Paper Details The System’s Motivation, Related Work, Software Requirements, Architecture, Data Flow, Risk Management, And Future Directions Including AI-based Recommendations And Automated Resume Screening.

Author: Prathamesh Shinde | Omkar Phapale | Saurabh Kanade | Sahil Zaware

Area of research: English

Early In His Career, Browning Began Writing Love Poems, And He Kept Doing So Until His Death. The Poet Communicates A Powerful, Sensual, Earthy, And Spiritual Love For The Woman In "A Pearl" And "By The Fireside," Which Opens Up An Endless Realm Of Love For The Lover. Browning Writes On Window Panes, Gloves, And Garden Walls Instead Of Challenges, Ideals, Or Broad Generalizations—things And Locations Connected To The Beloved Or To A Passionate Moment. Browning's Love Poems Don't Discuss Love Of The Truth, Love Of People, Or Love Of One's Country. His Love Is Merely A Passion That Attracts Men And Women Alike. Browning Doesn't Discuss The Beauty Of Women In His Love Poetry. There Is A Small Amount Of A Woman's Physical Allure. He Focuses On The Influence That A Woman Might Have Over A Guy In Their Relationship. Love Therefore Is Not A Goal Unto Itself. It Is A Technique Of Obtaining Heavenly Happiness. The Diversity Of Love Scenarios That Browning Tackled Was Extremely Broad, Making Him The Only English Poet To Have Dealt With Love In All Its Myriad Intricacies. Browning Did Not Hold Back When Describing Love, Which Was Frowned Upon By Tradition. He Also Writes Poetry About Unusual Couples In Love. The Foundation Of Browning's Success In Both His Personal And Literary Areas Rests On His Love, Which He Used As A Springboard To Success And Ultimately As A Mile-stone. The Numerous Stages And Types Of Love In All Societal Classes Are Covered In Browning's Love Poems. This Essay Aims To Explore Browning's Poetry's Treatment Of Love. It Also Sheds Light On How He Approaches Romantic Love, Both Physical And Spiritual, As Well As His Realism, The Strength Of Love, And How He Handles Unusual Or Atypical Love Situations.

Author: Dr.Shital Vipulkumar Chandak

Area of research: Commerce

NON-BANKING FINANCIAL INSTITUTIONS (NBFIs) Play A Pivotal Role In India’s Financial Ecosystem By Complementing Traditional Banks In Delivering Credit To Under Banked And Underserved Segments. This Study Undertakes A Comprehensive Study On The Financial Performance And Growth Of Select Prominent NBFIs In India Over The Period From 2020 To 2025. Indian Economy Undergoing Rapid Changes Due To The Pandemic, Policy Reforms, Digitization, And Evolving Market Dynamics, The Role Of NBFIs Has Gained Renewed Significance. This Study Aiming On Analyzing The Annual Returns And Associated Risks (measured Through Standard Deviation) Of Key NBFIs To Evaluate Their Profitability, Stability, And Growth Trends In A Changing Economic Landscape. The Institutions Analyzed Include Industry Leaders Such As Bajaj Finance, Muthoot Finance, Shriram Finance, And Others That Have Demonstrated Notable Financial Activity During The Study Period. These NBFIs Cater To Various Sectors, Including Consumer Finance, Gold Loans, Vehicle Loans, Microfinance, And Housing Finance, Reflecting The Diversity Of Services In The NBFC Domain. The Selection Of Institutions Was Based On Their Market Presence, Financial Transparency, And Relevance In The Indian Financial Market. The Study Applies Quantitative Tools To Examine Annual Return Data And Calculate Standard Deviation, Enabling A Comparative View Of Performance And Risk Across Different NBFIs. By Analyzing Yearly Trends, The Study Highlights How Each Company Responded To Challenges Such As The COVID-19 Pandemic, Liquidity Crises, Changing Regulatory Norms From The Reserve Bank Of India (RBI), And Digital Transformation Pressures. The Findings Show That While Some NBFIs Demonstrated Robust Financial Resilience And Continued Growth, Others Experienced Fluctuations And Increased Exposure To Risk. For Instance, Bajaj Finance Exhibited Strong Returns With Moderate Risk Levels, While Institutions Like Muthoot Finance Showcased Stability Due To Their Niche Gold Loan Market Positioning. Beyond Numerical Analysis, The Study Also Explores Qualitative Aspects, Such As The Regulatory Framework Governing NBFIs, The Impact Of RBI Guidelines On Credit Policies, And The Strategic Measures Adopted By Institutions To Remain Competitive. It Discusses How The Sector Is Becoming Increasingly Regulated, Ensuring Greater Transparency And Better Risk Management But Also Demanding Higher Compliance And Adaptability From The Institutions Involved. This Study Aims To Offer Valuable Insights Into The Structural And Financial Transformation Of NBFIs And Their Contribution To Economic Growth, Especially In Rural And Semi-urban Areas. It Also Seeks To Understand How These Institutions Can Balance Growth With Sustainability, Especially In The Face Of Economic Disruptions. By Doing So, The Study Aspires To Guide Future Financial Decisions, Policymaking, And Academic Inquiry Into India’s Evolving Financial Services Sector. In Conclusion, The Analysis Underlines The Critical Role That NBFCs Continue To Play In Democratizing Access To Credit In India. Despite Facing Regulatory Tightening And Market Competition, Their Ability To Innovate, Localize Financial Solutions, And Expand Outreach Makes Them An Indispensable Component Of India’s Financial Future. This Study Not Only Evaluates Past Performance But Also Provides A Lens Through Which To View Future Opportunities And Challenges For The NBFI Sector.

Author: Dr.V.Senthilkumar | Rekha R | Nimya M P

Area of research: Marketing

Brand Image Greatly Affects Customer Buying Behavior And Helps Companies Build Trust And Loyalty In The Dairy Industry. This Study Focuses On The Brand Image Of Vijay Dairy And Farm Products (P) Ltd, Aiming To Identify Factors That Influence Customer Preference And Satisfaction. Key Areas Of Focus Include Consumer Awareness, Product Quality, Price, Packaging, Hygiene, Promotional Strategies, And Overall Satisfaction With Vijay Dairy Products. We Used A Descriptive Research Design For This Study. Primary And Secondary Data Were Gathered From 120 Respondents Through A Structured Questionnaire And Convenience Sampling. Statistical Analysis Tools Such As Simple Percentage Analysis, Chi-Square Analysis, Correlation, And ANOVA Were Employed To Evaluate The Data. The Findings Show That Most Consumers View Vijay Dairy Products Positively, Especially Regarding Quality, Hygiene, Attractive Packaging, And Reasonable Pricing. TV And Promotional Campaigns Played A Significant Role In Raising Brand Awareness Among Consumers. The Study Concludes That Maintaining Quality Standards And Boosting Promotional Efforts Can Further Enhance Customer Satisfaction, Brand Loyalty, And The Overall Image Of Vijay Dairy And Farm Products (P) Ltd.

Author: M. Kalaivani | Dr. S. Senthil Kumar | Dr. S. Prakash

Area of research: HUMAN RESOURCE

This Study Examines The Level Of Employee Job Satisfaction At Rich Dairy Products (India) Private Limited, Namakkal. Employee Job Satisfaction Is A Critical Determinant Of Organizational Success, Affecting Productivity, Employee Retention, Morale, And Overall Workplace Effectiveness. The Research Aims To Evaluate The Satisfaction Levels Of Employees With Respect To Working Conditions, Compensation, Supervision, Interpersonal Relations, Career Growth, And Welfare Measures. A Descriptive Research Design Was Adopted, And Primary Data Were Collected From 110 Employees Using Structured Questionnaires. Secondary Data Were Obtained From Journals, Company Records, Books, And Digital Sources. Statistical Tools Including Percentage Analysis And Chi-square Test Were Employed For Data Analysis. The Findings Reveal That A Majority Of Employees Expressed Moderate-to-high Levels Of Satisfaction, While Certain Areas Such As Salary And Promotional Opportunities Indicated Scope For Improvement. The Study Concludes With Actionable Recommendations For Management To Foster A More Engaged, Productive, And Satisfied Workforce.

Author: Ms. S. Sushma Harshini | Mr. T. Krishna kumar | Mr. A. Aravinth

Area of research: BUSINESS ANALYTICS

Accurate Sales And Demand Forecasting Is A Strategic Imperative For Textile Manufacturing Enterprises Operating In Volatile, Seasonally Driven Markets. This Study Applies Time Series Analysis To Historical Sales And Demand Data (2022–2025) Of Ascent Textiles, A Home Textile Manufacturer Based In Karur, Tamil Nadu, To Evaluate Performance Patterns And Generate Forward-looking Projections For 2026. The Analysis Examines Trend, Seasonality, Cyclical Movement, And Irregular Variation Across Product Categories And Geographic Markets. Findings Reveal A Sustained Upward Sales Trajectory — From ₹9.8 Million In 2022 To ₹16.7 Million In 2025 — With Peak Demand Concentrated In The Final Quarter Of The Year. Domestic Markets Account For Approximately 61.78% Of Total Revenue, While Export Markets, Led By The USA And UK, Contribute The Remaining 38.22%. Tablecloths And Cushion Covers Emerge As The Highest-revenue Product Segments. Power BI-based Forecasting Confirms Continued Growth Through 2026, Tempered By Evidence Of Periodic Overproduction. The Study Demonstrates The Practical Value Of Data-driven Forecasting In Improving Production Planning, Inventory Management, And Strategic Decision-making In Textile SMEs.

Author: Poorani S | Dr.Suganya | Dr.R. Florence Bharathi

Area of research: Machine Learning & LabVIEW

Institutional Contexts Have Long Relied On Traditional Attendance Systems That Involve Taking Attendance, Using Roll Call, Paper Sign-in Sheets, And Manual Record Keeping, Which Have All Been Seen As Inefficient. They Are Labor Intensive, Subject To Human Error And Can Be Subject To Proxy Attendance Which Adds An Unnecessary Strain On Administrators And Undermines The Integrity Of Attendance Information. This Paper Presents A Real-time, Contactless Face Recognition Attendance Tracking System Based On The LabVIEW Platform That Can Overcome These Issues By Implementing Intelligent Automation With Computer Vision And Image Processing. The System's Core Is A Common Webcam Which Is Constantly Taking Snapshots From The Face Surrounding The User. The Images Are Then Processed Through A Rigidly Designed Pre-processing Chain, Which Starts With The Conversion Of The Images To Gray Scale And Then To A Set Of Images That Have Been Normalized Using A Histogram. This Is To Compensate For Various Lighting Conditions, Followed By Filtering To Removes Image Noise, And Finally A Region-of-interest Extraction That Selects The Most Relevant Areas Of The Face For Further Processing. This Pre-processing Pipeline Is Implemented In The LabVIEW Vision Framework. This Pre-processing Is What Gives The System The Accuracy It Needs Despite Less Than Optimal Conditions. After Images Have Been Prepared, The LabVIEW Vision Development Module Assumes Control, Running The Algorithms Necessary For Face Detection, Feature Extraction And Template Matching Against A User Database Of Registered Faces. The System Automatically Records Attendance When A Face Is Successfully Recognized, Adds A Timestamp, And Provides A Confidence Score, All Of Which Are Shown In Real Time On LabVIEW's Interactive Graphical Interface. The Experimental Evaluation Was Done Under Lab Conditions, And The Results Were Positive; 98.2% Face Detection Accuracy, 96.9% Recognition Accuracy And 97.4% Positive Result For Attendance Logging. Most Importantly, These Figures Were Maintained Under Different Lighting Conditions And Moderate Changes In Head Orientation, Which Are Typical Situations For Which Traditional Recognition Systems Struggle. A Confidence Threshold Mechanism Provides An Extra Degree Of Reliability, Which Decreases The Chances Of False-positive Reports And Guarantees That Attendance Is Accurate. The Value Of This System Is That All Components Of The System, From Image Acquisition To Preprocessing, Image Recognition, Decision-making, And Logging, Are Integrated In A Single Cohesive Real-time Platform. No Need For Post Session Check Or Manual Correction. Administrators Always Have Visibility Of The Identification Status, Attendance Logs And System Performance Statistics On A User-friendly Front Panel, Which Allows Them To Stay Informed And Stay In Control. The Outcome Is A Scalable, Cost-effective, And Easy-to-solve Solution That Could Be Adopted In Smart Schools, Universities, And Organizations Where Accurate And Automated Attendance Tracking Is Not Just A Better Practice, It's A Must.

Author: Shanmugam M | Magesh Kumar R | Nimal Dinesh M | Nitheesh Kumar R | Nithish kumar T

Area of research: Computer Application

The Rapid Proliferation Of Digital Content Creation Tools, Artificial Intelligence Platforms, And Advanced Image Editing Software Has Significantly Increased The Risk Of Digital Forgery In Images, Documents, And AI-generated Media. Traditional Forensic Methods Are Limited To Single-domain Analysis And Lack Integration, Centralized Evidence Management, And Automated Reporting. This Paper Presents, A Hybrid Digital Forgery Detection Framework Designed As A Unified, Full-stack Digital Forensic Intelligence Platform. The System Integrates Three Specialized Forensic Detection Modules: (i) Copy-Move Image Forgery Detection Using The Scale-Invariant Feature Transform (SIFT) Algorithm With FLANN-based Matching, (ii) Document Forgery Detection Using OCR-based Text Consistency Analysis And Structural Validation, And (iii) AI-Edited Image Detection Using Error Level Analysis (ELA), Noise Residual Analysis, And Compression Artifact Examination — All Within A Single Centralized Dashboard. The Platform Is Developed Using Next.js/TypeScript For The Frontend, FastAPI/Python For The Backend, And SQLite For Database Management. Experimental Evaluation Confirms Successful Execution Of Multi-category Forensic Analysis, Dashboard History Tracking, And Structured Forensic Report Generation With Tampered Region Localization. The Modular Architecture Supports Future Extensions Including Deepfake Video Detection, Blockchain-based Evidence Preservation, And Cloud-based Deployment.

Author: Sarathkumar L | Ms. Dharani A

Area of research: Healthcare Informatics

The Cloud-Based Diabetes Management System Using Machine Learning Is A Smart Healthcare Application That Helps Monitor And Manage Diabetes Efficiently. It Allows Patients To Register Online, Upload Health Data, Book Appointments, And View Reports Through A Cloud Platform. Machine Learning Algorithms Analyze Factors Like Glucose Level, Blood Pressure, Insulin, BMI, And Age To Predict Diabetes Accurately. The System Reduces Manual Work, Improves Healthcare Services, Supports Remote Monitoring, And Provides Secure And Reliable Patient Management..

Author: Dr.S.Saravanakumar | Gowtham.R | Hariharan.S | Praveen.P

Area of research: Civil Engineering

Infrastructure Development For A Country Is A Principle Development And Concrete Plays A Vital Role. Concrete Is The World’s Largest Consuming Material In The Field Of Construction. From Time Immemorial Research Over Concrete Has Been Going On To Enhance Its Performance And Strength. Nowadays, Most Concrete Mixture Contains Supplementary Cementious Material (SCM) Which Forms Part Of The Cementitious Component. These Materials Are Majority By-products From Other Processes. The Main Benefits Of SCMs Are Their Ability To Replace Certain Amount Of Cement And Still Able To Display Cementitious Property, Thus Reducing The Cost Of Using Portland Cement. The Fast Growth In Industrialisation Has Resulted In Tons And Tons Of By-product Or Waste Materials, Which Can Be Used As SCMs Such As Silica Fume, By Adding Steel Fibers And Carbon Fibers Etc. The Use Of These By-products Not Only Helps To Utilize These Waste Materials But Also Enhances The Properties Of Concrete In Fresh And Hydrated States. This Study Focuses On The Experimental Investigation Of Partial Replacement Of Cement With Ground Granulated Blast Furnace Slag (GGBS) In Hybrid Fibre Reinforced Concrete (HFRC) Using Steel And Carbon Fibres. The Main Objective Is To Evaluate The Mechanical And Durability Properties Of Concrete With Varying Percentages Of GGBS. Cement Is Partially Replaced With GGBS At Different Proportions Such As 0%, 20%, 40%, And 60%. Hybrid Fibres Are Added To Improve Tensile Strength, Crack Resistance, And Ductility. Steel Fibres Enhance Load-carrying Capacity, While Carbon Fibres Help In Controlling Micro-cracks. The Experimental Program Includes Tests On Compressive Strength, Split Tensile Strength, And Flexural Strength. The Results Indicate That GGBS Improves Long-term Strength And Durability, While Hybrid Fibres Significantly Enhance Mechanical Performance. The Study Concludes That GGBS-based Hybrid Fibre Reinforced Concrete Is A Sustainable And High-performance Construction Material.

Author: Mr Ashish Nitin Pawar

Area of research: Civil Engineering

Infrastructure Development For A Country Is A Principle Development And Concrete Plays A Vital Role. Concrete Is The World’s Largest Consuming Material In The Field Of Construction. From Time Immemorial Research Over Concrete Has Been Going On To Enhance Its Performance And Strength. Nowadays, Most Concrete Mixture Contains Supplementary Cementious Material (SCM) Which Forms Part Of The Cementitious Component. These Materials Are Majority By-products From Other Processes. The Main Benefits Of SCMs Are Their Ability To Replace Certain Amount Of Cement And Still Able To Display Cementitious Property, Thus Reducing The Cost Of Using Portland Cement. The Fast Growth In Industrialisation Has Resulted In Tons And Tons Of By-product Or Waste Materials, Which Can Be Used As SCMs Such As Silica Fume, By Adding Steel Fibers And Carbon Fibers Etc. The Use Of These By-products Not Only Helps To Utilize These Waste Materials But Also Enhances The Properties Of Concrete In Fresh And Hydrated States. This Study Focuses On The Experimental Investigation Of Partial Replacement Of Cement With Ground Granulated Blast Furnace Slag (GGBS) In Hybrid Fibre Reinforced Concrete (HFRC) Using Steel And Carbon Fibres. The Main Objective Is To Evaluate The Mechanical And Durability Properties Of Concrete With Varying Percentages Of GGBS. Cement Is Partially Replaced With GGBS At Different Proportions Such As 0%, 20%, 40%, And 60%. Hybrid Fibres Are Added To Improve Tensile Strength, Crack Resistance, And Ductility. Steel Fibres Enhance Load-carrying Capacity, While Carbon Fibres Help In Controlling Micro-cracks. The Experimental Program Includes Tests On Compressive Strength, Split Tensile Strength, And Flexural Strength. The Results Indicate That GGBS Improves Long-term Strength And Durability, While Hybrid Fibres Significantly Enhance Mechanical Performance. The Study Concludes That GGBS-based Hybrid Fibre Reinforced Concrete Is A Sustainable And High-performance Construction Material.

Author: Ms. Ahire Shweta Pradeep | Ms. AHIRE SHWETA PRADEEP

Area of research: Human Resource

Emotional Intelligence (EI) Has Emerged As A Critical Determinant Of Employee Performance And Organizational Effectiveness In The Contemporary Software Industry. This Study Investigates The Role Of Emotional Intelligence And Its Measurable Impact On Employee Performance At Rivvot Technologies, Coimbatore. A Structured Questionnaire-based Survey Was Conducted Among 150 Employees Across Various Job Roles Including Developers, Testers, HR, And Managers. Data Were Analyzed Using Simple Percentage Analysis, Chi-square Test, Pearson Correlation, And One-way ANOVA. Findings Reveal That A Significant Majority Of Employees Recognize The Positive Contribution Of EI To Job Performance, Stress Management, Interpersonal Communication, And Organizational Success. The ANOVA Results Confirm That Work Experience Significantly Influences Participation In EI Development Activities (F = 11.83, P < .001). The Study Underscores The Need For Organizations In The Software Sector To Systematically Invest In EI Training Programs To Foster A High-performance Work Culture.

Author: G. Sasireka | Mr.T. Krishna Kumar | Dr. R. Miyalvaganan

Area of research: Computer Science And Engineering

AI-TPEP (AI Trust & Performance Evaluation Platform) Is A Modular And Reproducible Framework For Comprehensive Evaluation Of Machine Learning Systems Across Key Trust Dimensions, Including Predictive Performance, Calibration, Fairness, Robustness, Explainability, Safety, And Privacy. The Platform Ingests Model Artifacts Or API Endpoints And Conducts Deterministic Benchmark And Stress Testing—such As Adversarial Attacks, Distribution Shifts, And Subgroup Analyses—within A Sandboxed Execution Environment That Captures Detailed Telemetry. Standardized Metrics Are Computed Using Explicitly Defined Normalization Transforms And Aggregated Into Configurable Subscores And A Composite AI Trust Score. AI-TPEP Also Generates Signed Evidence Packages Containing Raw Inputs And Outputs, Explanation Artifacts, Manifests, And Provenance Records To Support Independent Auditing. This Paper Presents The System Architecture, Formal Metric Formulations, And Parameterized Test Catalogs For NLP/LLM And Vision Models, Alongside A Reference Implementation With CI/CD Integration. Experimental Results Demonstrate Practical Trade-offs Among Trust Dimensions, And Deployment Guidance Is Provided To Help Organizations Make Transparent, Defensible, And Data-driven Model Governance Decisions.

Author: Niranchana V | Harikrishnan R | Sarveshwaran M | Suriyabharathi J

Area of research: Human Resource

The Effectiveness Of Training And Development (T&D) Programmes In Augmenting Employee Productivity Is A Central Concern In Contemporary Human Resource Management, Particularly Within The High-velocity Manufacturing Sector. This Study Empirically Investigates The Impact Of Structured Training Interventions—including On-the-job Training (OJT), Induction Programmes, Skill Enhancement Workshops, And SOP-based Instruction—on The Productivity Outcomes Of Employees At Padget Electronics Pvt Ltd (a Subsidiary Of Dixon Technologies), Oragadam, Tamil Nadu. Using A Structured Questionnaire Administered To 100 Respondents And Analysed Through Simple Random Sampling, The Study Employs Percentage Analysis, Likert Scale Rating, Chi-square Tests, And Pearson Correlation Analysis To Measure Training Effectiveness Across Five Key Performance Dimensions: SOP Compliance, Defect Identification, Machine Handling, Work Speed Improvement, And Training Relevance. Findings Reveal Significant Enhancements In SOP Adherence, Defect Rates, And Machine Handling Competencies Among Shop-floor Employees. A Strong Positive Correlation Was Identified Between Training Frequency And Employee Productivity Metrics. The Study Concludes With Strategic HR Recommendations Underscoring The Practical Relevance Of Need-based, Periodic Training In Driving Industrial Productivity And Workforce Capability.

Author: Ms. S. Rasika | Dr. S. Prakash | Dr. S. Senthil Kumar

Area of research: HUMAN RESOURCES MANAGEMENT

This Study Examines Employee Welfare Measures And Their Impact On Employee Satisfaction And Organizational Effectiveness At Rich Dairy Products (India) Private Limited, Namakkal. Employee Welfare Is An Important Component Of Human Resource Management That Contributes To Employee Motivation, Productivity, And Job Satisfaction. The Main Purpose Of The Study Is To Evaluate The Existing Welfare Facilities Provided By The Organization And Analyze Employees' Perceptions Regarding These Facilities. A Descriptive Research Design Was Adopted For The Study. Primary Data Were Collected From 110 Employees Through Structured Questionnaires, While Secondary Data Were Collected From Journals, Websites, Books, And Company Records. Statistical Tools Such As Percentage Analysis, Chi-square Analysis, Correlation, And ANOVA Were Used For Data Analysis. The Findings Indicate That Welfare Facilities Significantly Influence Employee Motivation, Productivity, Work-life Balance, And Retention. The Study Concludes That Improving Welfare Facilities Enhances Employee Satisfaction And Contributes Positively To Organizational Growth.

Author: Ms.C.V.Sathiyapriya | Mr. T. Krishna kumar | Ms.M.Rishivarthini

Area of research: HUMAN RESOURCES MANAGEMENT

Quality Of Work Life Is Referred To All The Organizational Inputs Which Aim At The Employee’s Satisfaction And Enhancing Organization Effectiveness. The Purpose Is To Develop Jobs And Working Condition That Is Excellent For Employees As Well As The Economic Health Of The Organization. It Also Refers To The Satisfaction, Motivation, Commitment, And Involvement Of An Individual Experience Concerning Their Line At Work. The Paper Aims To Study The Concept “Quality Of Work-Life” And The Role It Plays In Enhancing The Productivity And Performance In The Firm. The Purpose Of The Study Is Mainly To Understand The Quality Of Work Life Of The Employees With Significant Factors Like Working Environment, Training, And Development, Compensation & Rewards, Organizational Commitment, Job Satisfaction, Etc. The Research Includes 120 Employees Who Were Designated As Staff Employee, Technician, Executive And Manager In A Firm. The Primary Data Can Be Analysed Using The Statistical Tool Like ANOVA, Chi-Square, And Correlation

Author: G.Priyadharshini | Dr.S.Suganya | Dr. R.Florence Bharathi

Area of research: MBA

This Study Examines The Effectiveness Of Training And Development (T&D) Practices At Rich Dairy Products (India) Private Limited, Namakkal. The Study Adopts A Descriptive Research Design, Utilizing Primary Data Collected Through A Structured Questionnaire Administered To 110 Employees Via Convenience Sampling. Statistical Tools Including Percentage Analysis, Chi-square Test, Pearson Correlation, And ANOVA Were Applied. Findings Indicate That A Significant Majority Of Employees Are Satisfied With Learning Opportunities And Believe Training Programs Effectively Improve Job-related Skills And Career Growth. The Study Concludes With Strategic Recommendations To Strengthen T&D Practices For Enhanced Organizational Productivity.

Author: Ms. A.S. Agalya | Mr. T. Krishna Kumar | Dr. R. Miyal Vaganan

Area of research: HUMAN RESOURCES

This Study Examines The Impact Of Employee Involvement On Organizational Effectiveness In Chola Spinning Mills Private Limited, Erode. Employee Involvement Plays A Vital Role In Improving Productivity, Decision-making, Employee Commitment, And Organizational Performance. The Research Aims To Analyze Employee Engagement Practices, Identify Employee Satisfaction Levels, And Examine The Relationship Between Employee Involvement And Organizational Effectiveness. A Descriptive Research Design Was Adopted For The Study. Primary Data Were Collected From 150 Employees Through Structured Questionnaires, And Secondary Data Were Collected From Journals, Websites, And Company Records. Statistical Tools Such As Percentage Analysis, Chi-square Analysis, Correlation, And ANOVA Were Used For Data Analysis. The Findings Reveal That Employee Involvement Significantly Influences Productivity, Teamwork, Communication, And Organizational Performance. The Study Concludes That Organizations Should Encourage Participative Management, Effective Communication, Recognition Systems, And Employee Development Programs To Improve Organizational Effectiveness.

Author: Ms.K.Sathiyapriya | Mr. T. Krishna kumar | Ms.M.Rishivarthini

Area of research: Human Resourse

Employee Engagement Plays A Vital Role In Improving Organizational Performance And Employee Productivity. This Study Aims To Analyse The Level Of Employee Engagement And Identify The Factors Influencing Employee Commitment, Motivation, And Job Satisfaction Within The Organization. The Study Examines Various Aspects Such As Communication, Leadership Support, Recognition, Teamwork, And Work Environment. Primary Data Were Collected Through Questionnaires From Employees, And Statistical Tools Were Used For Analysis And Interpretation. The Findings Reveal That Effective Engagement Practices Positively Influence Employee Morale, Performance, Retention, And Organizational Growth. The Study Also Provides Suggestions To Improve Engagement Strategies And Create A Supportive And Productive Work Place Environment.

Author: P.Parkavi | Dr. S. Suganya | Dr. R. Florence Bharathi

Area of research: Computer Science And Engineering

The Used Vehicle Market Has Experi- Enced Significant Growth In Recent Years, Making It Challenging For Buyers And Sellers To Determine Accurate Vehicle Prices. Traditional Pricing Meth- Ods Often Rely On Personal Judgment, Which May Lead To Inconsistencies And Unfair Transactions [12]. In This Work, A Smart Web-based System Is Developed To Predict The Price Of Used Vehicles Using Machine Learning Techniques. The System Employs A Random Forest Regression Model To Analyze Key Features Such As Mileage, Brand, And Engine Capacity To Estimate A Fair Resale Value [5]. In Addition To Price Prediction, The Platform Provides Features Such As Vehicle Recommenda- Tions, Price Trend Analysis, And Location-based Search, Enhancing User Experience And Decision- Making [13]. Experimental Results Demonstrate That The Proposed Model Achieves High Accuracy And Is Suitable For Real-world Applications [14].

Author: Yash Kumawat | Om Panhale | Vishal Patil | Prajwal Shinde

Area of research: Civil Engineering

Building Information Modeling (BIM), Cloud Computing, Artificial Intelligence (AI), And The Internet Of Things (IoT) Are Some Of The Technologies That Are Driving The Construction Industry's Rapid Digital Transformation. Virtual Project Management (VPM), Which Enables Data-driven Decision-making, Real-time Communication, And Remote Coordination, Is Made Possible By These Technologies. This Study Uses A Structured Methodology Based On PRISMA Standards To Give A Systematic Review Of VPM In Construction. To Assess Important Technologies, Performance Results, And Implementation Issues, 147 Peer-reviewed Studies Were Examined. The Results Show That VPM Increases Safety Results, Lowers Project Costs By 10–20%, And Improves Schedule Performance By 15–25%. Effective Implementation Is Nevertheless Hampered By Obstacles Like Cybersecurity Concerns, Resistance To Change, Interoperability Problems, And A Lack Of Technical Skills. The Paper Highlights Important Research Gaps, Such As The Requirement For Scalable Digital Twin Applications, Explainable AI Models, And Integrated Frameworks. In Addition To Providing A Thorough Overview Of VPM, The Report Suggests Future Research And Industry Adoption Directions.

Author: Shreya Vivek Bharsakle

Area of research: Civil Engineering

Construction Site Safety Is A Serious Concern Due To The Complex And Risky Nature Of Construction Activities. Traditional Safety Monitoring Methods, Such As Manual Supervision And CCTV Systems, Are Commonly Used But Have Several Limitations, Including Dependence On Human Observation, Inconsistency, And Lack Of Real-time Response. With Recent Advancements In Computer Vision And Deep Learning, Automated Safety Monitoring Systems Have Started Gaining Attention. These Systems Can Detect Safety Violations, Such As Absence Of Personal Protective Equipment (PPE) Or Unsafe Worker Behavior, Directly From Images And Video Data In Real Time. This Paper Presents A Review Of Existing Research Related To Construction Safety Monitoring. It Covers Traditional Safety Practices, Computer Vision-based Approaches, And Deep Learning Models Such As YOLO[4], R-CNN[9], And Transformer-based Methods[10]. Different Studies Are Analyzed And Compared Based On Their Methodology, Performance, And Practical Use In Real-world Conditions. The Review Also Identifies Key Research Gaps, Including Limited Real-world Implementation, Lack Of Integration With Construction Management Systems, And Challenges Caused By Environmental Conditions Like Lighting And Occlusion. Finally, The Paper Discusses Possible Future Directions, Such As The Use Of Hybrid Systems, Integration With IoT Devices, And Development Of More Efficient Real-time Monitoring Solutions

Author: Sarvesh Rajendra Holey

Area of research: Human Resource

This Study Focuses On Work Stress Management Among Employees Of Dhanlaxmi Bank. In The Modern Banking Environment, Work Stress Has Become A Common Issue Due To Heavy Workload, Long Working Hours, Target Pressure, Customer Expectations, And Technological Changes. These Factors Affect The Physical And Mental Well-being Of Employees And Influence Their Job Performance And Satisfaction. The Main Objective Of The Study Is To Identify The Causes Of Work Stress, Examine Its Impact On Employee Health And Work Performance, And Evaluate The Stress Management Practices Followed In The Organization. The Study Adopts A Quantitative Research Design. Primary Data Were Collected From 100 Employees Using A Structured Questionnaire. The Findings Reveal That Workload, Target Pressure, And Customer Expectations Are The Major Sources Of Stress Among Employees. The Study Also Found That Effective Stress Management Practices, Supportive Work Environment, And Organizational Initiatives Help In Reducing Employee Stress And Improving Overall Well-being And Productivity.

Author: M.Samyuktha | Dr. S.Suganya | Dr R.Florence Bharathi

Area of research: IT

This Paper Presents The BuySense, A Machine Learning Application That Predicts Whether A Product Is Worth Purchasing Based On Amazon Review Text Data. The System Builds Two Independent Binary Classifiers—a Viability Model And A Regret Model—trained On The Amazon Review Polarity Dataset. Both Models Use A Text Vectorization Layer Combined With An Embedding And GlobalAveragePooling1D Architecture. The Trained Models Are Deployed In A Streamlit Web Application That Scrapes Amazon Product Pages In Real Time, Extracting Titles, Descriptions, And Structured Content To Generate Buy, Wait, Or Avoid Recommendations. This Dual-model Design Enables Nuanced Purchase Guidance Beyond Simple Positive/negative Polarity Classification.

Author: Shivaathmajan P | Ayswaryaa V | Gautham Siddarth | Nithya Roopa S

Area of research: MACHINE LEARNING

This Paper Presents A UPI Fraud Detection System Using Machine Learning Techniques To Identify And Prevent Fraudulent Digital Payment Transactions. With The Rapid Growth Of Unified Payments Interface (UPI) Transactions In India, Cyber Fraud And Unauthorized Payment Activities Have Increased Significantly. Traditional Fraud Detection Methods Often Fail To Detect Sophisticated Fraud Patterns In Real Time. The Proposed System Uses Machine Learning Algorithms To Analyze Transaction Behavior, Identify Suspicious Activities, And Classify Fraudulent Transactions Effectively. The System Improves Transaction Security, Reduces Financial Losses, And Enhances User Trust In Digital Payment Systems.

Author: Dr. S.Saravana Kumar | NaveenN, Kabilan T | Sarveshwar S

Area of research: MBA

Working Capital Management (WCM) Plays A Significant Role In Maintaining The Liquidity, Profitability, And Operational Efficiency Of An Organization. This Research Paper Analyzes The Working Capital Management Practices Of Femtosoft Technologies, Chennai, A Software Development And IT Solutions Company Specializing In Logistics And ERP Solutions. The Study Evaluates The Company’s Liquidity Position, Profitability, Solvency, And Operational Efficiency Using Financial Ratio Analysis, Working Capital Turnover Analysis, And Trend Analysis Over The Period From 2019–2020 To 2023–2024. The Findings Indicate That The Company Maintained A Strong Liquidity Position In Most Years, With Effective Utilization Of Working Capital Resources Despite Temporary Fluctuations During 2020–2021. The Study Concludes That Efficient Working Capital Management Has Positively Influenced The Company’s Financial Stability And Operational Performance. Recommendations Are Also Provided To Further Improve Receivables Management, Liquidity Planning, And Profitability.

Author: Neelakulali Senthil | Dr.S.Prakash | Dr.S. Senthil Kumar

Area of research: Artificial Intelligence

Student Health And Physical Fitness Are Increasingly Neglected Due To Academic Workload, Irregular Schedules, And Lack Of Personalized Guidance. This Paper Presents An AI- Powered Personalized Fitness Planning System Tailored Specifically For Students, Integrating Machine Learning, Deep Learning, And Computer Vision To Generate Adaptive Workout Recommendations. The Proposed System Collects Student-specific Parameters Such As Body Mass Index (BMI), Fitness Goals, Academic Schedule, Activity History, And Dietary Preferences To Build A Comprehensive User Profile. A Hybrid Model Combining A 1D-Convolutional Neural Network (1D-CNN) With A Gradient-boosted Classifier Generates Individualized Exercise Prescriptions, While A Real-time Pose Estimation Module Using MediaPipe Provides Form Feedback During Workouts. Experimental Evaluation On A Dataset Of 500 Undergraduate Students Demonstrates That The Proposed System Achieves 93.6% Accuracy In Fitness Level Classification And Yields Measurable Improvements In Student Physical Activity Adherence Over An Eight-week Trial Period. The System Represents A Practical, Low-cost Solution Deployable On Standard Smartphones, Making Personalized Fitness Coaching Accessible To The Student Popula- Tion Without Requiring Expensive Gym Memberships Or Personal Trainers.

Author: Dr.S.R.Patil | Yadav Karan Balasaheb | Sanket Bharat Gore | Junnare Prathamesh Suhas | Sapkal Vishwajeet Hanmant

Area of research: STRUCTURAL ENGINEERING

Ordinary Portland Cement (OPC) Is The Most Heavily Consumed Man-made Material On Earth. However, This Massive Scale Comes With A Severe Environmental Penalty: The 1:1 Carbon Ratio: As A Rule Of Thumb In Material Science, Manufacturing 1 Ton Of OPC Releases Approximately 0.8 To 1.0 Ton Of Carbon Dioxide Directly Into The Atmosphere. The Cement Industry Alone Is Responsible For Roughly 7% To 8% Of All Global Carbon Dioxide Emissions And Also Cement Manufacturing Is A Highly Energy-intensive Process, Requiring Temperatures Up To 1450°C Inside The Rotary Kiln. This Works Presents A Comprehensive Experimental Investigation Into The Mechanical Characterization And Acid-resistance Kinetics Of Sustainable Binary Blended Concrete Incorporating Agro-waste Rice Husk Ash (RHA) As A Partial Replacement For Ordinary Portland Cement (OPC). To Evaluate The Structural And Durability Performance, Cement Was Substituted With RHA At Varying Dosages Ranging From 0% To 25% At Increments Of 5%. M35 Grade Concrete Mixes Were Designed In Accordance With IS 10262 Guidelines, Keeping A Constant Water-binder Ratio. Mechanical Evaluation Was Conducted Via Compressive, Split-tensile, And Flexural Strength Tests At 7 And 28 And 56 Days Of Curing. The Core Focus Of This Study Centers On The Durability Kinetics Of The Blended Matrix When Subjected To Aggressive Chemical Environments. Concrete Specimens Were Fully Immersed In A 5% Hydrochloric Acid (HCl) Solution For Exposure Durations Of 7, 28 And 60 Days.

Author: M Maneesha | B Ganesh

Area of research: Computer Cyber Security

The Relentless Escalation Of Cyber Threats In Contemporary Network Environments Has Necessitated A Fundamental Rethinking Of Intrusion Detection Methodologies. Traditional Signature-based Systems, While Once Sufficient, Have Proven Structurally Incapable Of Addressing Polymorphic Attacks, Zero-day Exploits, And The Sheer Volume Of Anomalous Traffic Generated By Modern Enterprise Networks. This Paper Presents A Real-Time AI Intrusion Detection System (RT-AI-IDS) Developed Using A Hybrid Machine Learning Framework That Combines Unsupervised Anomaly Detection With Supervised Attack Classification. The Proposed System Employs The Isolation Forest Algorithm For Identifying Statistical Outlier Indicative Of Previously Unseen Attack Behaviour And A Random Forest Classifier For Precise Categorisation Of Known Attack Classes Using The Benchmark NSL-KDD Dataset. The System Is Architected As A Full-stack Web Application, With A FastAPI -powered Backend Exposing RESTful Endpoints For Real-time Prediction, Alert Generation, And Intrusion Logging, Coupled With A React And Vite-based Monitoring Dashboard Providing Live Visualisation Of Traffic Patterns And Detection Events. Persistent Storage Of Prediction Histories, Alert Records, And Intrusion Logs Is Managed Through A Lightweight SQLite Database, Enabling Retrospective Analysis And Operational Transparency. Experimental Evaluation Demonstrates High Classification Accuracy, Low False-positive Rates, And Real-time Latency Suitable For Production Deployment. The Architecture Is Explicitly Designed For Modularity And Horizontal Scalability, Allowing The System To Be Extended With Additional Detection Models, Data Sources, And Alerting Channels Without Architectural Disruption. The Proposed System Represents A Meaningful Contribution To The Practical Deployment Of AI -based Intrusion Detection In Resource-constrained And Enterprise-grade Environments Alike.

Author: Silambarasan B | Sarveswaran S | Sharan S.P | Pradeep N | Ravindra krishna chandar V

Area of research: Human Resources

Training And Development Play An Important Role In Improving Employee Performance, Productivity, And Organizational Growth In Modern Industries. This Study Focuses On Analyzing The Training And Development Practices Followed At Sundram Fasteners Limited, Chennai. The Main Objective Of The Study Is To Evaluate The Effectiveness Of Training Programs And Measure Employee Satisfaction Towards Training Initiatives. Primary Data Was Collected Through A Structured Questionnaire From 71 Employees Working In Different Departments. Statistical Tools Such As Percentage Analysis, Correlation Analysis, And Chi-square Test Were Used For Interpretation. The Findings Reveal That Most Employees Attended Training Programs And Considered Training Useful For Improving Efficiency, Confidence, And Work Performance. The Study Concludes That Effective Training Practices Contribute Significantly To Employee Development And Organizational Success.

Author: J.Thirisana | Mr. Aravinth | Mrs. R. Kokila

Area of research: Deep Learning-based Forest Fire Detection Using Explainable AI

This Study Proposes FireNet-CNN, A Lightweight And Efficient Deep Learning Model For Real-time Forest Fire Detection Using Convolutional Neural Networks (CNN) And Explainable AI (XAI) Techniques. The Model Was Trained And Evaluated On Augmented Datasets Containing Fire And Non-fire Images, Achieving High Performance With 99.05% Accuracy, 99.41% Precision, And 98.28% Recall. Stable Diffusion-based Synthetic Image Generation And Traditional Augmentation Methods Were Used To Improve Dataset Diversity And Reduce Class Imbalance. To Enhance Transparency And Reliability, Grad-CAM And Saliency Map Techniques Were Integrated To Visualize The Model’s Decision-making Process By Highlighting Fire-related Regions In Images. With A Compact Model Size And Fast Inference Time, FireNet-CNN Is Suitable For Deployment In Real-time Wildfire Monitoring Systems, Drones, And Embedded Devices For Early Forest Fire Detection And Disaster Management..

Author: Mrs.B.Sathya | Harihasudhanks | Karthik Ramanathan Sr | Madhushudanan V

Area of research: HUMAN RESOURCES MANAGEMENT

Author: R.Menaja | Dr.S. Prakash | Dr.S.Senthil Kumar

Area of research: MANAGEMENT/HUMAN RESOURCES

Employee Welfare And Job Satisfaction Are Essential Components Of Organizational Success And Employee Development In Modern Industries. Employee Welfare Includes All The Services, Facilities, And Benefits Provided By Employers To Improve The Physical, Mental, Social, And Economic Well-being Of Employees. The Present Study Aims To Analyse The Employee Welfare Measures And Their Impact On Job Satisfaction Among Employees Working At KMB Granites Pvt Ltd.The Study Focuses On Various Welfare Measures Such As Safety Facilities, Working Environment, Loan Facilities, Health Insurance, Working Conditions, Water Facilities, And Employee Relations. Welfare Measures Help Employees Feel Secure And Motivated, Which In Turn Improves Productivity And Organizational Efficiency. The Research Was Conducted Using Both Primary And Secondary Data. Primary Data Were Collected Through Questionnaires From 120 Employees Selected Through Convenience Sampling. Secondary Data Were Collected From Books, Journals, Company Reports, And Websites.Statistical Tools Such As Percentage Analysis, Correlation, Chi-square Test, And ANOVA Were Used To Analyse Employee Opinions Regarding Welfare Measures And Job Satisfaction. The Findings Of The Study Reveal That The Majority Of Employees Are Satisfied With Safety Measures, Supervision, Work Environment, And Welfare Facilities Provided By The Company. However, Employees Expressed Dissatisfaction In Certain Areas Such As Loan Facilities And Work Pressure.The Study Concludes That Employee Welfare Measures Have A Significant Positive Impact On Job Satisfaction And Employee Productivity. Organizations That Provide Effective Welfare Programs Can Improve Employee Morale, Reduce Absenteeism, And Maintain Better Industrial Relations. Therefore, Companies Should Continuously Improve Welfare Facilities To Achieve Organizational Growth And Employee Satisfaction.

Author: S. Poornisha | Dr . S.Suganya | Dr .R.Florance bharathi

Area of research: HUMAN RESOURCE MANAGEMENT

This Study Examines The Effectiveness Of Recruitment And Selection Processes In Organizations. The Research Focuses On Employee Perceptions Regarding Recruitment Transparency, Job Evaluation, Induction Process, And Training Programs. A Descriptive Research Design Was Adopted, And Primary Data Were Collected From 120 Respondents Through Structured Questionnaires. Statistical Tools Such As Percentage Analysis, Chi-square Analysis, Correlation, And ANOVA Were Used For Interpretation. The Findings Reveal That Most Respondents Are Satisfied With The Recruitment Process, While Some Employees Perceive Biasness In Recruitment Practices. The Study Concludes That Organizations Should Strengthen Transparency, Improve Communication, And Adopt Fair Recruitment Practices For Better Employee Satisfaction.

Author: M.Mythily | Mr.A.Aravinth | Dr .R. Florence Bharath

Area of research: Computer Science

The Internet Of Things Framework Has Seen A Rapid Change In Terms Of The Applications And Users Worldwide. However, The Need For Trustworthiness To Satisfactory Quality Of Service Is Of Utmost Importance Keeping In Mind The Nature Of Data Transfer In Wireless Media. The Advent Of High Compute Power Processors With Miniature Sizes And Low Power Consumption, Implementing Relatively Complex Algorithms Has Become Possible Which Is Necessity For Internet Of Things Applications. This Research Paper Focusses On The Design And Implementation Of The Code Blocks Of Turbo Codes Based On The BCJR Algorithm So As To Couple The Bits In The Code Blocks In The Composite Transport Block. The Information And Parity Bits Are To Be Coupled So As To Have More Information Sharing Within The Transport Block And Hence Reduce The Error Rate Steeply In Section Of The Error Waterfall. The Proposed Technique Attains Lower Bit Error Rate Performance Compared To The Conventional Un-coded And Hard Coded Counterparts. A Comparative Analysis With Respect To The Error Rate Has Been Done So As To Evaluate The Quality Of Service Of The Proposed Work. The Lower Error Rate Of The Proposed Work Ensures The High Quality Of Service And Trustworthiness Of The IoT System.

Author: Divya Chouhan | Prof. Kiran Ajmera

Area of research: AI

The Online Chatbot For Student Information System Is An Intelligent Web-based Application Designed To Provide Quick And Automated Responses To Student Queries Related To Academic And Institutional Information. Traditional Student Support Systems Often Require Manual Interaction, Which Can Be Time-consuming And Inefficient. The Proposed Chatbot System Uses Artificial Intelligence And Natural Language Processing (NLP) Techniques To Understand User Queries And Provide Accurate Responses In Real Time.The System Allows Students To Access Information Such As Course Details, Attendance, Examination Schedules, Results, Fee Details, And Academic Notifications Through An Interactive Chat Interface. It Also Includes Secure Login, User Management, Query Handling, And Admin Modules For Updating And Maintaining Information.

Author: S. Santhosh kumar | Aarthy R | Angel mahima J | Rubika P | Vaishnavi J

Area of research: Management-Finance

Financial Analysis Plays An Important Role In Evaluating The Financial Strength, Operational Efficiency, Liquidity, And Profitability Of An Organization. The Present Study Focuses On The Financial Analysis Of HDB Financial Services, Erode. The Study Aims To Analyze The Financial Performance Of The Company Over A Period Of Five Years From 2020-2021 To 2024-2025. Secondary Data Collected From Annual Reports, Journals, Company Records, And Financial Statements Were Used For The Study. Various Financial Tools Such As Ratio Analysis, Trend Analysis, And Comparative Analysis Were Applied To Evaluate The Company’s Liquidity Position, Profitability, Solvency, And Efficiency. The Study Reveals That The Current Ratio And Current Asset Ratio Showed An Increasing Trend During The Study Period, Indicating Improvement In Short-term Financial Strength. However, Profitability Ratios Such As Return On Assets And Return On Equity Remained Negative, Which Reflects Challenges In Generating Adequate Profits. The Debt Ratio Indicates High Dependency On Borrowed Funds, While Asset Turnover Ratios Remained Stable. The Study Concludes That The Company Possesses Strong Market Presence And Operational Growth But Needs To Improve Profitability And Reduce Financial Risk Through Better Cost Control And Efficient Utilization Of Resources. Suggestions Are Also Provided To Enhance The Financial Performance And Long-term Sustainability Of The Company

Author: D.Dharshini | Dr. S.Prakash | Dr.S. Senthil Kumar

Area of research: MBA

Financial Performance Analysis Plays An Important Role In Evaluating The Financial Strength And Operational Efficiency Of An Organization. The Present Study Focuses On Analyzing The Overall Financial Performance Of HDB Financial Services, One Of The Leading Non-Banking Financial Companies (NBFCs) In India. The Study Aims To Evaluate The Liquidity, Profitability, Solvency, And Operational Efficiency Of The Company Using Various Financial Analysis Tools. Secondary Data Has Been Collected From Annual Reports And Financial Statements For The Period From 2020–2021 To 2024–2025. Ratio Analysis And Trend Analysis Were Used To Interpret The Financial Position Of The Organization. The Study Identified Improvements In Liquidity Position And Asset Utilization, While Profitability Remained Comparatively Low During The Study Period. The Debt-equity Ratio Showed A Decreasing Trend, Indicating Better Financial Stability. The Research Also Highlights The Importance Of Proper Financial Planning And Management For Achieving Long-term Growth. Finally, Suitable Suggestions Have Been Provided To Improve Profitability, Liquidity Management, And Operational Efficiency.

Author: K. Ragavi | Dr. S. Prakash | Dr. S. Senthil Kumar

Area of research: Computer Science

The Potential For Data Collecting And Analytics-based Intelligent Traffic Systems (ITS) To Improve Traffic Systems Is Now Being Investigation. Predicting How Fast Traffic Will Be Moving Is One Of The Most Important Uses For This Technology. In Order To Improve Traffic Management And Maximize The Overall Efficiency Of Urban Mobility, Intelligent Transportation Systems Rely On Machine Learning For Traffic Speed Forecasts. Several Data Sources Are Used In This Procedure To Forecast Traffic Speeds, Including Past Trends, Data From Sensors In Real-time, And Environmental Conditions. The Ability Of Machine Learning Models To Sift Through Mountains Of Data In Search Of Hidden Patterns And Correlations And Provide Reliable Forecasts Makes Them Indispensable In This Field. A Neural Network Model For Traffic Speed Predictions Based On Particle Swarm Optimization (PSO) Is Presented In This Research. By Using The PSO, The Network Weights May Be Adaptively Updated, Which Is Different From Traditional Neural Network Models. Both The MAPE And The Forecasting Accuracy Metrics Show That The Suggested Method Is Superior Than The Current Baseline Methods.

Author: Jyoti Gawatiya | Prof. Kiran Ajmera

Area of research: MBA

The Hybrid Work Environment Has Become A Key Practice For Organizations After The Rapid Adoption Of Digital Technologies And Flexible Work Arrangements. This Study Aims To Examine The Impact Of Hybrid Work Environments On Employees' Work-life Balance At Accent Techno Soft In Coimbatore. The Research Looks At Employees' Views On Communication Technology, Organizational Support, Flexibility, Emotional Well-being, And Family Balance. The Study Uses Descriptive And Analytical Methods With A Structured Questionnaire Sent To 150 Employees. Statistical Tools Like Percentage Analysis, Chi-square Test, Correlation, And ANOVA Were Used For Data Analysis. The Findings Show That Hybrid Work Arrangements Improve Flexibility, Family Involvement, And Employee Comfort, While Challenges Persist In Technological Support And Work-related Stress. The Study Suggests That Organizations Should Strengthen Employee Support Systems, Provide Better Technological Infrastructure, And Maintain Effective Communication To Enhance Work-life Balance In Hybrid Settings.

Author: R.Manisha | Dr.S.SenthilKumar | Dr.S.Prakash

Area of research: Artificial Intelligence And Machine Learning

The AI-Based Drug Interaction Detection System Is A Smart And User-friendly Healthcare Application Developed To Improve Medication Safety By Detecting Harmful Drug Interactions. The System Allows Patients, Doctors, And Guests To Enter Medicine Details Through Text, Prescription Scanning, Or Voice Input. Using Artificial Intelligence Techniques Such As Machine Learning (ML) And Natural Language Processing (NLP), The System Analyzes Medicines With The Help Of A Medical Database And Displays Results As Safe, Caution, Or Dangerous. It Also Provides Features Like Drug Information, Alert Notifications, History Tracking, And Personalized Dashboards. By Automating Drug Interaction Analysis, The System Reduces Manual Errors, Supports Informed Medical Decisions, And Offers An Efficient And Scalable Solution For Better Healthcare Management.

Author: Mr. V.Gokulakrishnan | Ayesha Farveen A | Deepika R | Deepika S

Area of research: Cyber Security

Secure Digital File Sharing Has Become A Fundamental Requirement For Individuals, Enterprises, And Regulated Industries In An Era Characterized By Pervasive Data Breaches And Expanding Regulatory Obligations. Traditional File-Sharing Solutions—Consumer Cloud Drives, Email Attachments, And Ftp Servers—Consistently Fail To Provide End-To-End Encryption, Granular Access Control, Cryptographic File Integrity Verification, And Tamper-Evident Audit Trails. This Review Paper Analyses The Academic And Standards Literature Underpinning Modern Secure File Sharing And Uses That Analysis To Contextualise The Design Of Phantomvox, A Proposed Full-Stack Web Application That Integrates Aes-256-Gcm Encryption At Rest, Tls 1.3 In Transit, Role-Based Access Control (Rbac), Multi-Factor Authentication Using Totp, Sha-256 Integrity Verification, And An Append-Only Hash-Chained Audit Log Into A Unified Platform. Six Literature Sources Spanning Symmetric Encryption, Transport Security, Authentication, Access Control, And Compliance Logging Are Critically Reviewed, And Five Persistent Research Gaps Are Identified: Absence Of Unified End-To-End Encryption In Accessible Tools, Weak Access Control Defaults, Insufficient Auditability, Lack Of File Integrity Enforcement, And Poor Usability Of Secure Sharing Workflows. The Phantomvox System Directly Addresses Each Gap Through Its Modular Architecture, Built On Node.Js, Express.Js, React, And Postgresql, And Is Demonstrated To Satisfy All Functional, Security, And Performance Requirements Through Comprehensive Testing.

Author: Hidhesh A | Kandha Prasanth S | Sindhura Selvam | Mrs. Sasikala

Area of research: Computer Science And Business Systems

Electricity Theft Is One Of The Major Challenges Faced By Power Distribution Systems, Especially In Developing Countries Like India. Traditional Electricity Monitoring Systems Mainly Depend On Manual Inspection And Basic Monitoring Techniques, Which Are Inefficient In Identifying Real-time Power Theft And Cyber-attacks. Smart Meters Are Widely Used For Monitoring Electricity Consumption, But The Communication Between Smart Meters And Electricity Providers Is Vulnerable To Data Tampering And Unauthorized Access. This Paper Proposes GridShield, A Secure Smart Meter Communication And Intelligent Energy Theft Detection System Using Encryption And Machine Learning Techniques. The Proposed System Encrypts Smart Meter Data Using AES Encryption Before Transmitting It To The Server, Thereby Ensuring Secure Communication. Machine Learning Algorithms Are Used To Analyse Electricity Consumption Patterns And Identify Abnormal Usage Behaviour That May Indicate Electricity Theft. The System Provides Real-time Alerts To Electricity Authorities When Suspicious Activity Is Detected. The Proposed Solution Improves Smart Grid Security, Reduces Power Loss, Minimizes Manual Monitoring, And Increases Theft Detection Accuracy. The System Is Scalable And Suitable For Smart City And Smart Grid Environments.

Author: Mrs. P. Uma Maheshwari | Swetha S | Riyashika R | Dhanupriya RD

Area of research: Computer Science And Business Systems

The Digital Memory Map System Is An AI-based Intelligent Object Tracking Application Designed To Help Users Locate Frequently Misplaced Personal Items Such As Mobile Phones, Wallets, Keys, Bottles, And Other Daily-use Objects. Traditional Methods Of Searching For Misplaced Objects Are Time-consuming And Inefficient. The Proposed System Uses Real-time Object Detection With YOLOv8 And OpenCV To Identify Objects Through A Webcam And Store Their Last Detected Location In A Database.The System Divides The Camera Frame Into Spatial Zones And Records Object Positions With Timestamps Using SQLite. Users Can Later Query The System Through A Flask-based Web Interface To Retrieve The Last Known Location Of An Object. The Proposed Solution Improves Object Management, Reduces Search Time, And Demonstrates The Practical Application Of Artificial Intelligence And Computer Vision In Smart Assistance Systems.

Author: Dinesh kumar C | Lukman Ahamed M | Sarukesh J

Area of research: Cyber Security

Phishing Attacks Continue To Represent One Of The Most Prevalent And Operationally Effective Cybersecurity Threats Targeting Everyday Internet Users. Adversaries Craft Malicious URLs That Closely Imitate Legitimate Websites With The Intent To Steal Credentials, Banking Details, And Sensitive Personal Data. This Paper Presents PhishShield — An Intelligent Phishing Detection Platform Developed Using Python And FastAPI. The System Subjects Any Submitted URL To A Multi-layer Detection Pipeline Comprising Heuristic Rule Evaluation, Shannon Entropy Analysis, Suspicious Top-level Domain (TLD) Identification, Deep Subdomain Inspection, Digit Randomisation Scoring, And Live Registration Data Access Protocol (RDAP) Domain-age Intelligence. All Indicators Are Aggregated Into A Single Weighted Risk Score That Classifies The URL As SAFE, SUSPICIOUS, Or PHISHING. Scan Results Are Persisted In A Local SQLite Database, And A Real-time Dark-themed Web Dashboard Enables Users To Submit URLs And Instantly Obtain A Verdict, Risk Meter Visualisation, And Historical Scan Data. Empirical Testing Across A Dataset Of 100 URLs — Comprising 50 Known-phishing Samples And 50 Legitimate URLs — Demonstrated 92% Accuracy On Phishing Samples And 96% Accuracy On Legitimate URLs, With An Average Scan Time Below One Second. The Platform Is Fully Modular, Lightweight, And Deployable On Any Standard Python Environment Without Requiring Expensive External APIs Or Cloud Subscription Services.

Author: Simpson R | SarathiKannan K | Sarveshwaran P | Varun S | Ravindra Krishna ChandarV

Area of research: Human Resource

Employee Onboarding Plays A Significant Role In Helping Newly Recruited Employees Adjust To Organizational Culture, Job Responsibilities, And Workplace Expectations. New Employees Often Experience Anxiety, Stress, Confusion, And Uncertainty During The Initial Stages Of Employment Due To Unfamiliar Work Environments And Unclear Role Expectations. This Study Examines The Impact Of The Onboarding Process On Reducing New Employee Anxiety At Femtosoft Technologies, Chennai. The Research Focuses On Key Onboarding Practices Such As Communication, Orientation And Training Programs, HR Support, Supervisor Guidance, Role Clarity, And Interaction With Team Members. Primary Data Were Collected From 100 Employees Using A Structured Questionnaire. Statistical Tools Such As Percentage Analysis, Correlation, And ANOVA Were Used To Analyze The Collected Data. The Findings Reveal That Structured Onboarding Programs Significantly Reduce Employee Anxiety By Improving Role Clarity, Confidence, Workplace Comfort, And Emotional Adjustment. The Study Concludes That Effective Onboarding Practices Contribute Positively To Employee Satisfaction, Adaptation, And Organizational Commitment. The Research Highlights The Importance Of Continuous HR Support And Employee-centered Onboarding Strategies For Improving Organizational Effectiveness And Reducing Employee Turnover.

Author: Ms. R. Roshini | Dr. S. Suganya | Dr. R. Florence Bharathi

Area of research: Machine Learning

Healthcare Is One Of The Most Important Research Fields With The Rapid Improvement Of Technology And Increase In Data. It Is Difficult To Handle Huge Amounts Of Patient Data. Big Data Analytics Can Be Used To Handle Such Data. There Are A Lot Of Procedures For The Treatment Of Multiple Diseases Across The World. Machine Learning Is A Prominent Approach That Helps In Prediction And Diagnosis Of A Disease.In The Existing System,diagnosis At An Early Stage Can Be Difficult As A Lot Of Diseases Might Have Very Common Symptoms And Require A Professional To Identify The Illness. Besides, A Lot Of Patients Delay Visiting A Doctor Because Of The Lack Of Information And Availability. Therefore, To Tackle This Problem, A System That Will Be Able To Give Some Hints About The Patient's Health Issues Is Developed. The Proposed System, “Enhancing The Recommendation Model For Disease Prediction Based On User Symptoms Using Machine Learning,” Presents An Effective Approach For Predicting Diseases And Providing User-centered Health Recommendations. By Integrating Multiple Machine Learning Algorithms Such As Random Forest, Support Vector Machine, Naïve Bayes, Decision Tree, And K-Nearest Neighbors Through An Ensemble Voting Model, The System Achieves Improved Prediction Accuracy, Robustness, And Reliability Compared To Traditional Single-model Approaches. The Implementation Of Pre-processing And Feature Engineering Techniques Enhances Data Quality And Optimizes Model Performance, Enabling The System To Handle Noisy And Diverse Symptom Data Effectively. In Addition To Disease Prediction, The Inclusion Of A Recommendation Module Provides Meaningful Precautions And Lifestyle Suggestions, Increasing The Practical Usefulness Of The System For End Users. The Developed Interface Ensures A User-friendly And Accessible Environment For Symptom Analysis And Prediction. Performance Evaluation Demonstrates That The Proposed Approach Produces Accurate And Consistent Results While Supporting Continuous Improvement Through Feedback Mechanisms.

Author: Sasikala M | Dr.Josephine Mary L

Area of research: Engineering

This Research Focuses On Accuracy Enhancement In The Detection Of Deepfakes Using The ResNet50 Algorithm Designed Through Deep Learning. It Analyzes Anomalies In Artificial Facial Images. Materials And Methods: The Two Implemented Deep Learning Models Include MobileNetV2 (Group 1) And ResNet50 (Group 2), Each Trained And Tested With 40 Image Samples, Comprising 20 Real Images And 20 Deepfake Images. Here, A Facial Irregularity Detector Based On ResNet50 Was Trained Against One Whose Model Was Created Through MobileNetV2. Result: ResNet50 Was Shown To Have A Detection Accuracy Of 91.81 % To 97.87 % For Distinguishing Between Real And Fake Photographs. Its Effectiveness For Real-time Applications Is Demonstrated. Statistical Study Revealed A Significant Improvement In Detection Accuracy Than The MobileNetV2 Model (p-value < 0.05). Conclusion: According To The Study's Results, The ResNet50 Algorithm Is Very Good At Identifying Deepfake Photos And Real Photos With A Low Mistake Rate And High Accuracy. Due To Its Efficiency In Processing Synthetic And Genuine Images, It Can Be A Dependable Tool For Handling The Problems Created By Deepfake Media.

Author: S.Pavithra, | R.Sridevi | Mahalakshmi .N | Devika R

Area of research: Engineering

This Research Aims To Develop A New Framework For The Early Identification Of Oral Cancer Through Multimodal Data Fusion With LSTM Networks And CNNs. The Performance Of The Proposed Framework Is Tested For Sensitivity, Specificity, And Accuracy In Detecting Oral Cancer And Has A High Operational Accuracy Of 93 %. This Research Uses Two Groups Of Data Sets.One Method Involves Using Clinical Image Data With 30 Samples Processed By The CNNs For The Extraction Of Spatial Features From Possible Malignancies.The Other Strategy Has Medical Image Data Analyzed With CNN- LSTM Networks Capturing Both The Temporal Dependencies And The Contextual Information With Samples Of 45. The Proposed Framework Shows High Sensitivity And Specificity In The Detection Of Oral Cancer, Which Identifies Early-stage Lesions With The Accuracy Of 93% And Subtle Abnormalities And Shows Significance Below That Of 0.05.The Existing Concept With An CNN Was Replaced With The Proposed Concept Of Enhanced Early And Accurate Identification Of Oral Cancer, Improving Patient Outcomes And Revolutionizing Healthcare Diagnostics.

Author: Dharani G | Shevak SriP

Area of research: Marketing

The Study Examines Customer Satisfaction Towards Milka Wonder Cake And Identifies The Level Of Satisfaction Among Consumers. The Research Is Based On A Descriptive Research Design And Uses Both Primary And Secondary Data. Primary Data Was Collected From 100 Respondents Through A Structured Questionnaire, While Secondary Data Was Gathered From Books, Journals, Websites, And Previous Research Studies Related To Customer Satisfaction. Statistical Tool Such As Percentage Analysis, Chi-square, Correlation, And ANOVA Was Used For Data Analysis. The Finding Sreveal That Most Respondents Are Satisfied With The Taste And Quality Of The Product, While Availability And Value For Money Also Influence Customer Satisfaction. This Study Provides Useful Suggestions To Improve Customer Satisfaction And Strengthen Market Performance.

Author: M. Santhiya | Mr. T. Krishna Kumar | Dr. R. Miyal Vaganan

Area of research: Cyber Security

Emergency Response Systems In Modern Urban And Disaster Environments Continue To Rely On Conventional Alert Mechanisms — Physical Buttons Or Active Network Calls — That Are Inaccessible To Individuals In Distress Due To Physical Incapacitation, Environmental Noise, Limited Network Connectivity, Or Psychological Shock. Traditional SOS Systems Require Deliberate Manual Interaction, Fundamentally Reducing Their Effectiveness In Real-world Crisis Scenarios. This Paper Presents PhantomVox, A Voice And Gesture-activated SOS Web Application Designed To Autonomously Trigger Emergency Alerts Based Entirely On Passive User Input Signals. The System Operates As A Browser-native Progressive Web Application (PWA), Eliminating The Need For Proprietary Hardware, Dedicated Installation, Or Platform-specific Dependencies. PhantomVox Utilises The Web Speech API For Continuous Distress Keyword Monitoring, The Device Motion API For Accelerometer-based Fall Detection, And The Geolocation API For Real-time GPS Coordinate Capture. Upon Trigger Detection, Structured SOS Alerts Are Dispatched To Up To Three Pre-registered Emergency Contacts Via Email Or SMS. Experimental Validation Confirms An Average Alert Latency Below 8 Seconds, A False-positive Rate Below 2%, And Voice Keyword Detection Accuracy Exceeding 92% In Quiet Environments. Through Comparative Analysis Of Existing Personal Safety Systems — Life Alert, BSafe, Apple Watch Fall Detection, And Voice Assistant SOS Features — Five Persistent Research Gaps Are Identified And Addressed Through PhantomVox's Design.

Author: Bradley Paulcat P | Dinesh M | Keerthivasan S | Saihitesh C | Saihitesh C | Mrs. K. Sivaselvi (M.E)

Area of research: ARTIFICAL INTELLIGENCE &DATA SCIENCE

Customer Churn Prediction Is An Important Research Problem In Business Analytics That Helps Organizations Identify Customers Who Are Likely To Discontinue Their Services. This Paper Proposes A Machine Learning-based Churn Prediction Framework Using Decision Tree, Random Forest, Logistic Regression, And XGBoost Algorithms. The Dataset Is Preprocessed Through Missing-value Handling, Categorical Encoding, Normalization, And Feature Selection Techniques To Improve Prediction Accuracy. Experimental Results Show That The XGBoost Model Provides Higher Performance Compared With Other Algorithms. The Proposed System Supports Organizations In Identifying High-risk Customers At An Early Stage And Enables Proactive Retention Strategies That Improve Customer Satisfaction And Organizational Profitability. This Paper Focuses On Customer Churn Prediction Using Artificial Intelligence And Machine Learning Techniques To Identify Customers Who Are Likely To Discontinue Services In Advance. The Proposed System Uses The Telco Customer Churn Dataset And Applies Machine Learning Algorithms Such As Decision Tree, Random Forest, Logistic Regression, And XGBoost To Analyze Customer Behavior Based On Attributes Like Tenure, Contract Type, Monthly Charges, And Payment Methods. Data Preprocessing Techniques Including Handling Missing Values, Encoding Categorical Variables, And Feature Selection Are Performed To Improve Prediction Accuracy. Among All The Algorithms, XGBoost Achieved The Highest Performance, Making It The Most Effective Model For Churn Prediction. The Developed System Helps Organizations Take Proactive Retention Strategies, Reduce Customer Loss, And Improve Business Decision-making Through Data-driven Insights.

Author: M.Sheeba | k.Srivalli | Y.Venkataprasanna | CH.Devaki

Area of research: Data Science

Objective: Thyroid Cancer Is A Growing Global Health Concern, With Early Detection And Accurate Diagnosis Playing A Pivotal Role In Improving Patient Outcomes. This Study Aims To Develop And Evaluate Machine Learning Models For Predicting Thyroid Cancer Malignancy Using Structured Clinical And Demographic Patient Data. Methods: We Utilized A Dataset Comprising 212,691 Anonymized Patient Records, Featuring Demographic Information (age, Gender), Clinical Indicators (family History Of Thyroid Disease), And Biochemical Markers (TSH, T3, T4 Levels). The Dataset Was Preprocessed To Address Missing Values, Encode Categorical Variables, And Standardize Numerical Features. Eight Machine Learning Algorithms—Logistic Regression, Random Forest, Gradient Boosting, AdaBoost, Decision Tree, Naive Bayes, XGBoost, And LightGBM—were Trained And Evaluated Using Accuracy, Precision, Recall, And F1-score. Results: The Top-performing Models—Logistic Regression, Gradient Boosting, And AdaBoost—achieved An Accuracy Of 82.5%, Demonstrating Strong Predictive Capability For Classifying Benign And Malignant Thyroid Cancer Cases. The Decision Tree Model Underperformed With An Accuracy Of 70.2%, Likely Due To Overfitting. Our Findings Were Contextualized With A 2024 Journal Of Medical And Health Sciences (JMHS) Study, Which Reported A 92.3% Accuracy For Predicting Thyroid Cancer Recurrence Using Logistic Regression, Underscoring The Potential Of Machine Learning In Clinical Settings. Clinical Implications: The Results Highlight The Utility Of Ensemble Machine Learning Models As Decision-support Tools For Clinicians, Facilitating Early Risk Assessment And Personalized Treatment Planning. Integration With Electronic Health Records (EHR) Could Further Streamline Diagnostic Workflows And Enhance Patient Care. Conclusion: This Study Validates The Effectiveness Of Machine Learning In Predicting Thyroid Cancer Malignancy, With Ensemble Models Showing Particular Promise. Future Research Will Focus On Hyperparameter Optimization, Deep Learning Techniques, And Real-world Clinical Deployment To Refine Accuracy And Practical Applicability.

Author: Musa Idris | Yusuf Ibrahim Yusuf

Area of research: Cyber Security

Cybersecurity Threats Have Grown Exponentially In Complexity And Volume, Rendering Traditional Rule-based Monitoring Tools Increasingly Inadequate For Modern Enterprise And Institutional Environments. Existing Security Platforms Are Often Siloed, Lack Intelligent Reasoning Capabilities, And Fail To Integrate Cross-domain Threat Intelligence With Automated Response Workflows. This Review Paper Surveys The Evolution Of AI-based Cybersecurity Monitoring Systems, Examining Machine Learning-based Intrusion Detection, Threat Intelligence Platforms, Dark Web Surveillance Systems, MITRE ATT&CK-aligned Detection Frameworks, And Large Language Model Applications In Security Operations. Through Systematic Examination Of Eight Significant Research Contributions, Five Persistent Research Gaps Are Identified: Absence Of Unified Multi-domain Threat Correlation, Exclusion Of AI-driven Red Team Simulation, Reliance On Static Signature-based Rule Databases, Lack Of Natural-language Explainability In Threat Analysis, And The Absence Of Integrated Zero-trust Policy Management. These Gaps Collectively Justify The Conceptual Design Of SecurAI Sentinel, A Proposed Full-stack AI-powered Cybersecurity Web Application Integrating Eight Intelligence Modules—CVE Intelligence Hub, Dark Web Monitor, MITRE ATT&CK Mapper, Incident Response Playbook Generator, Forensics Timeline Builder, Packet Capture Analyzer, AI Red Team Agent, And Zero Trust Policy Builder—within A Unified Glassmorphic Interface Powered By Google Gemini AI, React, Node.js, And Express.js. The Paper Concludes With A Discussion Of The System's Feasibility, Societal Impact, And Directions For Future Research.

Author: Arikaran P | Yogendra Sai P | Kalluri Nagalakshmi | Yeruva Sai Hanumantha Reddy | Sivaselvi K

Area of research: Human Resource

This Study Examines The Employee Health And Safety Measures Implemented At Current Electro Mech Private Limited. The Research Aims To Evaluate Existing Safety Practices, Employee Awareness, Workplace Hazards, And Areas For Improvement. Both Primary And Secondary Data Were Used For The Study. Primary Data Were Collected From 90 Employees Through A Structured Questionnaire Using Simple Random Sampling, While Secondary Data Were Obtained From Company Records And Relevant Literature. Statistical Tools Such As Percentage Analysis, Chi-square Test, Correlation, Weighted Average Method, And ANOVA Were Applied For Analysis. The Findings Indicate That Most Employees Are Aware Of Safety Policies, Regularly Use Personal Protective Equipment (PPE), And Are Satisfied With The Company’s Safety Measures And Training Programs. The Study Concludes That The Company Maintains Effective Health And Safety Practices; However, Continuous Monitoring, Regular Training, And Improved Safety Facilities Are Essential For Sustaining A Safe And Healthy Workplace.

Author: V.Vishvini | Dr. S. Senthil Kumar | Dr. S. Prakash

Area of research: Cyber Security

The Increasing Prevalence Of Spam And Scam Calls Poses A Significant Security Threat, Particularly In Multilingual Regions Such As India. This Paper Presents ShieldCall AI, An AI-driven Spam Call Detection System That Integrates Machine Learning, Natural Language Processing (NLP), And Real-time Audio Processing To Identify Fraudulent Calls Across Eight Major Indian Languages. The System Employs A Multinomial Naive Bayes Classifier With TF-IDF Feature Extraction To Classify Call Transcripts As Spam Or Legitimate. Live Audio Is Transcribed Via Google Speech Recognition, While Language Detection And Translation Modules Support Hindi, Tamil, Telugu, Kannada, Malayalam, Bengali, Marathi, And Gujarati By Translating Regional Language Transcripts To English Prior To Classification. The Backend Is Implemented Using Python And Flask, And The Frontend Is A Mobile-first Web Application Deployed On Render. A Distinctive Feature Is The Live AI Screener, An Autonomous Conversational Agent That Interacts With Suspected Spam Callers Using Voice Synthesis, Collects Evidence, And Dynamically Escalates Risk Scores. Testing Across All Supported Languages Achieved High Spam Detection Accuracy With Low False-positive Rates, Demonstrating The Practical Effectiveness Of The System For Real-world Personal And Organizational Security Applications

Author: Adithya Varshan A | Ashvanthgopal Baskaran | Ranjith V | Dr. V. Ravindra Krishna Chandar

Area of research: MARKETING

The Purpose Of This Study Is To Examine The Customer Buying Behaviour Towards Milka Wonder Cake And Identify The Key Factors Influencing Consumers’ Purchasing Decisions. The Research Investigates Demographic Influences, Brand Perception, Taste Preferences, Pricing, Promotional Strategies, And Product Availability. A Structured Questionnaire Was Administered To Customers Selected Through Purposive Sampling In Urban Retail Outlets. Data Were Analyzed Using Descriptive Statistics And Chi-square Tests To Determine Relationships Between Variables. The Findings Reveal That Taste Quality, Brand Loyalty, And Promotional Offers Are Significant Determinants Of Purchase Intention. Additionally, Price Sensitivity And Product Packaging Also Affect Consumer Choice. The Study Concludes That Effective Marketing Strategies And Improved Product Positioning Can Enhance Customer Satisfaction And Increase Market Share For Milka Wonder Cake. Recommendations For Marketers Include Focusing On Targeted Promotions, Maintaining Competitive Pricing, And Strengthening Distribution Channels To Better Meet Customer Needs.

Author: M. Rithika | Dr. S. Senthil Kumar | Dr. S. Prakash

Area of research: Cyber Security

Maintaining Visibility Over An Expanding Network Infrastructure Is One Of The Biggest Challenges Faced By Modern Cybersecurity Teams. Without Regular Auditing, Open Ports, Outdated Services, And Unpatched Operating Systems Become Easy Targets For Attackers. For This Mini-project, Our Team Focused On Automating Network Reconnaissance Using Nmap, A Powerful And Widely Used Open-source Security Tool. Running Nmap Commands Manually For Multiple Targets Is Tedious, Time-consuming, And Prone To Human Error. To Solve This, We Developed A Complete Python Automation Script That Runs Comprehensive Network Scans And Generates Structured Security Audit Reports With A Single Click. By Utilizing The 'python-nmap' Library, Our System Seamlessly Performs Host Discovery, TCP/UDP Port Scanning, Service Version Detection, OS Fingerprinting, And Automated Vulnerability Checking Using The Nmap Scripting Engine (NSE). Instead Of Presenting Raw, Complex Terminal Data, Our Script Intelligently Categorizes Open Ports Into SAFE, RISK, Or VULNERABLE Statuses And Outputs An Easy-to-read Text Report. Testing In A Controlled Virtual Environment Against Vulnerable Targets Like Metasploitable 2 Proved That Our Tool Accurately Identifies Active CVEs (Common Vulnerabilities And Exposures) Much Faster Than Manual Auditing, Making It Highly Effective For Routine Network Security Management.

Author: Harinarayanan R | Himanis Hariharha Kawsikan K S | Haswin T | Shiva Shankaran S | Dr. V. Ravindra Krishna Chandar

Area of research: Cyber Security

ARP (Address Resolution Protocol) Spoofing Is A Very Common And Dangerous Network Attack. In This Attack, A Hacker Sends Fake ARP Messages Over A Local Area Network To Link Their Own MAC Address With The IP Address Of A Legitimate Device, Usually The Default Gateway Or Router. Once This Is Done, All The Data Meant For The Internet Goes Through The Attacker's Computer First, Allowing Them To Steal Passwords And Read Private Messages. For Our Mini-project, We Developed A Tool Called ARPTM (ARP Spoofing Detection And Network Trust Monitor) Using Python. Our Main Goal Was To Create A System That Not Only Detects The Attack But Also Stops It Automatically. We Designed A Unique 'Trust Score' System Out Of 100 For Every Device On The Network. The Tool Uses The Scapy Library To Read Network Packets In Real-time. If A Device Acts Suspiciously, Like Changing Its MAC Address Or Sending Too Many Packets At Once, Its Score Drops. If The Score Goes Below 60, Our Python Script Automatically Runs Firewall Commands To Block The Attacker And Sends Out Healing Packets To Fix The Network. Testing Showed That Our Project Detects And Stops Attacks In Less Than A Second, Making It Very Effective For Securing Local Networks.

Author: Mohamed Thowfiq A | Jawahar P S | NasreynAbinash A | Dr.N. SundaraRajulu

Area of research: Cyber Security

Crypto-ransomware Remains One Of The Most Destructive Categories Of Malware, Exploiting Strong Symmetric Encryption To Render Victim Data Inaccessible Until A Ransom Is Paid. Differential Area Analysis (DAA), Introduced By Davies Et Al., Analyzes Shannon Entropy Variations Within File Headers To Discriminate Ransomware-encrypted Files From Compressed And Legitimately Encrypted Content. Despite Its Efficacy, DAA Is Susceptible To Adversarial Header Manipulation. This Paper Presents Three Novel Header-injection Attack Strategies—designated Attack-I, Attack-II, And Attack-III—that Exploit The Header-dependency Of DAA To Systematically Suppress Detectable Entropy Signatures. To Counteract These Evasion Vectors, We Propose Three Enhanced Countermeasure Techniques, Namely 2-Fragments (2F), 3-Fragments (3F), And 4-Fragments (4F), Which Partition File Headers Into Multiple Non-overlapping Segments And Compute Differential Entropy Across Each Fragment To Improve Detection Sensitivity. Machine Learning Classifiers, Including Logistic Regression (LR), Support Vector Machine (SVM), And XGBoost, Are Trained On Entropy-derived Feature Vectors Extracted Via The Proposed Fragmentation Schemes. Extensive Experiments On A Dataset Comprising Over 130,000 Files—including Real-world Ransomware Samples From WannaCry, Ryuk, Phobos, Sodinokibi, And NetWalker—demonstrate That Multi-fragment Analysis Substantially Improves Detection Robustness, Achieving F1-scores Exceeding 96% While Maintaining High Throughput In Files-per-second Benchmarks. The System Is Validated For Resilience Against Low-entropy Data Injection And Operates Effectively Under Adversarial Conditions Where Vanilla DAA Fails.

Author: Dr. Ravindra Krishna Chandar V | Kaliyamoorthi B | Shakthi aravinth M | Sekar C | Mohamed Ismail Anas M

Area of research: Data Science

Recommender Systems Often Reinforce Preference Homogeneity And Remain Vulnerable To Rating Manipulation. This Paper Proposes A Credibility-weighted Framework Integrating Review Usefulness Signals Into (1) Thematic Recommendation Modeling, (2) Weighted Rating Aggregation, And (3) Influence-based Community Ranking. Unlike Conventional Systems That Treat All User Interactions Equally, The Proposed Method Weights Contributions Based On Credibility Derived From Community Feedback. Experimental Evaluation On Benchmark Datasets Demonstrates Improved Diversity And Robustness While Maintaining Competitive Ranking Accuracy.

Author: Musa Idris | Yusuf Ibrahim Yusuf

Area of research: 3D Printing

This Paper Presents The Design, Fabrication, And Experimental Evaluation Of A Lightweight, Modular Robotic Supporting Arm Developed To Assist Individuals With Upper-limb Paralysis. To Balance Affordability And Structural Integrity, The Prototype Was Constructed Using Fused Deposition Modeling (FDM) 3D-printed Polylactic Acid (PLA) Components, Driven By Servo And DC Actuators, And Controlled Via An Arduino Uno Architecture. The Fabricated Assembly Features An Total Mass Of [Insert Weight] And Exhibits High Manufacturing Precision With A Dimensional Accuracy Of $\approx 0.2\text{ Mm}$. Electromechanical Evaluations Demonstrated That The Robotic Arm Achieves $\approx 120^\circ$ Of Elbow Flexion And $\approx 90^\circ$ Of Wrist Articulation. Functional Testing Under Loaded Conditions Confirmed Smooth Operational Kinematics And Repeatable Trajectory Tracking, With Minor Structural Deformation Observed During Prolonged Continuous Stress. Finally, We Discuss Critical Design Optimizations, Current System Limitations, And Future Work, Which Includes Transitioning To High-torque Actuation, Substituting Materials With PETG Or Aluminum Alloys, And Conducting Clinical User Trials.

Author: N Mansoor Basha | N Abhivishnu Naik | G Kiranmai | P Ugandhar

Area of research: Computer Science (Cybersecurity)

This Paper Presents BehaveGuard, A Security-focused System Designed To Protect Sensitive Files And Folders From Unauthorized Access. The System Continuously Monitors A Designated Folder And Detects Suspicious Activities Such As Unauthorized File Access, Modification, Or Deletion. Upon Detecting Such Actions, The System Automatically Locks The Device To Prevent Further Intrusion. To Enhance Security, The System Captures The Image Of The Unauthorized User Through The Webcam And Sends Real-time Alerts Via Telegram And Email. A Multi-layer Authentication Mechanism Is Implemented, Where The System Can Only Be Unlocked Using A Valid Password Followed By Approval Through Telegram Verification. The Proposed System Integrates File Monitoring, Anomaly Detection, Alert Mechanisms, And Access Control Into A Unified Backend Solution. This Approach Significantly Improves Data Security And Ensures Proactive Threat Detection And Response.

Author: Jeyadharshni.R | Harshadha Princy.P | Harshadha Princy.P | Devika R

Area of research: Cybersecurity, Password Strength

With The Increasing Risk Of Account Takeovers And Credential Misuse, Passwords Remain The Primary Factor That Gets Compromised In Data Breaches. This Project Presents PSAWBD — A Password Security Assessment Tool With Breach Detection, Developed Entirely In Python. The Tool Evaluates Password Strength Using Common Criteria Including Length, Character Diversity. It Checks The User's Password Against A Locally Stored Database Of Previously Breached Password Hashes Using SHA-1 Hashing To Protect The User's Credentials During Comparison. Additionally, It Simulates Real-world Attack Methods — Dictionary Attacks And Rockyou-list-based Attacks — Using Multi-threaded Execution To Test The Actual Resilience Of The Password Against Offline Cracking. The Results Are Delivered Through An Animated, Text-based Command-line Interface, Making The Tool Lightweight, Portable, And Easy To Use. This Project Promotes Better Password Hygiene And Raises Cybersecurity Awareness Among Everyday Users.

Author: Naveen Daniel M | Livin Jospher | Kamalesh P | Anbuselvan B | Devika R

Area of research: IoT-Based Robotics And Automation

This Paper Presents A Comprehensive Design And Implementation Of A Wi-Fi Transmission And Voice-Controlled Pickup And Place Robotic Arm Vehicle. The System Leverages An ESP8266 Microcontroller As The Master Control Unit (MCU), A Node.js Server Domain Hosted On A Local Network, And A Web Speech APIenabled Browser Interface To Facilitate Realtime Voice Command Processing. The Vehicle Combines A Multi-axis Servo-driven Robotic Arm With A Four-wheel DC Motor Drive Chassis, Enabling Remote Pickup And Precise Placement Of Objects Up To 500g. Wi-Fi Communication Is Routed Through A Dedicated Server Domain, Enhancing Reliability, Range, And Scalability Over A Direct IP-based Approach. The System Demonstrates The Convergence Of Embedded Systems, Wireless Networking, Server-side Computing, And Natural Language Voice Control Into A Single, Cost-effective Robotic Platform Suitable For Industrial, Humanitarian, And Educational Use Cases.

Author: Varun Mishra | Nikhil Hase | Kaustubh Deshmukh | Prof. Reena Asati

Area of research: Artificial Intelligence And Data Science

Mental Health Disorders Affect Over One Billion People Globally, Yet Timely Professional Support Remains Limited. This Paper Presents MindCare, An AI-powered Mental Health Chatbot Integrating Natural Language Processing (NLP), Machine Learning (ML), And Deep Learning (DL) For Empathetic, Context-aware Conversational Support. The System Employs A Fine-tuned BERT-based Transformer For Emotion And Intent Detection, A Bidirectional LSTM (BiLSTM) For Sentiment Analysis, And A Retrieval-Augmented Generation (RAG) Framework For Clinically Informed Responses. A Crisis Detection Module (CDM) Triggers Emergency Escalation When Suicidal Ideation Is Detected. Evaluation On Empathetic Dialogues And DAIC-WOZ Datasets Demonstrates 94.7% Emotion Accuracy, 91.3% Intent F1-score, And 87.2% User Satisfaction Across 500 Clinical Trial Participants Over 12 Weeks (p < 0.001).

Author: Ms. J Isaraelin Insulata | V Devendar Reddy | T Sai Narasimha Reddy | T Nandha

Area of research: Information Technology

The Aim Of This Research Is To Improve The Security Of ATMs By Creating ATM Models That Make Use Of Convolutional Neural Networks (CNN) And Recurrent Neural Networks (RNN) For Facial Recognition Of Users And User Approval.Materials And Methods: The Study Comprises Two Groups: Group 1 Has A CNN-only Model With A Sample Size Of 26 Samples, And Group 2 Uses A Hybrid CNN-RNN Model With The Same Sample Size Of 26. The Statistical Analysis Of The Study Was Conducted With A G Power Of 80%, A Significance Level Of 0.05%, And A Confidence Interval Of 95%. Results: The Findings Indicate That The Hybrid CNN-RNN Model Outperforms The CNN-only Model By A Significant Margin. The Accuracy Of The Hybrid Model Was Between 91.8% And 98.7%, While That Of The CNN-only Model Was Between 85.3% And 92.4%.The Maximum Accuracy Found Was 98.7% At P-value 0.0480, With P = 0.005. But The Dataset For This Research Was Only 26 Samples Per Group, Which Is A Concern Regarding The Model's Generalizability To Larger, Real-world Datasets. More Research Is Required To Evaluate How This Model Would Scale And Perform On Larger Datasets To Establish Its Wider Applicability. Conclusion: The Results Indicate That The Hybrid CNN-RNN Model Performs Better Than The CNN-only System In Facial Recognition And Enhances ATM Security. Yet, Scalability And Generalizability Of The Model Would Require Further Investigation With Larger Datasets.

Author: S.sasipriya | V.Ramesh Kannan

Area of research: Thermal Interface Materials

The CPU Thermal Paste Is A Kind Of Material That We Put Between The Processor And The Heat Spreader Or Heat Sink. This Helps To Reduce The Resistance To Heat Flow Between These Two Parts. The CPU Thermal Paste Is Not Meant To Be Better Than Copper Or Aluminum At Conducting Heat. It Helps To Fill In The Tiny Gaps Between The Surfaces With A Material That Can Spread The Heat Easily.The CPU Thermal Paste Works By Replacing The Air In These Gaps With A Material That Can Wet The Surface And Create A Thin Path For The Heat To Flow. Many People Have Studied The CPU Paste, Like Prasher In 2006 And Razeeb And Others In 2018 And Wei And Others In 2024. They Found That The Performance Of The CPU Paste Depends On Many Things, Such As How Well The Filler Material Conducts Heat And The Thickness Of The Paste. The Pressure On The Paste And How Well It Wets The Surface.It Is Not About How Well The CPU Thermal Paste Conducts Heat, But Also About Its Long-term Stability. Some Other People, Like Goel And Others In 2008 And Prasher And Shipley And Others In 2003 Have Also Studied This. This Article Will Talk About How The CPU Thermal Paste's Made, What Materials Are Used, Strengths, Weaknesses And How It Has Changed Over Time From Old Grease-based Systems To New Nanostructured And Hybrid Systems. We Will Also Talk About Alternatives To The CPU Thermal Paste, Such As Phase-change Materials, Graphite Sheets, Liquid Metals And Metal-based Bonded Interfaces, Which Have Been Studied By People Like S Chen And Others In 2020 Hoffmeyer And Others In 2017 Lee And Kim In 2024 And Razeeb And Others, In 2018.

Author: Yogesh Bataw | Dr. Priya Joshi | Rushikesh Pansare | Tanmay Pagar | Khush Patil

Area of research: CSE

With The Increasing Risk Of Account Takeovers And Credential Misuse, Passwords Remain The Primary Factor That Gets Compromised In Data Breaches. This Project Presents PSAWBD — A Password Security Assessment Tool With Breach Detection, Developed Entirely In Python. The Tool Evaluates Password Strength Using Common Criteria Including Length, Character Diversity. It Checks The User's Password Against A Locally Stored Database Of Previously Breached Password Hashes Using SHA-1 Hashing To Protect The User's Credentials During Comparison. Additionally, It Simulates Real-world Attack Methods — Dictionary Attacks And Rockyou-list-based Attacks — Using Multi-threaded Execution To Test The Actual Resilience Of The Password Against Offline Cracking. The Results Are Delivered Through An Animated, Text-based Command-line Interface, Making The Tool Lightweight, Portable, And Easy To Use. This Project Promotes Better Password Hygiene And Raises Cybersecurity Awareness Among Everyday Users.

Author: Mr. Shabeer V.V | Avula Vamsi Krishna Reddy | Busupalli Uma Maheswara Reddy | Butreddy Chaitanya Reddy

Area of research: Computer Science

Tomato Leaf Diseases Pose A Persistent Challenge To Sustainable Crop Production, Requiring Accurate And Computationally Efficient Diagnostic Solutions For Real-time Field Applications. This Study Proposes A Hybrid Deep Learning Framework That Combines A Lightweight Convolutional Neural Network (CNN) With Lightweight Transformer Architecture For Robust Classification Of Tomato Leaf Diseases. The CNN Component Is Employed To Extract Fine-grained Local Spatial Features, While The Transformer Module Captures Global Contextual Dependencies, Enabling Improved Discrimination Between Visually Similar Disease Categories. The Model Was Trained And Evaluated On A Multi-class Tomato Leaf Dataset Comprising Healthy And Diseased Samples, Including Bacterial Spot, Early Blight, Late Blight, Leaf Mold, And Septoria Leaf Spot. Data Augmentation And Transfer Learning Strategies Were Applied To Enhance Generalization And Mitigate Overfitting. The Proposed Hybrid Model Achieved An Overall Classification Accuracy Of 99.08%, With A Precision Of 98.93%, Recall Of 98.95%, And F1-score Of 98.94%. Comparative Analysis Indicates Superior Performance Over Standalone Lightweight CNN And Transformer Models, While Maintaining Reduced Computational Complexity Suitable For Resource-constrained Environments. To Improve Model Interpretability, Gradient-weighted Class Activation Mapping (Grad-CAM) Was Utilized To Visualize Disease-relevant Regions, Confirming That The Model Focuses On Meaningful Pathological Features. The Results Demonstrate That The Proposed Approach Provides A Reliable, Efficient, And Interpretable Solution For Automated Plant Disease Detection, Supporting Its Applicability In Precision Agriculture And Smart Farming Systems.

Author: Rupesh Kumar | Dr. Ranu Pandey

Area of research: Cyber Security

The Pervasive Adoption Of Digital Platforms Has Precipitated A Concomitant Escalation In Phishing Attacks, Wherein Malicious Uniform Resource Locators (URLs) Serve As The Primary Vector For Credential Theft, Financial Fraud, And Malware Dissemination. Conventional Detection Paradigms, Including Blacklist-based Filtering, Signature Matching, And Rule-based Heuristics, Are Demonstrably Insufficient Against Zero-day Attacks And Polymorphic Phishing Campaigns That Continuously Mutate To Evade Static Defenses. This Paper Presents An Integrated Framework For Real-time Malicious URL Detection Leveraging A Multi-layered Artificial Intelligence Architecture Comprising Logistic Regression (LR), Decision Trees (DT), Random Forest (RF), Support Vector Machines (SVM), Convolutional Neural Networks (CNN), And Recurrent Neural Networks (RNN) With Long Short-Term Memory (LSTM) Cells. A Comprehensive Feature Engineering Pipeline Extracts URL-lexical, Domain-metadata, Content-structural, And Behavioral Attributes From Heterogeneous Data Sources Including Kaggle And PhishTank Repositories. The Proposed System Is Deployed Within A Django-based Full-stack Web Application Integrated With The VirusTotal API For Real-time Threat Intelligence Augmentation. A Secure Continuous Integration And Continuous Deployment (CI/CD) Pipeline Ensures Automated Testing, Vulnerability Assessment, And Deployment Integrity. Experimental Evaluation Demonstrates That Ensemble And Deep Learning Models Achieve Superior Classification Performance Across Accuracy, Precision, Recall, F1-score, And ROC-AUC Metrics Compared To Legacy Detection Methods. The System Addresses Identified Research Gaps Including Concept Drift Resilience, Dataset Imbalance, And Adversarial Robustness. Future Directions Encompass Transformer-based Architectures, Federated Learning, And Explainable AI (XAI) Integration.

Author: Dr. Ravindra Krishna Chandar | Keerthi Kumar A | Satheesh Kanna S | Bavithran C | Nikesh S

Area of research: Computer Science And Engineering

Traditional Healthcare Systems Rely Heavily On Manual Monitoring Of Patient Vital Signs, Which Can Result In Delayed Medical Response, Inaccurate Readings, And Inefficient Healthcare Management. To Overcome These Challenges, This Project Proposes An IoT Driven Name-Based Patient Monitoring System For Real-Time Healthcare. The System Uses IoT Sensors To Continuously Monitor Important Health Parameters Such As Heart Rate, Body Temperature, And Oxygen Level (SpO2). The Collected Data Is Transmitted To A Cloud Server Through A Microcontroller Like ESP32/ESP8266 For Real-time Monitoring And Storage. Each Patient Is Uniquely Identified Using A Name-based Record System, Reducing Data Mismatch And Improving Patient Management. Doctors And Healthcare Providers Can Access Patient Information Remotely Through A Web Or Mobile Application And Receive Automatic Alerts During Abnormal Health Conditions. The Proposed System Aims To Improve Healthcare Efficiency,

Author: Mrs. K Pradeepa | Aalamul Irfan I | Aathavan S | Sabarish K

Area of research: Machine Learning In Healthcare Prediction Systems

Heart Disease Is One Of The Leading Causes Of Death Worldwide, And Early Prediction Plays An Important Role In Reducing Severe Health Risks. Traditional Diagnostic Methods Such As ECG, Blood Tests, And Imaging Require Expert Interpretation, More Time, And Higher Cost. This Paper Proposes An Efficient Heart Disease Prediction System Using Machine Learning Techniques To Predict The Possibility Of Heart Disease Based On Clinical Data Such As Age, Blood Pressure, Cholesterol, Chest Pain Type, ECG Results, And Heart Rate. The System Applies Algorithms Such As Logistic Regression, K-Nearest Neighbors, Support Vector Machine, Random Forest, And XGBoost. Among These Models, Random Forest And XGBoost Provide Better Accuracy And Reliability. Explainable AI Techniques Such As SHAP And LIME Are Also Used To Make The Prediction Results Transparent And Understandable For Doctors And Patients.

Author: Mr. P. Ranjith | Krithik Ragul S | Moses Praveen R | Vigneshwaran M

Area of research: Computer Science And Business Systems

Credit Card Fraud Has Become A Significant Challenge In The Digital Economy, Resulting In Substantial Financial Losses And Reduced Trust In Online Transactions. This Paper Presents A Hybrid Adaptive Sampling With Risk-Based Fraud Detection (HAS-RFD) Framework Integrated With Explainable Artificial Intelligence (XAI) Techniques To Improve Fraud Detection Performance. The Hybrid Sampling Method Combines SMOTE-based Oversampling With Clustering-based Adaptive Undersampling To Address Class Imbalance While Preserving Important Data Patterns. Machine Learning Models Such As Random Forest And Logistic Regression Are Trained On The Balanced Dataset To Classify Transactions As Fraudulent Or Legitimate. The System Incorporates A Risk-based Classification Mechanism That Categorizes Each Transaction Into Low, Medium, Or High Risk Levels. Explainability Techniques Provide Clear Insights Into Fraud Predictions, Enhancing Transparency And User Trust. The Proposed System Achieves Improved Detection Accuracy, Reduced False Positives, And Enhanced Interpretability, Making It Suitable For Real- World Financial Applications.

Author: Mrs.M.Karthika | Rubhavashni LR | Muthamil E | Lavanya Shri R

Area of research: Civil Engineering

Industrial Wastewater Generated From Manufacturing And Processing Industries Contains High Concentrations Of Organic Matter, Dissolved Solids, Nutrients, Oil And Grease, And Other Pollutants That Pose Serious Environmental And Public Health Risks. This Study Experimentally Investigates The Effectiveness Of Water Hyacinth (Eichhornia Crassipes) As A Low-cost And Eco-friendly Phytoremediation Technique For Improving The Quality Of Industrial Wastewater. Wastewater Samples Were Collected From An Industrial Source And Treated Using A Tank-based Water Hyacinth System Over A Period Of Six Months. Key Water Quality Parameters Such As PH, Total Dissolved Solids (TDS), Biochemical Oxygen Demand (BOD), Chemical Oxygen Demand (COD), Oil And Grease, And Nitrate Concentration Were Analyzed At Monthly Intervals. The Results Indicated A Gradual And Significant Improvement In Water Quality, With Reductions Of Approximately 47.6% In TDS, 52.5% In BOD, 54.2% In COD, 48.3% In Oil And Grease, And 44.3% In Nitrate Concentration. The Findings Demonstrate That Water Hyacinth Is An Effective, Economical, And Sustainable Treatment Option For Industrial Wastewater, Particularly Suitable For Decentralized Treatment Systems.

Author: Sarthak Thakare | Piyush Patil | Prem Patil | Mohit Mali | Prof. Dr. R. K. Pawar

Area of research: Deep Learning

The Banking Appointment System Is A Web-based Application Designed To Improve Customer Service Efficiency In Banking Institutions. Traditional Banking Systems Often Face Issues Such As Overcrowding, Long Waiting Times, And Inefficient Customer Handling. The Proposed System Allows Customers To Schedule Appointments Online By Selecting Banking Services, Preferred Dates, And Available Time Slots. The System Helps Reduce Waiting Time, Improves Customer Satisfaction, And Enables Better Management Of Banking Operations. The Application Uses Modern Web Technologies And Database Management Systems To Provide Secure And Reliable Appointment Scheduling.

Author: Gayathri N G | Athithya p | Manikandan M | Riyash M

Area of research: Civil Engineering

This Project Explores Innovative, Eco-friendly Solutions To Manage Riverbank Erosion In India Using Construction And Demolition (C&D) Waste. By Repurposing Waste Materials Like Concrete And Bricks, The Approach Aims To Reduce Environmental Pollution, Cut Costs, And Restore River Habitats. Although Still In Early Stages, This Sustainable Method Holds Great Promise For Protecting Rivers While Promoting Circular Use Of Resources. Riverbank Erosion And Habitat Degradation Are Pressing Environmental Issues That Threaten Ecosystems And Human Settlements Alike. This Project Explores The Innovative Utilization Of Construction And Demolition (C&D) Waste As A Sustainable And Cost-effective Solution For River Erosion Control And Habitat Restoration. Traditionally Viewed As A Disposal Challenge, C&D Waste Comprising Concrete, Brick, Metal, And Asphalt Can Be Repurposed To Stabilize Riverbanks, Reduce Sediment Load, And Create Microhabitats For Aquatic And Riparian Species. This Study Assesses The Mechanical Properties And Ecological Compatibility Of Various C&D Materials And Demonstrates Their Effectiveness Through Case Studies, Lab Testing, And Field Pilot Installations. Emphasis Is Placed On Modular Gabion Structures, Artificial Reefs, And Vegetated Revetments Made From Processed Waste Materials.

Author: Patil Rubee | More Kartika | Patil Rutuja | Chaudhari Pallavi | Prof. G.V.Tapkire

Area of research: Construction Materials

This Project Focuses On Designing An M40 Grade Concrete Mix Used In High-strength Structures Like Bridges And High-rise Buildings. The Aim Is To Achieve A Compressive Strength Of 40 MPa With Good Workability, Durability, And Cost Efficiency.The Mix Design Was Carried Out As Per IS 10262:2019 And IS 456:2000. Material Tests Were Conducted To Determine Key Properties Required For Accurate Mix Proportions.Trial Mixes Were Prepared And Evaluated, And The Final Mix Met The Required Strength And Workability. Compared To M20 And M30, M40 Showed Better Performance For Heavy-load Structures.Overall, The Study Highlights The Importance Of Proper Testing And Quality Control In Concrete Mix Design.

Author: Aniket Thakur | Prathamesh Shinde | Chetan Wadile | Rohit Patil | Dr. Hemraj Kumavat

Area of research: Maharashtra Institute Of Pharmacy D- Pharm Betala

The Study Aimed To Formulate An Herbal Powder Shampoo And Evaluate Its Physicochemical Properties. The Herbal Liquid Shampoo Was Formulated By Adding The Fine Powder Of Shikakai, Methi, Hibiscus, Neem, Ashwagandha, Reetha, Amla, Cinnamon, Kalonji And Rose. Several Tests Such As Organoleptic Character (Odor, Color And Texture), Bulk Density, Tapped Density, Moisture Content, Dirt Dispersion, PH, Water Solubility And Foaming Index Were Performed To Determine The Physicochemical Properties. All The Evaluation Parameters Give The Satisfactory Results.

Author: Ashwini Chandrahashya Balpande | Payal Kuldip Indurkar

Area of research: Cyber Security

Source Code Similarity Measurement, Which Involves Assessing The Degree Of Difference Between Code Segments, Plays A Crucial Role In Various Aspects Of The Software Development Cycle. These Include But Are Not Limited To Code Quality Assurance, Code Review Processes, Code Plagiarism Detection, Security, And Vulnerability Analysis. Despite The Increasing Application Of ML Technique In This Domain, A Comprehensive Synthesis Of Existing Methodologies Remains Lacking. This Paper Presents A Systematic Review Of Machine Learning Techniques Applied To Code Similarity Measurement, Aiming To Illuminate Current Methodologies And Contribute Valuable Insights To The Research Community. Following A Rigorous Systematic Review Protocol, We Identified And Analysed 84 Primary Studies On A Broad Spectrum Of Dimensions Covering Application Type, Devised Machine Learning Algorithms, Used Code Representations, Datasets, And Performance Metrics, As Well As Performance Evaluations. A Deep Investigation Reveals That 15 Applications For Code Similarity Measurement Have Utilised 51 Different Machine Learning Algorithms. Additionally, The Most Prevalent Code Representation Is Found To Be The Abstract Syntax Tree (AST). Furthermore, The Most Frequently Employed Dataset Across Various Code Similarity Research Applications Is BigCloneBench. Through This Comprehensive Analysis, The Paper Not Only Synthesises Existing Research But Also Identifies Prevailing Limitations And Challenges, Shedding Light On Potential Avenues For Future Work.

Author: Mrs.D.Sasikala | Y.Vishnuvardhan Reddy | G. Jaswanth | M.Bhanu Prakash

Area of research: Computer Science

In Many Situations, Citizens Are Required To Carry Multiple Identity Proofs Such As Driving License, PAN Card, Voter ID, And Ration Card For Verification. Managing Several Physical Documents Can Be Inconvenient And Increases The Risk Of Loss Or Damage. This Paper Proposes A System Called White Card Application, Which Integrates Multiple Government Identity Proofs Into A Single Unified Digital Identity System.The System Stores Citizen Identity Information In A Centralized Database And Generates A Unique 12-digit White Card Number Along With A QR Code For Quick Verification. Government Officers From Different Departments Such As RTO, IT, Voting, And Ration Can Verify And Update The Respective Identity Details Through The System. When The QR Code Is Scanned, All Verified Identity Information Can Be Accessed Instantly.This Approach Reduces The Need To Carry Multiple Physical Cards, Simplifies Identity Verification, And Improves The Efficiency And Security Of Citizen Identity Management.

Author: Dr.N.Dhivya | Dhivya N, MCA.,M.Phil.,PhD.,

Area of research: Mechanical Engineering

This Project Explores The Design And Fabrication Of An Elliptical Trammel (Trammel Of Archimedes) Using Fused Deposition Modeling (FDM) To Demonstrate The Inversion Of A Double Slider-crank Chain. The Mechanism, Which Translates Constrained Linear Motion Into A Precise Elliptical Trajectory, Was Digitally Modeled Via CAD Software With A Specific Focus On Parametric Geometry, Including A Semi-major Axis Of 136.5 Mm And A Semi-minor Axis Of 84 Mm. The Components Were Fabricated Using Polylactic Acid (PLA) On A QIDI TECH Q1 Pro 3D Printer, Utilizing A 0.2 Mm Layer Height And 20% Grid Infill To Balance Structural Rigidity With Production Efficiency. Critical Attention Was Paid To Dimensional Tolerances (0.2 Mm To 0.4 Mm) To Ensure Smooth Slider-to-track Interaction And Minimize Mechanical Backlash. The Final Assembly, Completed In 2 Hours And 29 Minutes, Validates The Effectiveness Of Additive Manufacturing For Rapid Prototyping And Low-cost Educational Tools. This Work Serves As A Functional Benchmark For Kinematic Analysis And Illustrates The Practical Application Of 3D Printing In Mechanical Engineering Education.

Author: M Rajesh | S MD Shahid

Area of research: Computer Engineering

The Modern Mental Health Management Tools Need Something More Sensitive To The Actual Condition Of A Person, Beyond What They Are Willing To Disclose Via Their Communication With The System. The Currently Existing Solutions, Like Wysa And Woebot, Are Exclusively Based On The Text Input. Therefore, A User Typing “I Am Fine” While Actually Being In Serious Trouble, Cannot Be Identified. To Improve That Situation, HealMind AI Was Created To Analyze Three Parallel Streams Of Information: The Facial Expression Analyzed By A Webcam, The Tone Of Voice Analyzed By A Microphone, And Text Generated Via Journaling Or During Chat. There Is A Neuro-Symbolic Hybrid Logic Engine Serving As A Backbone Of This Technology, Combining Machine Learning Algorithms With A Rule-Based System And Providing All Results Within The Scope Of Clinical Safety. With A Help Of A Feature Fusion Layer, It Is Possible To Resolve Contradictions Between Different Input Channels. Thus, When A User's Facial Expression Seems Troubled, But His/her Voice Seems Calm, The Level Of His/her Stress Is Classified As Moderate Instead Of Reaching The Extremes. After The Stress Is Detected, A Stress Dashboard Triggers A Search For Nearby Clinics Using The Google Maps API, Allowing Users To Find A Suitable Help In Allopathy, Homeopathy, Or Ayurveda. The Overall Efficiency Of The Combination Equals 88%, Which Is Much Better Than Any Of The Single Channel Solutions.

Author: Dr. S.S. Patil | Prof. S.S. Shinde | Sanika Ekshette | Jayashree Devkar | Payal Dhasade | Kalyani Sul

Area of research: Artificial Intelligence & Data Science

Traditional Handicrafts Are Culturally Significant But Face Challenges In Modernization, Digital Exposure, And Market Accessibility. This Project Presents An AI-Assisted Craft Design Customization And Marketplace Platform That Bridges Traditional Craftsmanship With Intelligent Digital Tools. The System Enables Users To Select A Craft Style And Customize Design Parameters Such As Motif Density, Colour Themes, Pattern Type, And Traditional-modern Blend Ratio. Using Pretrained Artificial Intelligence Models For Image Generation, The Platform Produces Craft-inspired Design Variations Based On User Inputs. To Enhance Usability, Generated Designs Can Be Previewed On Product Mock-ups Such As Textiles, Pottery Surfaces, And Carved Panels. The Platform Also Includes A Marketplace Module Where Users Can Upload, Describe, Price, And List Their Generated Or Handcrafted Designs For Browsing And Simulated Purchase. This Integrates Creativity With A Digital Commerce Ecosystem. The Project Emphasizes Accessibility, Cultural Preservation, And Digital Empowerment Of Artisans By Combining AI-assisted Creativity With A Scalable Web-based Marketplace. The System Demonstrates How Intelligent Tools Can Support Traditional Art Forms Without Replacing Human Craftsmanship, Offering A Practical Model For Technology-enabled Craft Innovation.

Author: Nuthalapati Bhagyasri Lakshmi | Gujjala Sindhu | Vannemreddi Tejasree | Ms. M.Sivasangari

Area of research: ENT

Coronavirus Disease 2019 (COVID‑19) Caused An Unprecedented Global Health Crisis And Was Associated With Multiple Secondary Infections, Among Which Mucormycosis Emerged As One Of The Most Severe Opportunistic Fungal Diseases. During The Second Wave Of The Pandemic, Especially In India, A Dramatic Rise In COVID‑19 Associated Mucormycosis (CAM) Cases Was Reported. CAM Predominantly Affected Patients With Uncontrolled Diabetes Mellitus, Prolonged Corticosteroid Therapy, Immune Dysregulation, Hypoxia, And Prolonged Hospitalization. Rhino‑orbito‑cerebral Mucormycosis Was The Most Common Presentation In India, Whereas Pulmonary Mucormycosis Was More Frequently Observed In Western Countries. The Present Review Summarizes The Epidemiology, Etiopathogenesis, Risk Factors, Clinical Manifestations, Diagnosis, Treatment, And Preventive Strategies Related To CAM Globally With Special Emphasis On India. Literature From Peer‑reviewed Journals, WHO Reports, PubMed, And Google Scholar Was Reviewed To Understand The Burden And Outcomes Of This Emerging Fungal Infection. The Findings Indicate That Irrational Steroid Use, Poor Glycemic Control, Endothelial Dysfunction, And Impaired Immunity During COVID‑19 Significantly Contributed To The Outbreak. Early Diagnosis, Surgical Debridement, Prompt Antifungal Therapy, And Management Of Comorbid Conditions Remain Essential For Reducing Mortality. The Review Also Highlights The Need For Awareness Among Healthcare Professionals Regarding Rational Steroid Use And Early Recognition Of Warning Symptoms.

Author: Satpinder Kaur | Vishal salhotra

Area of research: Net Zero Energy Buildings

The Building Sector Significantly Contributes To Energy Consumption, Particularly In Hot And Humid Regions Where Cooling And Dehumidification Demands Are High. In Rapidly Growing Cities Like Mumbai, Pune, And Nagpur, High Temperatures (30°C–38°C) And Humidity Levels (70–85%) Increase Reliance On Mechanical Cooling, Leading To Higher Energy Use. This Study Presents A Structured Workflow For Climate-responsive Building Design To Achieve Net Zero Energy Building (NZEB) Performance. A Seven-phase Methodology Is Adopted, Including Climate Analysis Using Typical Meteorological Year (TMY) Data, Baseline Model Development, And Building Energy Modelling (BEM) Using EnergyPlus And DesignBuilder. Passive And Active Strategies Are Integrated To Optimize Performance, Followed By Solar Photovoltaic System Implementation. Results Show Significant Energy Reduction, Enabling NZEB Achievement And Supporting A Climate-specific Framework Aligned With ECBC Guidelines.

Author: Shreyas Vijay Kadam

Area of research: CYBER SECURITY

Blockchain Technology Offers A Robust And Innovative Solution For Certificate Verification And Validation, Addressing Prevalent Issues Such As Fraud, Forgery, And Data Manipulation. By Leveraging The Decentralized, Immutable, And Transparent Nature Of Blockchain, Educational Institutions, Certification Bodies, And Employers Can Ensure The Authenticity And Integrity Of Digital Certificates. Each Certificate Is Recorded On The Blockchain As A Unique Digital Asset, Secured Through Cryptographic Hashing. This Process Guarantees That Any Attempt To Alter Or Counterfeit The Certificate Can Be Easily Detected, As Even The Slightest Modification Would Change The Hash Value. Furthermore, Blockchain's Decentralized Ledger Allows Multiple Stakeholders To Access And Verify Certificates In Real-time Without Relying On A Central Authority, Enhancing Trust And Reducing Administrative Overhead. The Use Of Smart Contracts Automates The Validation Process, Ensuring That Only Verified And Authorized Certificates Are Added To The Blockchain. As A Result, Blockchain-based Certificate Verification And Validation Provide A Secure, Efficient, And Transparent Framework That Significantly Enhances The Reliability Of Credentialing Systems. Counterfeit Academic And Professional Certificates Continue To Erode Institutional Trust And Compromise Merit-based Selection Worldwide. This Paper Proposes A Decentralized Digital Credentialing Framework Anchored In Blockchain Immutability, SHA-256 Cryptographic Hashing, And Programmable Ethereum Smart Contracts. Upon Certificate Issuance, A Canonical JSON Credential Object Is Hashed And Its Digest Is Permanently Inscribed On A Distributed Ledger Through An Auditable Smart Contract Transaction, Rendering Retroactive Modification Computationally Infeasible. Verification By Any Relying Party—employer, Licensing Authority, Or Academic Registrar—requires Only A Client-side Hash Recomputation And A Read-only Blockchain Query; Any Discrepancy Between The Submitted And Stored Digest Triggers Immediate Rejection Without Institutional Intermediation. Role-differentiated Access Control Governs Issuance, Retrieval, And Revocation Across All Stakeholder Classes. The Proposed Architecture Furnishes A Secure, Transparent, And Globally Accessible Credentialing Infrastructure Applicable To Universities, Certification Bodies, And Cross-border Employers.

Author: Dr.N.Sundararajulu | M.Ravi Bhaskar | S.Venkata siva | P.Koteswara Rao

Area of research: Civil Engineering

Concrete Is One Of The Most Widely Used Construction Materials Because Of Its Strength, Durability, And Versatility. However, Large-scale Cement Production Creates Serious Environmental Problems Due To Excessive Carbon Dioxide Emissions And Depletion Of Natural Resources. At The Same Time, Marble Industries Generate Huge Quantities Of Waste Marble Powder During Cutting, Polishing, And Finishing Operations. Improper Disposal Of This Waste Causes Air Pollution, Soil Contamination, And Environmental Degradation. The Present Study Focuses On The Utilization Of Marble Powder As A Partial Replacement Of Cement In Concrete Production To Achieve Sustainable And Eco-friendly Construction Practices. In This Investigation, Marble Powder Was Used As A Partial Replacement Of Cement In Different Proportions For M20 Grade Concrete. Various Tests Were Conducted To Evaluate The Mechanical Properties Of Concrete, Including Compressive Strength, Split Tensile Strength, And Flexural Strength. The Study Observed That The Addition Of Marble Powder Improved The Strength Characteristics Of Concrete Up To An Optimum Replacement Level Due To The Filler Effect And Improved Particle Packing Within The Concrete Matrix. Beyond The Optimum Percentage, The Strength Gradually Decreased Because Of Reduction In Cementitious Bonding Properties.

Author: Dr.Yogesh Sonawane | Shivam Ramesh Wadile | Yash Pravin Samudre | Gaurav Jagan Chavan | Vivek Dipak Kakade

Area of research: Computer Science And Engineering

The Proliferation Of AI-generated Images Poses Pressing Threats To Information Integrity, Cybersecurity, And Public Trust. This Paper Presents DeepScan AI, A Multi-layer, Heuristic-driven Image Authenticity Verification System Engineered Without Reliance On Large-scale Labeled Training Datasets. The Proposed System Extracts And Fuses Seven Distinct Feature Channels—EXIF Metadata Integrity, Luminance Noise Distribution, Texture Complexity (variance-based), RGB Channel Entropy, Edge-gradient Regularity, Frequency-domain Artifacts, And Aspect-ratio Consistency—to Generate A Probabilistic Authenticity Score In The Range [0, 100]. Deployed As A Lightweight Web Application Using Python And Streamlit, DeepScan AI Achieves An Experimental Detection Accuracy Of 85.5% With A Mean Inference Latency Of 2–4 Seconds, Operating Entirely On-device To Preserve User Privacy. Unlike Cloud-dependent Or Deep-learning-heavy Alternatives, The System Is Zero-cost, Transparent, And Immediately Accessible To Non-technical Users. Results Demonstrate That Multi-feature Fusion Significantly Outperforms Single-channel Heuristic Baselines And Establishes A Practical Foundation For Further Deep-learning Augmentation.

Author: Gayatri P. Nandavadekar | Shrenika V. Salunkhe

Area of research: Artificial Intelligence And Data Science

This Project Proposes An Integrated Web Application Designed To Provide Realtime Emergency Assistance And Efficient Resource Navigation During Critical Situations. The System Functions As A Centralized Platform That Delivers Instant Emergency Alerts, Safety Instructions, And Important Updates To Users,enabling Them To Respond Promptly To Crises. By Offering Timely Information And Connectivity To Essential Services, The Application Supports Effective Decision-making And Improves Situational Awareness During Emergencies. The Application Integrates A Location-based Mapping System To Help Users Quickly Identify And Access Nearby Services Such As Hospitals, Ambulances,medical Stores, Mechanics, Petrol Stations, And Towing Facilities. Secure Data Handling Is Ensured Through The Use Of The AES Encryption Algorithm, While SQL-based Database Operations Enable Efficient Storage And Retrieval Of Information. Overall, The System Enhances Coordination Between Users And Service Providers, Ensuring Reliable, Secure, And Rapid Emergency Support.

Author: G. Keerthnasri | G. Guna Vardhan | K. Venkata Ganesh | P. Prabhudeep

Area of research: ARTIFICIAL INTELLIGENCE

The AI-Based Cyber Threat Prediction System Improves This Paper Presents A Telemedicine Platform With AI-Powered Diagnostics Designed To Improve Healthcare Accessibility And Remote Patient Monitoring. Traditional Healthcare Systems Often Face Challenges Such As Limited Medical Access In Rural Areas, Delayed Diagnosis, And Shortage Of Healthcare Professionals. The Proposed System Integrates Telemedicine Services With Artificial Intelligence (AI) Techniques To Provide Remote Consultation, Disease Prediction, Symptom Analysis, And Real-time Health Monitoring. The Platform Enables Patients To Consult Doctors Online While AI Algorithms Assist In Early Diagnosis And Decision-making. The System Enhances Healthcare Efficiency, Reduces Consultation Delays, Minimizes Operational Costs, And Supports Digital Healthcare Transformation.

Author: Mrs. M. Kirithika Devi | Sowmiya M | Srineeha SS | Teena Sherly P | Vijayalakshmi M

Area of research: Computer Science And Business System

This Paper Presents An Intelligent Credit Scoring And Loan Approval System Using Behavioral Analytics To Improve The Accuracy And Efficiency Of Loan Approval Decisions. Traditional Systems Rely Mainly On Credit History And Financial Records, Which May Not Fairly Evaluate Users With Limited Credit Background. The Proposed System Uses Machine Learning Techniques To Analyze Applicant Details Such As Income, Loan Amount, Employment Status, Existing Loans, And Behavioral Data From Transaction Patterns And Payment History. A Random Forest Algorithm Is Used To Predict Loan Approval Status And Generate A Credit Score. The System Also Provides Secure Login, Loan Application Processing, Document Upload, And An Admin Dashboard For Verification And Approval. This Approach Improves Decisionmaking, Reduces Manual Effort, Reliability Of Loan Processing

Author: Dr.PShanmuga Priya | Anantha Lakshme V | Nivitha M | Pradeepa S

Area of research: Computer Science And Business Systems

This Paper Presents A Unified Object Recovery System For Campus Communities Designed To Simplify The Management Of Lost And Found Items Within Educational Institutions. Traditional Methods Such As Notice Boards And Verbal Communication Are Inefficient And Unorganized. The Proposed System Provides A Centralized Web-based Platform Where Users Can Report Lost Or Found Items With Details Such As Description, Location, Date, And Images. The System Supports Searching, Matching, And Notification Features To Improve Item Recovery Efficiency. It Reduces Manual Effort, Enhances Transparency, And Ensures Secure Access Through User Authentication

Author: Karthikam | Asha R | Mekala M | Sivasangari S

Area of research: Artificial Intelligence And Data Science

Source Code Similarity Measurement, Which Involves Assessing The Degree Of Difference Between Code Segments, Plays A Crucial Role In Various Aspects Of The Software Development Cycle. These Include But Are Not Limited To Code Quality Assurance, Code Review Processes, Code Plagiarism Detection, Security, And Vulnerability Analysis. Despite The Increasing Application Of ML Technique In This Domain, A Comprehensive Synthesis Of Existing Methodologies Remains Lacking. This Paper Presents A Systematic Review Of Machine Learning Techniques Applied To Code Similarity Measurement, Aiming To Illuminate Current Methodologies And Contribute Valuable Insights To The Research Community. Following A Rigorous Systematic Review Protocol, We Identified And Analysed 84 Primary Studies On A Broad Spectrum Of Dimensions Covering Application Type, Devised Machine Learning Algorithms, Used Code Representations, Datasets, And Performance Metrics, As Well As Performance Evaluations. A Deep Investigation Reveals That 15 Applications For Code Similarity Measurement Have Utilised 51 Different Machine Learning Algorithms. Additionally, The Most Prevalent Code Representation Is Found To Be The Abstract Syntax Tree (AST). Furthermore, The Most Frequently Employed Dataset Across Various Code Similarity Research Applications Is BigCloneBench. Through This Comprehensive Analysis, The Paper Not Only Synthesises Existing Research But Also Identifies Prevailing Limitations And Challenges, Shedding Light On Potential Avenues For Future Work.

Author: Mrs.K.Gowri | Mr.E.Naveen | Mr.M.Vijay | Mr.M.Logeswaran

Area of research: Civil Engineering

The Construction Industry Continuously Seeks Advanced Technologies To Reduce Construction Cost, Project Duration, Labor Dependency, And Material Wastage. Concrete 3D Printing (C3DP), Also Known As Additive Manufacturing In Construction, Has Emerged As An Innovative Technology Capable Of Transforming Conventional Construction Practices. This Study Evaluates The Time And Cost Efficiency Of Concrete 3D Printing Compared To Traditional Construction Methods. The Comparative Analysis Indicates That Concrete 3D Printing Significantly Reduces Construction Time, Labor Requirements, And Material Wastage While Improving Construction Accuracy And Sustainability. The Study Concludes That C3DP Has Strong Potential For Affordable Housing And Sustainable Infrastructure Development In India Despite Challenges Such As High Initial Investment And Limited Technical Expertise.

Author: Rohit Shinde | Dr. P. M. Alandkar | Ms.Swati Kshirsagar

Area of research: Civil Engineering

The Construction Industry Is One Of The Major Consumers Of Natural Resources And Construction Materials, Generating Significant Quantities Of Construction Waste And Environmental Pollution. Conventional Reinforced Cement Concrete (RCC) Construction Involves High Consumption Of Cement, Steel, Aggregate, And Formwork Materials, Resulting In Increased Project Cost, Material Wastage, And Carbon Emissions. Ferro-cement Technology Has Emerged As A Lightweight And Sustainable Alternative Construction Technique Due To Its Reduced Material Requirement, Thin Structural Sections, And Improved Construction Efficiency. The Present Study Focuses On The Analysis Of Material Efficiency And Waste Reduction Achieved Using Ferro-cement Construction Technology In Comparison With Conventional RCC Construction. A Comparative Study Of RCC And Ferro-cement Wall Panels Was Carried Out Based On Quantity Estimation, Material Consumption Analysis, And Construction Waste Evaluation. The Results Obtained From The Study Indicated That Ferro-cement Construction Achieved Approximately 70.53% Reduction In Cement Consumption, 68.15% Reduction In Steel Reinforcement Usage, And 75% Reduction In Structural Volume Compared To RCC Construction. The Analysis Also Revealed That Material Wastage Generated In Ferro-cement Construction Was Significantly Lower, Ranging From 3–5%, Whereas RCC Construction Generated Approximately 8–10% Wastage. Reduced Formwork Requirement, Controlled Mortar Application, And Lightweight Reinforcement Systems Contributed To Better Material Utilization And Lower Construction Debris Generation. The Study Concludes That Ferro-cement Technology Provides Significant Environmental And Sustainability Benefits And Can Serve As An Effective Alternative Construction Technique For Low-cost Housing, Prefabricated Structures, And Sustainable Construction Applications.

Author: Rohit Shinde | Dr.Sushmaawad | Dr. P. M. Alandkar

Area of research: Computer Engineering

The Proliferation Of Digital Technologies Has Catalyzed A Fundamental Transformation Within The Public Transportation Sector, Compelling The Replacement Of Conventional, Counter-dependent Ticketing Processes With Scalable, Web-enabled Reservation Platforms. Passengers Now Demand The Ability To Search Routes, Compare Schedules, Choose Seats, And Complete Payments Through Accessible Online Interfaces—without The Inconvenience Of Physical Queues Or Restricted Service Hours. This Paper Presents The Design, Architecture, And Implementation Of BookMyRoute, A Web-based Bus Ticket Reservation Portal Engineered To Deliver An Intuitive User Experience Alongside Core Capabilities Including Real-time Seat Availability Management, Integrated Payment Processing, And Comprehensive Booking Administration. The Platform Is Constructed Upon A Modern, Three-tier Technology Stack Comprising ReactJS For The Presentation Layer, Spring Boot For Application Logic, And MySQL For Relational Data Management, With Deployment Hosted On AWS Cloud Infrastructure To Ensure High Availability And Horizontal Scalability. A Critical Review Of Existing Solutions And Implementation Challenges Reveals Several Differentiating Advantages Of This Approach—namely Operational Efficiency, Improved Passenger Satisfaction, And Optimized Capacity Utilization. The Study Additionally Outlines Prospective Enhancement Directions, Including Artificial Intelligence-driven Demand Forecasting, Native Mobile Application Support, And Real-time Vehicle Tracking Integration. Evaluation Outcomes Validate That BookMyRoute Successfully Bridges The Gap Between Passenger Convenience And Operational Efficiency, Demonstrating The Transformative Potential Of Purpose-built Digital Platforms In Modernizing Public Transportation Services.

Author: Sahil Patil | Gayatri Lonkar | Shruti Kasbe | Dr. Shubhangi R. Patil | Prof. Shreyas S. Shinde

Area of research: Civil Engineering

The Structural Performance Of Bridge Deck Slab Systems Is Strongly Influenced By Deck Geometry, Load Transfer Mechanism, Stiffness Distribution, And Stress Concentration Under Vehicular Loading. The Present Study Investigates The Comparative Behaviour Of Different Deck Slab Configurations In An MNB Bridge Using ANSYS Finite Element Software. Three Bridge Deck Systems Are Considered: RCC T-beam Bridge, Rectangular Box Girder Bridge, And Trapezoidal Box Girder Bridge. The Models Are Analysed Under IRC Class A And IRC Class AA Loading Conditions. The Main Response Parameters Considered In The Study Are Shrinkage, Equivalent Creep, Normal Creep, Shear Creep, Equivalent Stress, Normal Stress, And Shear Stress. The Bridge Models Are Developed With Identical Span And Width Conditions To Obtain A Rational Comparison Between The Selected Deck Slab Systems. The Results Show That Deck Slab Geometry Has A Considerable Effect On Stress Distribution And Time-dependent Deformation Behaviour. Under Class A Loading, The Trapezoidal Box Girder Bridge Shows Higher Shrinkage, Creep, Equivalent Stress, And Shear Stress Compared With The Rectangular Box Girder And RCC T-beam Bridge. Under Class AA Loading, The Trapezoidal Box Girder Still Shows Higher Shrinkage And Creep, While The Rectangular Box Girder Shows Higher Equivalent Stress And Shear Stress. The Study Confirms That ANSYS-based Finite Element Modelling Is Effective For Understanding Bridge Deck Behaviour And For Selecting A Suitable Bridge Deck Configuration During The Preliminary And Analytical Design Stage.

Author: Robin Pandita | prof. Abhijeet Undre | Dr. Atul Pujari

Area of research: Information Technology

CivicCMS Is An AI-integrated Civic Complaint Management Platform Built On Spring Boot 3.2, MySQL, And The Claude AI API. It Enables Citizens To Report Civic Problems Such As Potholes, Water Leaks, Power Outages, Garbage Overflow, Drainage Blockages, And Street-light Failures Through An Intelligent Chatbot Interface. The System Automatically Classifies Each Complaint Into A Category, Assigns It To The Correct Government Department, Detects Geo-based Duplicate Complaints, And Enforces Service Level Agreement (SLA) Deadlines With Automated Escalation. Department Heads Manage Their Assigned Complaints Through A Dedicated Portal, While Administrators Monitor All Activity, Chaos Alerts, And Analytics Through A Real-time Dashboard. The Platform Supports Multi-language Output In English, Tamil, And Hindi, Offers Google OAuth And OTP-based Authentication, And Provides A Public Feedback Wall For Community Transparency. Experimental Evaluation Demonstrates Consistent Complaint Classification Accuracy And Measurable Reduction In Resolution Turnaround Time.

Author: Gulzar Begam J | Baranidharan ASB | Deepa kumar S | Gokuldharsan S

Area of research: Computer Science And Engineering

The Competitive Employment Landscape Demands Not Only Technical Proficiency From Candidates But Also Confident Communication, Composed Behavior, And Structured Articulation During Interviews. A Significant Proportion Of Job Seekers Underperform In Interviews Due To Limited Access To Realistic Practice Environments And The Absence Of Personalized, Real-time Feedback. This Paper Presents The Design And Implementation Of An Intelligent Virtual Interview Preparation & Skill Evaluation System (IVIPSES)—a Web-based, AI-powered Platform That Simulates Authentic Interview Scenarios And Delivers Comprehensive Multi-modal Performance Assessment. The System Integrates Quork AI For Adaptive, Profile-driven Question Generation; Whisper ASR For High-accuracy Speech-to-text Transcription; Natural Language Processing (NLP) For Evaluating Verbal Response Quality Across Dimensions Of Relevance, Fluency, Grammar, And Keyword Coverage; And Computer Vision-based Behavioral Analytics To Assess Non-verbal Cues Including Eye Contact, Facial Expressions, And Confidence Indicators. The Technology Stack Comprises Next.js (frontend), Node.js (backend), And Firebase (real-time Cloud Database). Functional Validation Confirms End-to-end Operational Effectiveness Across All System Modules, With The Integrated Multi-modal Pipeline Demonstrating Superior Assessment Breadth Compared To Existing Single-dimension Interview Tools.

Author: Pratik Pravin Bodake | Ankit Surendra Swami | Vaibhav Keshav Magar

Area of research: Cyber Security

The AI-Based Cyber Threat Prediction System Improves Cybersecurity Using Artificial Intelligence, Machine Learning, And Deep Learning Techniques. The System Continuously Monitors Network Traffic, System Logs, And User Activities To Identify Suspicious Behavior And Predict Cyber Threats In Real Time. CNN, RNN, And LSTM Models Are Used For Anomaly Detection And Threat Prediction With Improved Accuracy.

Author: Mrs.P.Uma Maheswari | Lekhashreel | Mahalakshmi S | Sreenithi V

Area of research: Structural Design

This Report Covers The Structural Analysis, Design, And Planning Of A G+3 Multi-storey Hospital Building In [Shirpur], Designed To Provide Good Healthcare Services. The Plan Includes Important Areas Like Outpatient Clinics, Patient Rooms, Emergency Rooms, Operating Theaters, Labs, And Offices. The Layout Makes It Easy To Move Around, Keeps Patients Safe, And Follows The National Building Code And Hospital Design Rules. The Structure Was Analyzed Using STAAD Pro And Drawn With AutoCAD. The Building Is Built To Handle Its Own Weight, People, Wind, And Earthquakes, According To IS 875 And IS 1893 Standards. Reinforced Concrete Is Used For The Main Structure, Making It Strong, Stable, And Long-lasting. Fire Safety, Natural Light, Air Flow, Space For Future Growth, And Ways To Move Up And Down The Building Are Also Included. The Approach Focuses On Making Sure The Hospital Works Well, Is Safe, Doesn’t Cost Too Much, And Is Good For The Environment. This Leads To A Complete And Realistic Design For A Modern Hospital.

Author: Girase Unmesh Sudamsing | Mali Ganesh Prakash | Rajput Chintan Hitendra | Nikum Darshan Pravin | Otari Keshav Dagdu | Dr. Mahesh. N. Patil

Area of research: Civil Engineering

Highway Bridges Are Important Elements Of Transportation Infrastructure, And The Selection Of A Suitable Bridge Superstructure Directly Affects Structural Safety, Serviceability, Durability, And Economy. In Indian Bridge Construction, Reinforced Cement Concrete (RCC) Deck Bridges And Prestressed Concrete (PSC) Girder Bridges Are Commonly Used For Small And Medium-span Applications. The Present Research Paper Focuses On The Comparative Structural Analysis Of An RCC Deck Bridge And A PSC Girder Bridge Using STAAD-Pro Under IRC Loading Conditions. A Two-span Continuous Bridge Of 40 M + 40 M Span Arrangement With A Total Length Of 80 M, 10.5 M Carriageway Width, 250 Mm Deck Slab, And Three Longitudinal Girders Is Considered For Analysis. Both RCC And PSC Models Are Developed With Identical Geometry, Support Conditions, And Loading Arrangement To Ensure A Fair Comparison. Dead Load, Superimposed Dead Load, IRC Class AA Tracked Vehicle Load, And IRC Class A Wheeled Vehicle Load Are Applied Using STAAD-Pro Moving Load Generation. The PSC Model Includes Prestressing Force Applied To The Longitudinal Girders To Study The Effect Of Post-tensioning. The Structural Responses Are Compared In Terms Of Bending Moment, Shear Force, Support Reaction, Plate Stress, And Mid-span Deflection. The Results Indicate That The PSC Girder Bridge Provides Better Serviceability, Reduced Tensile Stress, And Improved Deflection Control Compared With The RCC Bridge. Therefore, PSC Is Found More Suitable For The Selected Medium-span Continuous Bridge Configuration.

Author: Gorale Saptarshi Ram1 | Gorale Saptarshi Ram

Area of research: Civil Engineering

This Research Investigates The Analysis And Structural Design Of A G+9 Reinforced Concrete Residential Building By Utilizing Integrated Engineering Software Tools, Including AutoCAD, ETABS, And RCDC. The Structure Is Subjected To Different Types Of Loads, Namely Dead, Live, Wind, And Seismic Loads, In Compliance With The Recommendations Of IS 875 And IS 1893:2016. The Design Of Essential Structural Members Such As Slabs, Beams, Columns, And Footings Is Performed In Accordance With IS 456:2000, Adopting The Limit State Method. The Study Highlights That The Use Of Integrated Software Systems Leads To Improved Computational Accuracy, Reduces Dependence On Manual Calculations, And Increases Overall Productivity In The Design Of Multi-storey Reinforced Concrete Structures.

Author: Aaditya Patil | Amansing Jadhav | Aakash Pawar

Area of research: Artificial Intelligence And Data Science

Autonomous Human Violence Detection Has Become An Essential Component In Modern Smart Surveillance Systems Due To The Increasing Demand For Public Safety And Automated Monitoring. Traditional Surveillance Methods Rely Heavily On Continuous Human Observation, Which Is Inefficient And Prone To Errors When Monitoring Multiple Video Streams For Long Durations. To Address This Limitation, This Paper Presents An Autonomous Human Violence Detection System For Smart Surveillance Using Deep Learning And Real-time Video Analysis. The Proposed System Uses A Laptop Camera To Capture Live Video And Analyzes Human Activities Using A Trained Deep Learning Model Developed From Two Different Datasets Consisting Of Normal Human Actions And Violent Activities. Convolutional Neural Networks (CNN) And Object Detection Techniques Are Used To Extract Features From Video Frames And Classify The Actions As Normal Or Violent. The Trained Model Compares Live Video Input With Learned Patterns And Generates Real-time Predictions To Identify Suspicious Behavior. Experimental Results Show That The Proposed System Can Detect Violent Activities With Good Accuracy And Can Be Applied In Public Surveillance, Educational Institutions, And Security Monitoring Environments, Reducing The Need For Continuous Human Supervision And Improving The Efficiency Of Intelligent Surveillance Systems.

Author: Vennila P | Sanjeev babu M | Anbumani M | Yogesh R

Area of research: Computer Science And Engineering

Smart Community & Service Hub Is A Mobile-based Application Developed To Provide A Secure, Efficient, And Digitally Connected Platform For Residential Society Management. Traditional Society Management Systems Are Often Manual, Time-consuming, And Lack Transparency. The Proposed System Introduces A Centralized Solution That Integrates Essential Services Such As Notice Management, Complaint Handling, Service Booking, Maintenance Tracking, QR-based Visitor Access, And SOS Emergency Alerts. The System Supports Dual Access For Residents And Administrators, Enabling Seamless Communication And Efficient Management. The Application Is Developed Using Flutter For The Frontend, Node.js With Express.js For Backend Services, And MongoDB For Data Storage. The Integration Of QR-based Security And Real-time Alert Systems Enhances Safety And Operational Efficiency. The System Simplifies Daily Operations, Improves Transparency, And Promotes A Smart And Connected Living Environment.

Author: Sakshi Arvind Patil | Akanksha Kishor Kurale | Vaishnavi Vilas Jadhav

Area of research: Computer Science

Agriculture Plays A Vital Role In India’s Economy, Yet Farmers Face Multiple Challenges Such As Lack Of Real-time Information, Poor Access To Markets, And Limited Technological Support. This Paper Presents Krishi AI, An Intelligent AI-based Chatbot Designed To Provide Agricultural Advisory Services, Real-time Market Insights, And Multilingual Interaction. The System Integrates Artificial Intelligence, Natural Language Processing, And Location-based Services To Enhance Decision-making For Farmers. It Aims To Bridge The Gap Between Traditional Farming Practices And Modern Digital Solutions By Offering A User-friendly Platform That Supports Both Online And Offline Functionalities. The Proposed System Acts As An Intelligent Virtual Assistant Capable Of Responding To Farmers’ Queries Related To Crop Selection, Soil Health, Weather Forecasts, Pest Control, Irrigation Techniques, And Market Prices (APMC Rates). The Chatbot Is Designed With A User-friendly Interface Supporting Regional Languages To Ensure Accessibility For Rural Users With Varying Literacy Levels. It Also Integrates Location-based Services To Deliver Personalized Recommendations Based On State, District, And Local Agricultural Conditions. The System Architecture Combines Machine Learning-based Intent Classification With A Knowledge Base Of Agricultural Data Sourced From Government Portals And Agricultural Research Institutions. Additionally, The Chatbot Can Be Deployed On Mobile And Web Platforms, Ensuring Wide Accessibility. The Expected Outcome Of This System Is To Enhance Decision-making Efficiency For Farmers, Reduce Dependency On Intermediaries, And Promote Smart Agriculture Practices. By Bridging The Gap Between Modern Agricultural Knowledge And Rural Farmers, The AI-based Chatbot Aims To Contribute Toward Sustainable Farming And Improved Crop Productivity. Keywords — Artificial Intelligence, Chatbot, Smart Agriculture, Natural Language Processing, Farmer Assistance, Machine Learning, APMC, Precision Farming.

Author: Dhanashri Dhondiram Maske | Sanjana Rajendra Misal | Sambodhi Pradeep Kamble

Area of research: Computer Acience

Predicting Price Of Any Stock Is Difficult To Achieve Because Of Volatility Of Market Conditions. In Fraction Of Seconds Markets Go Up And Down, Fluctuations And Mood Of The Investor These Makes Traditional Methods Less Accurate To Predict. Normal Statistical Methods Do Not Detect Small Changes In Stock Market Even Though That Are Important For Understanding The Market And Many Of The Times Single Ml Model Become Unstable With Market Changes. In This Study, We Tried A Prediction Method That Combines The ‘Long Term Short Memory’ Network With The Gradient Boosting Procedure (XG Boost). Long Term Short Memory (LSTM) Helps To Understand How The Price Changes With The Time And It Finds The Underlying Timely Based Pattern In Past Data(historical) Related To Stocks. These Observations, Combined With A Group Of Produced Market Measures Are Given Or Passes To The XG Boost, Which Helps Show Patterns And Irregular Movements That The LSTM Model Alone Skips Or Failed To Notice. The Objective Of This Mixed Approach Is To Form Predictions That Remain Accurate In Various Or Changing Conditions Of The Market And Give More Accurate Changing Pattern(nature) Of Stock Data. Tests Conducted On NIFTY 50 Index And Various NSE- Listed Stocks Reveal That The Integrated Approach Outperforms Isolated LSTM, XG Boost, And Baseline Models, As Measured By Lower RMSE And MAE Values, Underscoring Its Reliability And Real-world Potential.

Author: Krishna Mittal | M Akhil | Mohammed Bande Nawaz | Mohammed Mujeeb

Area of research: Civil Engineering

Road Under Bridges (RUBs) Are Critical Infrastructure For Improving Traffic Safety And Continuity, Particularly Where Level Crossings Disrupt Smooth Flow Of Rail Or Road Traffic. Conventional Open-cut Construction Of RUBs Often Causes Significant Traffic Disruption, Temporary Embankment Instability, And High Construction Costs. The Box Pushing Technique Offers A Trenchless Alternative, Wherein A Reinforced Concrete (RCC) Box Is Fabricated Adjacent To The Embankment And Hydraulically Pushed Through The Soil While Excavation Is Performed At The Leading Face. This Method Minimizes Surface Disturbance And Allows Traffic To Remain Operational During Construction. The Present Study Evaluates The Structural Behavior And Construction Feasibility Of A Single-cell RCC Box (6.0 M × 4.5 M Clear Opening) Using ANSYS Finite Element Analysis. Detailed Engineering Calculations, Including Hydraulic Jack Capacity, Soil-structure Interaction, Earth And Surcharge Loads, And Lateral Pressures, Were Incorporated Into The Model. Simulation Results Indicate A Maximum Equivalent Stress Of 18.7 MPa, Well Below The M35 Concrete Compressive Strength, And Total Deformation Ranging From 1.8 Mm To 4.6 Mm, With Vertical Deflection Dominating Due To Soil Cover And Live Load Surcharge. The Required Installed Jacking Capacity Of 12,000 KN Ensures Safe And Controlled Advancement Of The Box, Accounting For Base Friction (6,693 KN) And Face Resistance (1,800 KN). Safety Evaluation Confirms A Factor Of Safety Of 1.5, Validating The Design Under Both Construction-stage And Service-stage Loads. The Study Demonstrates That The Box Pushing Technique, Combined With FEM Analysis, Allows Efficient, Safe, And Minimally Disruptive RUB Construction. Recommendations Include Careful Jack Alignment, Staged Excavation, And Monitoring Of Settlement And Stress To Optimize Safety And Performance In Urban And Railway Environments.

Author: Shravya Shetty | prof. Abhijeet Undre | Dr. Atul Pujari

Area of research: Civil Engineering

This Study Investigates The Use Of Eggshell Powder (ESP) As A Partial Replacement For Cement In M30 Grade Concrete. With The Increasing Demand For Sustainable Construction Materials, ESP Offers An Eco-friendly Alternative Due To Its High Calcium Carbonate Content, Which Is Chemically Similar To Cement. The Primary Objective Of This Research Is To Evaluate The Impact Of Varying ESP Replacement Levels (5%, 7.5%, And 10%) On The Mechanical Properties And Durability Of M30 Concrete. The Study Assesses Compressive Strength, Water Absorption, Chloride Penetration, And Acid Resistance At Different Curing Periods (7, 14, And 28 Days). Results Show That ESP Enhances Both The Strength And Durability Of The Concrete, With 7.5% ESP Replacement Providing The Best Balance Of These Properties. The Findings Confirm That ESP Can Significantly Reduce The Environmental Impact Of Concrete Production By Replacing Cement, Which Is A Major Contributor To Carbon Emissions. This Study Provides A Deeper Understanding Of The Role Of ESP In Concrete, Offering Insights For Sustainable Construction Practices And Waste Material Utilization In The Industry.

Author: Akshay Laxman Kamthe | prof. Abhijeet Undre | Dr. Atul Pujari

Area of research: Civil Engineering

High-rise Buildings In Seismic Zones Are Vulnerable To Significant Damage During Earthquakes Due To Their Height And Stiffness. To Mitigate Seismic Risks, Base Isolation Has Emerged As A Vital Engineering Solution. This Research Aims To Evaluate The Seismic Performance Of High-rise Buildings With Varying Aspect Ratios, Both With And Without Base Isolation. The Study Focuses On G+30 And G+40 Storey Buildings Modeled In ETABS 2016, Considering Aspect Ratios Ranging From 0.25 To 2.0. The Research Compares Three Building Shapes: Square, T-shape, And C-shape, To Assess The Effectiveness Of Base Isolation In Reducing Seismic-induced Displacement, Drift, And Shear Forces. Using A Combination Of Response Spectrum And Time History Analysis, The Study Simulates Earthquake Forces Using Real Seismic Records (Bhuj Earthquake). Base Isolation, Implemented With Lead Rubber Bearings (LRB), Is Used To Decouple The Structure From Ground Motion, Thereby Reducing Seismic Forces Transmitted To The Building. The Results Show That Base Isolation Significantly Reduces Lateral Displacement And Drift, With Improvements Ranging From 30% To 60%, Depending On The Building Shape And Aspect Ratio. Additionally, Base Shear Is Reduced By Up To 60%, Demonstrating The System’s Efficiency In Minimizing Lateral Forces. The Study Concludes That Base Isolation Is Highly Effective In Enhancing The Seismic Performance Of High-rise Buildings, Particularly Those With Larger Aspect Ratios And Irregular Geometries. This Research Provides Valuable Insights Into The Application Of Base Isolation In Seismic Design And Offers Recommendations For Improving Building Resilience In Earthquake-prone Regions.

Author: Nikhil Uday Jadhav | prof. Abhijeet Undre | Dr. Atul Pujari

Area of research: Artificial Intelligence And Data Science

Data Leakage Is One Of The Most Critical Challenges In Machine Learning Systems, Leading To Unrealistic Model Performance And Poor Generalization In Real-world Applications. Leakage Occurs When Information From Outside The Training Dataset Is Inadvertently Used During The Model Training Process, Causing Biased Predictions And Overly Optimistic Evaluation Metrics. Detecting Such Leakage Before Model Development Is Essential For Building Reliable And Robust Machine Learning Systems. This Paper Proposes A Data Leakage Detection And Intelligent Data Preprocessing System That Automatically Identifies Potential Leakage Sources In Datasets Prior To Model Training. The System Integrates Dataset Profiling, Leakage Detection, Preprocessing Techniques, And Visualization Tools Within A Flask-based Web Application. Users Can Upload Datasets, Analyze Data Quality, Detect Different Types Of Leakage Such As Target Leakage And Temporal Leakage, And Apply Safe Preprocessing Operations. The System Also Provides Interactive Data Visualizations And Exports A Cleaned Dataset Ready For Machine Learning Tasks. By Combining Leakage Detection With Automated Preprocessing, The Proposed Solution Improves Model Reliability, Reduces Human Error, And Enhances The Overall Machine Learning Workflow.

Author: Pakirathan K | Murshith Ahamed Eithirish | Gokul | Gokul R

Area of research: Forensic Science

Earlier, Numerous Methods For The Synthesis Of Latent Fingerprints On Numerous Surfaces Were Documented. This Research Describes A Simple And Non-toxic. The Powder Dusting Approach Is The Most Practical Approach To Develop Latent Fingerprints At A Crime Scene. Latent Fingerprint Is Prone To Damage And Destruction Because Of Their Fragile Nature. There Are So Many Powders Present In The Market For Lifting A Fingerprint Which Are Toxic And Expensive. Such Chemical Methods Are Not Impermeable And Permanent On Surface. Replacing Old Methods, We Have Formulated A New Powder With The Help Of Cement And Beetroot Which Is Non- Toxic And Easily Available. This Novel Formulation Is Efficient To Make Latent Fingerprint Permanent And Water Resistant On Different Surface Which Benefits At Outdoor Crime. We Have Tested This On Surfaces Like Wall, Aluminium, Wood, Plastic, Foil-paper, Leather, Glass, Paper Etc.

Author: Arbaz B khan | Sudha Shetty

Area of research: ARTIFICIAL INTELLIGENCE - MACHINE LEARNING : DEEP LEARNING

The Web-Based Plant Disease Detection And Fertilizer Recommendation System Presents A Web-based System For Detecting Plant Leaf Diseases. Users Can Upload Leaf Images Through A Simple Interface For Analysis.An Image Enhancement Module Improves Low-quality Images Before Processing. The Enhanced Images Are Analyzed Using A Deep Learning Model For Accurate Disease Detection. The Model Is Optimized Using Quantization Techniques To Improve Speed And Reduce Computational Complexity. The System Displays The Detected Disease Along With Prediction Confidence And Suitable Fertilizer Recommendations.A Chatbot Module Is Integrated Into The System To Answer User Queries Related To Plant Diseases, Fertilizers, And Crop Care. Overall, The System Supports Early Disease Detection, Improves Agricultural Productivity, And Assists Farmers In Maintaining Healthy Crops.

Author: Dr .P. Shanmuga Priya | Cladias Wilfred L | Krishnan J | Suhail H

Area of research: Computer Science And Engineering

This Paper Presents Smart Medicare Hub, An Ad- Vanced Integrated Hospital Management System Developed Using Flutter And Firebase. The System Automates Hospital Operations By Connecting Departments Such As Patient Registration, Phar- Macy, Laboratory, Billing, And Administration Under One Unified Digital Platform. Leveraging Artificial Intelligence(AI), Machine Learning (ML), And Cloud Computing, It Enables Predictive Ana- Lytics, Real-time Data Access, And Seamless Communication Across All Healthcare Departments. The System Includes A Convolutional Neural Network (CNN)-based Pneumonia Detection Module Inte- Grated Via A Flask REST API. Experimental Results Demonstrate Dashboard Load Times Of 1.2–1.8 Seconds On A Standard 4G Connection, Fulfilling The Stated Performance Requirements. The Proposed System Transforms Traditional Hospital Management Into A Smart, Technology-driven, And Patient-centered Environ- Ment.

Author: Nayana Suresh Mali | Sanika Prakash Thombare | Shravani Keshav Patil | Bharati Shankar Hegade

Area of research: Machine Learning

Lung Diseases Such As Viral Pneumonia And Smoking-induced Lung Damage Are Major Global Health Concerns Responsible For Millions Of Deaths Each Year. Early And Accurate Detection Of These Conditions Is Essential For Timely Medical Intervention And Treatment. This Project Presents A Deep Learning–based Image Classification Model For Automated Identification Of Viral Pneumonia And Lung Damage Caused By Smoking Using Chest X-ray And CT Images. The Proposed System Leverages Transfer Learning With The EfficientNetB0 Architecture, Which Extracts High-level Visual Features From Lung Images And Classifies Them Into Two Categories. The Dataset Is Preprocessed Through Normalization And Image Augmentation To Enhance Generalization And Reduce Overfitting. The Model Is Trained Using Binary Cross-entropy Loss And Optimized With The Adam Optimizer To Achieve High Accuracy And Robustness. Experimental Results Demonstrate The Model’s Capability To Distinguish Between Viral Pneumonia And Smoker-affected Lungs Effectively, Supporting Radiologists In Diagnostic Decision-making. This System Offers A Reliable, Efficient, And Scalable AI-driven Approach For Medical Imaging Analysis And Contributes To The Advancement Of Computer-aided Diagnosis In Pulmonary Healthcare.

Author: Mrs. Lavanya | J. Karthik | G. Sudheer | G. Rohith Yadav

Area of research: Computer Science And Business Systems

This Paper Presents An Internet Of Things (IoT)-based Smart Bin Level Detection System Designed To Improve Waste Management Efficiency. Traditional Waste Collection Methods Rely On Fixed Schedules, Leading To Inefficiencies Such As Overflowing Bins And Unnecessary Fuel Consumption. The Proposed System Uses Ultrasonic Sensors To Monitor Waste Levels In Real Time And Transmits Data To A Cloud Platform. Alerts Are Generated When Bins Reach Threshold Levels, Enabling Optimized Collection Routes. The System Reduces Operational Costs, Improves Cleanliness, And Supports Smart City Development.

Author: Mrs.N. G. Gayathri | Visvasugin N. R | Hariprakash .R | Charukesh.R

Area of research: IT Engineering

Plastic Waste Is An Issue For Our Environment Because It Takes A Long Time To Break Down. It Just Keeps Piling Up In Cities, Rivers And Natural Areas Causing Pollution And Hurting Animals. The Main Goal Of This System Is To Find Waste In Pictures And Videos. It Uses Lots Of Images Of Waste To Teach A Computer Model Called YOLOv8. After Teaching The Model Can Quickly And Accurately Spot Waste In Photos, Videos And Even Live Camera Feeds. When The System Finds Waste It Can Help Sort. Track It. This Makes It Easier To Deal With Waste In A Way And Make Sure It Is Thrown Away In A Way That Is Friendly, To The Environment. Overall This System Helps Find Waste Easily Makes Waste Management Better And Keeps Our Environment Clean.

Author: Prof. A.P.Kulkarni | Omkar Bangar | Dnyaneshwari Gore | Karan Kawale | Sayali Gund

Area of research: Electrical And Electronics Engineering

The Proliferation Of Landmines And Explosive Devices In Conflict Zones Poses A Significant Threat To Human Safety, Necessitating The Development Of Automated Detection Systems. This Paper Presents The Design And Implementation Of A Smart Robotic Platform Integrated With A Metal Detector And A Passive Infrared (PIR) Sensor For The Dual Purpose Of Landmine Detection And Human Presence Monitoring. Controlled Wirelessly Via A Bluetooth Module, The System Utilizes An Arduino Uno Microcontroller To Process Sensor Data And Execute Navigation Commands. The Metal Detector Identifies Metallic Components Common In Explosive Casings, While The PIR Sensor Detects Infrared Radiation Emitted By Living Beings. Upon Detection Of A Threat, The System Triggers An Audible Alarm And Provides Real-time Feedback To The Operator. Experimental Results Demonstrate The Robot’s Effectiveness In Hazardous Environments, Offering A Cost-effective And Reliable Solution For Reducing Human Risk In Surveillance And Demining Operations.

Author: M.Guru Prasath | V.Kalaivanan | R.Kowsihk Prabu | S.Perarasan | Mrs.S.Suganya

Area of research: Bharamapuri

This Abstract Presents The Development And Evaluation Of Herbal Tooth Powders With The Aim Of Providing A Natural Alternative For Oral Care. With Growing Concerns About The Potential Risks Associated With Synthetic Chemicals And Additives Commonly Found In Commercial Toothpastes, The Demand For Herbal And Natural Oral Care Products Is On The Rise. A Blend Of Carefully Selected Botanical Substances With Proven Efficacy In Improving Oral Health Was Used To Create The Herbal Tooth Powder. These Blends Contained The Antibacterial, Anti-inflammatory, And Refreshing Properties Of Herbs. The Powder Formulation Was Created To Ensure Ease Of Use And Efficient Cleaning While Improving General Oral Health. The Evaluation Of Herbal Tooth Powder Included Various Parameters Including Physical Characteristics, Microbial Analysis And Sensory Evaluation. The Physical Properties Of The Powders Was Evaluated To Meet The Desired Specifications. Microbial Analysis Was Performed To Verify The Safety Of The Product And Confirm The Absence Of Harmful Bacteria. Additionally, Sensory Evaluation Was Performed To Assess Factors Such As Taste, Aroma And Overall User Experience.

Author: Monali Devrao Nagrikar | Payal Kuldip Indurkar

Area of research: Computer Science And Engineering

Traditional Automated Guided Vehicle (AGV) Systems Rely On Centralized Controllers And Rigid Programming, Causing Collisions, Torque Overloads, And Operational Downtime When Factory Conditions Change. Nobody Checks If An AGV Mission Is Physically Safe Before Execution—if An Operator Assigns An Impossible Task, The Drivetrain Suffers Damage. We Built BlueFactory Copilot To Solve These Problems. It Is A Lightweight Orchestration Platform For Bonfiglioli-powered AGVs That Runs Four Intelligent Modules In The Background: An LLM-based Natural Language Mission Designer, A Physics-accurate Digital Twin Validator, A Mesh-network Swarm Coordinator, And An IoT-driven Predictive Maintenance Engine. When An Operator Issues A Command, The System Parses Intent Via Meta's Llama-3.3 70B Through Groq Cloud, Simulates The Mission Against Torque/thermal/battery Constraints, Negotiates Paths Peer-to-peer With Other AGVs, And Predicts Mechanical Wear. If A Threat Is Detected—like Torque Overload Or Path Conflict—Copilot Takes Action Autonomously: Flattening Acceleration Curves, Rerouting Via Cooperative A*, Or Scheduling Maintenance. We Tested BlueFactory Copilot Against Dynamic Factory Scenarios That Traditional AGV Controllers Failed To Handle. Our System Reduced Simulated Fleet Collisions By 95%, Cut Unplanned Downtime By 30%, And Maintained Average Inference Latency Under 300ms. The App Uses Zero Local GPU And Only Standard CPU Resources.

Author: Mrs. G. Nandhini | Harish C | Harish V | Hyagreevan G | Elumalai P

Area of research: Internet Of Things (IoT)

The Increasing Demand For Efficient Transportation And Logistics Management Has Led To The Development Of Intelligent Fleet Monitoring Systems. This Paper Presents A Smart Fleet Management System Using Internet Of Things (IoT) Technology To Enable Real-time Tracking, Monitoring, And Analysis Of Vehicle Operations. The Proposed System Integrates GPS Modules, Onboard Sensors, And An ESP32 Microcontroller To Collect And Transmit Data Such As Vehicle Location, Speed, Fuel Level, And Driver Behavior.The Collected Data Is Sent To A Cloud-based Platform For Storage, Processing, And Visualization Through A User-friendly Dashboard. The System Provides Real-time Alerts For Abnormal Conditions Such As Over Speeding, Fuel Theft, And Unauthorized Vehicle Usage. It Improves Operational Efficiency, Enhances Safety, And Reduces Maintenance Costs. The Proposed System Is Cost-effective, Scalable, And Suitable For Modern Smart Transportation Systems.

Author: Mrs.M.Saranya | Sowmiya A | Roshan K H | Sridhar T | Vetrivel R

Area of research: Mechanical Engineering

Industrial Automation Has Become An Important Part Of Modern Manufacturing Systems To Improve Productivity, Precision, And Efficiency. This Paper Presents The Design And Fabrication Of A 5-Degree Of Freedom (5-DOF) Robotic Arm For Pick-and-place Operations. The Robotic Arm Is Designed To Automate Repetitive Material Handling Tasks Commonly Performed In Industrial Environments. The System Consists Of A Mechanical Arm Structure, Servo Motors, Gripper Mechanism, Arduino Uno Microcontroller, And Control Circuitry. CAD Modelling Of The Robotic Arm Was Carried Out Using CATIA Software, And The Components Were Fabricated Using 3D Printing Technology With ABS Plastic Material. Kinematic Analysis And Torque Calculations Were Performed To Ensure Proper Movement And Stability Of The Robotic Arm. The Developed Robotic Arm Is Capable Of Accurately Picking And Placing Lightweight Objects Within A Specified Workspace. The System Provides A Cost-effective And Flexible Solution For Small-scale Industrial Automation And Educational Applications.

Author: Abhijeet Nikam | Deep Parab | Yash Salunke | Jidnyasa Magar

Area of research: Computer Science And Engineering

Construction Sites Are Hazardous Environments For Anyone Working Within Them With Various Dangers Present Due To The Presence Of Heavy Machinery, Unsafe Working Practices, And Inadequate Safety Measures. Personal Protective Equipment (PPE) Like Hardhats, Safety Vests, Gloves, Boots, And Masks Can Be Used To Minimize Injuries And Accidents. It May Be Challenging Manually To Monitor The Compliance Of Workers Regarding Their Adherence To Wearing PPE Since There Are Many Individuals At A Construction Site, And Supervision May Not Be Feasible. The Aim Of This Project Is To Design An Intelligent PPE Detection System Using An Enhanced YOLOv11 Deep Learning Model To Analyze Real-time Data From Cameras At A Construction Site To Verify If Workers Are Wearing Appropriate PPE. The Intelligent PPE Detection System Will Detect Various PPEs Worn By Individual Workers In Real-time And Classify Workers Based On Whether They Comply With PPE Usage Safety Guidelines Or Not. In Case Of Non-compliance By Any Individual, The System Will Automatically Alert The Respective Supervisor Through SMS And Notifications. This Particular System Has Been Created In Such A Way That It Will Work Under Difficult Circumstances That Are Usually Found On Construction Sites, Including Poor Lighting, Crowded Environments, And Partially Blocked Lines Of Sight. Through This Particular System, Continuous And Automatic Monitoring Of Compliance Regarding PPE Will Be Made Possible, Which Will Lead To Improved Safety In The Workplace, Ensuring Regulatory Compliance, Minimizing Workplace Accidents, And Improving Safety On The Construction Site.

Author: Lavanya E | Menaka R | Deepika S | Aruna Devi R | Mrs.M.Agalya

Area of research: Computer Sceince And Engineering

Agricultural Crop Damage Due To Herbivorous Animal Intrusion Causes Annual Losses Exceeding USD 2.5 Billion In India Alone. This Paper Presents An IoT-based Smart Herbivore Detection And Deterrence System That Autonomously Monitors An Agricultural Field Boundary Using A Passive Infrared (PIR) Motion Sensor, Activates An Acoustic Buzzer Deterrent With An Irregular Non-repetitive Alarm Pattern, Drives An L298N Motor Module To Animate A Physical Scare Mechanism, Captures Evidence Images Using An OV7670 CMOS Camera, And Persists All Event Records To A Micro SD Card — All On An Arduino Uno (ATmega328P) Microcontroller Platform. Laboratory Validation Across 300 Trials Confirmed An Overall PIR Detection Rate Of 91.2%, Deterrence Response Latency Under 80 Ms, And SD Card Logging Reliability Of 99.8% Over 8-hour Endurance Tests. The System Is Deployable From A USB Power Bank With A Total Bill Of Materials Cost Under INR 1,800.

Author: Pooja K | Priyadharshini G | Sowmiya K | Suganya B

Area of research: Computer Science And Engineering

Gastrointestinal (GI) Diseases Such As Polyps, Ulcers, And Tumors Require Early Detection For Effective Treatment. Traditional Manual Inspection Of Endoscopy Images Is Timeconsuming And Prone To Human Error. This Project Proposes An AIbased System Using ResNet50 For Automatic Multiclass Classification (Normal, Polyp, Ulcer, Tumor) Of GI Diseases From Endoscopy Images. A Comprehensive Comparative Analysis Of VGG16, MobileNetV2, EfficientNetB0, And ResNet50 Is Conducted On The Kvasir Dataset (8,000 Images). ResNet50 Achieves The Highest Performance: 94.2% Accuracy, 93.8% Precision, 94.1% Recall, And 93.9% F1score, Outperforming VGG16 (89.4%), MobileNetV2 (86.7%), And EfficientNetB0 (91.3%). The System Reduces Diagnosis Time From 8–10 Minutes Per Patient To Under 2 Seconds Per Image, Improving Diagnostic Consistency And Early Detection.

Author: Sharmilaa S | Udhayakumari S | Sona M | Vishnu Devi R

Area of research: Mechanical Engineering

Waste Cooking Oil Biodiesel Blends Are A Low-cost, Environmentally Friendly, And Cost-effective Alternative To Conventional Combustion Fuels. In This Work, The Effects Of Oxygen Enriched Combustion On Performance, Combustion, And Emission Characteristics Of A Combustion Engine Are Investigated. The Results Indicate That Increasing Oxygen Concentration Generally Enhances Combustion Efficiency, Reduces Incomplete Combustion Products (CO, HC), And Lowers Particulate Emissions, But Often Increases NOx Emissions Due To Higher Flame Temperatures. This Work Provides A Practical Combustion Control Approach That Can Be Further Developed For Industrial Waste Oil Burners, Thermal Systems And Emission Control Applications.

Author: Gaurav Pandit | Nikhil Ujjainkar | Rushikesh Ithape | Vaibhavi W. | Dr. D.M. Mate Sir

Area of research: Computer Science And Engineering

With The Rise In Internet Usage, The Dangers Brought About By Malicious URLs And Phishing Websites Have Also Escalated, Thus The Need For Internet Security Becoming Paramount. Existing Solutions To Detect Malicious URL/website Include The Blacklisting Method And The Rule-based Technique, Which May Not Be Efficient In Detecting New Attacks. This Research Aims At Proposing A Privacy-Preserving Safe Browsing (PPSB) Scheme, Which Is An Amalgamation Of Machine Learning And Cryptographic Technology, With The Objective Of Increasing Both Security And Privacy Of The Users. The SVM Algorithm Is Used To Detect Malicious Websites Through Classification Based On The Features Extracted. In Addition To Detection Precision, The Suggested Framework Stresses Its Capability For Offering A High Level Of Privacy Guarantees With The Integration Of AES Encryption In Order To Protect User Searching Histories And Browsing Habits. In Such A Way, Sensitive Data Are Guaranteed To Be Safe From Any Third-party Analysts As Well As Service Operators. Moreover, The Framework Includes A Dynamic Blacklisting Function, Which Allows Users To Update And Manage The Blacklisted Websites On An Ongoing Basis. Finally, It Offers The Possibility Of Selectively Aggregating The Information About The Users' Activities While Providing Differential Privacy Guarantees.

Author: Ms.U.SathyaM.E | K.Athisri | A.Abinaya | A.Archana | G.K.Bavidra

Area of research: Computer Science And Engineering

The Accurate Forecasting Of Stock Prices Remains One Of The Most Persistent And Challenging Problems In Quantitative Finance, Owing To The Inherently Noisy, Nonstationary, And Nonlinear Nature Of Financial Time Series. Traditional Forecasting Methods Struggle With The Dual Challenges Of Capturing Both Linear Periodicities And Nonlinear Behavioral Dynamics Within A Unified Framework. This Paper Presents A Combined Approach To Enhance Stock Price Forecasting Accuracy. The Proposed Model Integrates Three Key Components: Autoregressive Integrated Moving Average (ARIMA) For Effectively Modeling Linear Patterns And Repeating Cycles In Stock Price Data; Long ShortTerm Memory (LSTM) Networks For Identifying Complex, Longterm Dependencies In Stock Price Movements; And Market News Sentiment Analysis To Capture Investor Behavior And Emotional Influences. By Synergistically Combining Statistical Timeseries Modeling, Deep Learning, And Sentimentdriven Insights, The Approach Aims To Overcome The Limitations Of Individual Methods. Experimental Evaluation On Three Largecap Stocks (AAPL, JPM, TSLA) Over A 12month Testing Period Demonstrates That The Hybrid Model Achieves A 41.9% Reduction In RMSE Compared To Standalone ARIMA And A 30.1% Reduction Compared To Standalone LSTM, With Directional Accuracy Improving From 52.9% And 59.9% To69.5%. The Hybrid Framework Provides More Robust And Precise Stock Price Predictions, Supporting Better Informed Financial Decision Making.

Author: Reshma RB | Swathika P | Sackcini M | Saadana R

Area of research: Computer Science And Engineering

Silent Performance Degradation In Network Systems Is A Critical Operational Challenge Where Servers Gradually Lose Efficiency Without Explicit Failures Or Alarms, Leading To Reduced Reliability, Increased Downtime Risk, And Higher Maintenance Costs. Existing Monitoring Systems Depend On Threshold Alerts And Reactive Logs, Missing Gradual Performance Drift And Producing False Alarms During Spikes. They Also Lack Interpretability And Do Not Offer Clear Root-cause Insights Or Actionable Guidance. This Project Proposes A Machine Learning–based Predictive Analysis Framework For The Early Detection Of Such Hidden Degradation Using Multivariate System Metrics Including CPU Usage, Memory Consumption, Disk I/O Wait, Network Latency, And Process Behavior Collected As Continuous Time-series Data. The Core Detection Engine Employs OmniAnomaly (Variational Autoencoder With GRU) To Learn Normal Temporal Behavior And Joint Distribution Of System Metrics In An Unsupervised Manner. Instead Of Relying On Labeled Failure Data, The Model Identifies Subtle Deviations Through Rising Reconstruction Error, Enabling Early Identification Of Gradual Performance Drift. To Make Predictions Interpretable, A SHAP-based Explainability Layer Quantifies Each Metric’s Contribution To The Anomaly Score, Revealing The Root Cause Of Degradation. A Rule-driven Recommendation Engine Maps Explainable Causes To Precise Corrective Actions. An Intelligent Alert System Validates Anomalies Using Adaptive Thresholds, Drift Trend Analysis, And SHAP Consistency Checks Before Issuing Multi-level Notifications. This Approach Enables Administrators To Detect, Understand, And Resolve Silent Performance Issues Proactively, Improving System Stability, Uptime, And Operational Efficiency.

Author: Padma Nivedha M | Monisha M | Praddeepa R.K | Sasikumari S | Sowmiya P

Area of research: Artificial Intelligence And Machine Learning

Financial Firms Must Embed AI Governance And Risk Management Up Front, Treating AI Not As An Isolated Project But As Integral To Enterprise Risk And Compliance Frameworks. Recent Regulatory Signals (Treasury/FSSCC In The U.S., APRA In Australia, EBA In The EU, MAS In Singapore) Emphasize Integrating AI Oversight Into Existing Structures. Industry Studies Likewise Report That Banks With Centralized, Well-defined AI Governance Outperform Peers. This Paper Surveys Regulatory Guidance (e.g., NIST AI RMF, EU AI Act Mapping, PRA/FCA Feedback), Industry Frameworks (FSSCC’s FS AI RMF, FINOS AI Governance Framework), And Case Studies (JPMorgan’s Platform Approach, Industry Consortia) To Distill Three Governance-first Approaches: (1) Centralized AI Governance Office/CoE, (2) Federated/Embedded Governance, (3) Platform-Centric (Pipeline) Governance. For Each, We Outline The Structure, Processes, Roles, Pros/cons, Investment, And Maturity, And Map Suitability To Enterprise Personas (e.g., Global Banks Vs. Fintechs, CIO Vs. CRO/CDO). We Then Propose A Reusable “governance-first” Framework And Step-by-step Implementation Roadmap (including Templates/checklists, KPIs/KRIs, Monitoring/audit Plans, Third-party Controls, And Change Management). Illustrative Case Studies (e.g., JPMorgan Chase, FINOS Consortium, A Hypothetical Regional Bank) Highlight Practical Trade-offs. We Conclude With Recommended Approaches For Different Organizational Types. All Statements Are Supported By Recent Regulatory And Industry Sources.

Author: Dr. Rajan Nagarajan

Area of research: Artificial Intelligence And Data Science

The Rapid Advancement Of Deep Learning And Generative Models, Particularly Generative Adversarial Networks (GANs), Has Enabled The Creation Of Highly Realistic Synthetic Media Known As Deepfakes. These Manipulated Media Forms, Including Images, Videos, Audio, And Text, Pose Significant Threats To Digital Trust, Cybersecurity, And Social Stability. Deepfakes Are Increasingly Used For Misinformation Campaigns, Identity Theft, Political Manipulation, And Financial Fraud, Making Their Detection A Critical Research Challenge. This Paper Proposes A Multimodal Deepfake Detection System That Integrates Advanced Artificial Intelligence Techniques To Analyze And Classify Content Across Multiple Data Modalities. The System Employs BERT-based Natural Language Processing (NLP) For Text Analysis, Convolutional Neural Networks (CNNs) For Image And Audio Classification, And Long Short-Term Memory (LSTM) Networks For Temporal Video Analysis. The Proposed System Is Evaluated Using Benchmark Datasets Such As Celeb-DF, FaceForensics++, And ASVspoof, Achieving High Accuracy Across All Modalities. Furthermore, The System Is Implemented As A Web-based Platform That Enables Real-time Detection Of Deepfake Content. The Results Demonstrate That The Proposed Approach Significantly Improves Detection Performance And Provides A Scalable Solution For Combating Misinformation In Digital Ecosystems.

Author: Ms. S. Mahalakshmi | C. Akash | R. Vasanthakumar | R. Kalyanamoorthy | M. Sanjai

Area of research: Computer Science And Engineering

Card-based ATM Authentication Has Been A Primary Target For Skimming, Cloning, And PIN Theft Attacks, Resulting In Billions Of Dollars In Losses For Account Holders And Financial Institutions Every Year. This Paper Presents A Secure Cardless Withdrawal System That Eliminates The Physical Card And Routes Every Transaction Through A Two-step Authentication Process Based On User Type. Local Citizens Authenticate Using Their Face, Captured Live By The ATM Webcam, Along With A Six-digit One-time Password Sent To Their Registered Mobile Number Via The Twilio SMS Gateway. Foreign Visitors Undergo The Same Face Capture And Mobile OTP Steps, Along With An Additional Verification Of Their Passport Number. This Matches The Submitted Document Identifier Against The Credentials Stored In The Bank Database At The Time Of Registration, Providing Foreign Users With A Stronger Three-factor Authentication Path That Meets Their Higher Identity Assurance Needs. Both Branches Merge At A Decision Engine That Allows A Transaction Only When All Required Checks Are Completed. A Gemini-powered Conversational Assistant Included In The Interface Supports Tamil, Hindi, Telugu, Kannada, Malayalam, And English, Making The Platform Easier To Use For Individuals Who Struggle With English-only Terminals. The System Uses A Python-Flask Backend, An SQLite Database, And An HTML/CSS Frontend, Demonstrating That It Is Possible To Develop A Secure, User-friendly, And Fully Cardless ATM Authentication Platform Using Common Web Technologies.

Author: Mrs. Sindhu Biravi S | Abinaya S | Dharani D | Malini S | Keerthana L

Area of research: Computer Science And Engineering

Managing Institutional Queries Efficiently Remains A Persistent Challenge For Colleges And Universities, Particularly When Student Numbers Are Large And Available Support Staff Are Limited. Existing Approaches Such As Physical Helpdesks, Email Threads, And Telephone Helplines Are Restricted In Availability, Inconsistent In Quality, And Unable To Scale During High-demand Periods Such As Admissions Or Examination Seasons. This Paper Presents The Design And Development Of A Web-Based College Query Chatbot System Tailored For Vivekanandha College Of Technology For Women. The Proposed System Combines A React.js Frontend, A Python Flask Backend, And A Microsoft SQL Server Database To Deliver An Always-available, Automated Query-resolution Platform. A Lightweight Natural Language Processing Pipeline Handles Query Understanding Through Lowercase Normalization, Stop Word Removal, And Keyword-based Intent Matching, Without The Use Of Any Machine Learning Model Or Deep Learning Framework. Responses Are Retrieved From A Structured Database Of Predefined Intent-response Pairs. Queries That Cannot Be Matched Automatically, Or That Involve Sensitive Matters, Are Escalated To Appropriate Human Staff Through A Built-in Escalation Mechanism. Verification And Validation Testing Confirmed That The System Correctly Handles All Defined Intent Categories, Provides Consistent And Accurate Responses, And Appropriately Escalates Unrecognized Queries. The System Significantly Reduces The Routine Workload On Administrative Personnel While Ensuring Round-the-clock Student Access To Institutional Information.

Author: Hemapriya P | Maghema R.A. | Madhumathi R | Dhanushya L | Mrs. M. Agalya

Area of research: Computer Science And Engineering

Public Transportation Systems Serve As The Backbone Of Urban Mobility, Yet They Frequently Suffer From Service Quality Degradation Due To The Absence Of Efficient, Real-time Passenger Feedback Mechanisms. Traditional Complaint Channels Such As Physical Complaint Boxes, Telephone Hotlines, And Manual Report Forms Are Plagued By Delays, Data Loss, And Administrative Inefficiencies That Prevent Timely Resolution Of Passenger Grievances. This Paper Proposes And Presents The Design, Development, And Evaluation Of A Smart Public Transport Complaint And Feedback System (SPTCFS) — A Digital Platform That Leverages QR Code Technology, Cloud-based Centralized Database Management, And An Intelligent Administrative Dashboard To Streamline The End- To-end Complaint Lifecycle. Passengers Scan A Unique QR Code Affixed Inside Buses Or At Transit Stations Using Their Smartphones To Access A Dynamic Complaint And Feedback Form Without Requiring Any Dedicated Application Installation. Submitted Data Is Instantly Stored In A Structured Centralized Database, Where Administrators Can Review, Categorize, Assign, Escalate, And Resolve Complaints In Real Time. The System Further Incorporates Automated Status Notification, Complaint Analytics, And A Priority-based Routing Mechanism To Ensure High-severity Issues Receive Prompt Attention. Experimental Evaluation And Usability Studies Conducted Across A Pilot Deployment Involving 300 Participants Demonstrate That The Proposed System Reduces Average Complaint Resolution Time By Approximately 62% Compared To Conventional Methods, Achieves A System Usability Scale (SUS) Score Of 84.6, And Improves Overall Passenger Satisfaction Ratings By 47%. The Proposed Platform Is Lightweight, Scalable, And Universally Deployable Across Any Urban Public Transport Network Regardless Of City Size Or Existing Infrastructure Maturity.

Author: Shamyuktha K | Pratheesha S | Vidhyasree M | Umavathi B | SHAMYUKTHA K

Area of research: Computer Science And Engineering

Grocery Retailing At The Small And Medium Scale Continues To Suffer From Operational Fragmentation, Where Order Intake, Stock Monitoring, Delivery Dispatch, And Customer Communication Are Handled Through Disconnected Channels. The Resulting Delays, Transcription Mistakes, And Opaque Fulfillment Timelines Collectively Erode Both Operational Efficiency And Consumer Trust. This Paper Introduces The Grocery Shop Application (GSA), A Native Android Application Engineered In Kotlin With A PHP Back End And A MySQL Relational Database, That Consolidates All Store-facing Activities Within A Single Tri-module Platform. The Admin Module Enables Centralized Control Over Product Cataloging, Delivery Staff Enrollment, Order Review, Task Assignment, And Progress Monitoring. The Customer Module Provides Account Registration, Product Browsing, Order Placement, Payment Processing, And Shipment Tracking. The Delivery Employee Module Equips Field Personnel With A Structured Interface For Retrieving Assigned Deliveries And Submitting Real-time Status Updates. Controlled Evaluation Over Fifty Simulated Customer Accounts, Ten Delivery Agents, And One Hundred Twenty Catalog Items Demonstrates That The GSA Reduces Order-processing Time By 78.5 Percent, Decreases Delivery-assignment Latency By 85.4 Percent, And Lowers The Order Error Rate From 18.6 Percent To 2.1 Percent Relative To Fully Manual Operations, Achieving An Aggregate F1-score Of 0.89 Across Core Workflow Quality Metrics. The Proposed Framework Offers Grocery Store Operators A Scalable, Locally Deployable, And Cost-effective Solution For End-to-end Retail Management Automation.

Author: Padma Nivedha M | Jaya Priya S | Lathika N | Dharshini E | Harshini D

Area of research: Computer Science And Engineering

Deepfake Technology Threatens Information Integrity By Enabling The Creation Of Highly Realistic Synthetic Media That Can Mislead Viewers, Manipulate Public Opinion, And Compromise Biometric Authentication Systems. Existing Detection Models Struggle To Generalize Across Unseen Manipulation Techniques, Suffering From Severe Performance Degradation When Tested On Deepfakes Generated By Methods Not Present In Their Training Data. This Project Leverages The Semantic Understanding Of CLIP (Contrastive LanguageImage Pretraining) With ParameterEfficient FineTuning (PEFT) To Detect Deepfakes More Robustly. Unlike Traditional Deep Learning Approaches That Require Full Model Retraining, Our Method Finetunes Only A Small Fraction Of Parameters (less Than 1%) While Preserving CLIP's Powerful Zeroshot Capabilities. The Framework Processes Face Images Through A Dualencoder Architecture That Compares Visual Features Against Learned Textual Prototypes Of "real" And "fake" Classes. Evaluation Is Performed Across Three Benchmark Datasets—FaceForensics++, CelebDFv2, And DFDC—to Validate Generalization Performance. Experimental Results Demonstrate That Our Approach Achieves Average Crossdataset Accuracy Of 89.7%, Significantly Outperforming Stateoftheart Methods By 8–12% On Unseen Manipulation Types. ParameterEfficient FineTuning Reduces Training Memory Footprint By 85% Compared To Full Finetuning, Making The Solution Practical For Deployment On Standard Hardware. This Work Establishes CLIPPEFT As A Scalable, Generalizable, And Resourceefficient Framework For Deepfake Detection And Authenticity Verification.

Author: Padma Nivedha M | Yaazhini P S | Sweetha Shree P | Ruvaitha M

Area of research: Computer Science And Engineering

Most Antivirus Programs Today Work By Matching Files Against A List Of Known Viruses. This Approach Fails When A New Virus Appears Or When Malware Changes Its Code. Also, Nobody Checks What Users Copy To Their Clipboard. If Someone Copies A Phishing Link From A Website Or A Scam Message From WhatsApp, Their Antivirus Ignores It Completely.We Built Net Immune To Solve These Problems. It Is A Security App For Windows 10 And 11 That Runs Six Small Monitoring Programs In The Background. These Watch Your Clipboard, Downloads Folder, USB Drives, Browser Tabs, Running Processes, And A Special Drop Folder. When Something Looks Suspicious, The App Sends The Data To Meta's Llama-3.3 70B AI Through Groq Cloud. The AI Reads The Text And Tells Us If It Is Dangerous. If It Finds A Threat, Net Immune Takes Action By Itself - Closing Bad Browser Tabs, Ejecting Infected USB Drives, Or Renaming Dangerous Files.We Tested Net Immune Against New Fake Attacks That Normal Antivirus Missed Completely. Our System Caught 95% Of Them. The Average Warning Time Is Less Than 2 Seconds. The App Uses Almost No GPU And Only 0.3% CPU On Average.

Author: Mrs. G. Nandhini | Kamalesh S | John Peter V | Junaid Ahmed J | Lingesh M

Area of research: Computer Science And Engineering

Financial Trading Continues To Serve As A Cornerstone Of The Global Economy, Yet Traditional Trading Systems Remain Largely Optimized For Skilled Professionals And Rely Heavily On Manual Interfaces Such As Keyboards And Touch-based Inputs. These Interfaces Impose Accessibility Barriers For Novice Traders, Elderly Individuals, And Persons With Disabilities, Limiting Their Active Participation In Financial Markets. Recent Advancements In Artificial Intelligence (AI), Automatic Speech Recognition (ASR), And Natural Language Processing (NLP) Have Opened New Opportunities For Human-computer Interaction Through Natural Speech. Leveraging These Technologies, This Paper Presents Smart Voice Trade (SVT) — An Intelligent Voice-driven Trading Platform Designed To Execute Financial Transactions Through Spoken Commands In Real Time. The Proposed System Integrates A Multi-layered Architecture Combining ASR For Speech Transcription, NLP For Intent And Entity Extraction, And A Secure Broker API Interface For Order Placement And Validation. A Built-in Risk Verification Module Ensures Trade Integrity By Screening For Abnormal Parameters, Market Anomalies, And Execution Inconsistencies. The Platform Also Features Adaptive Speech Models Capable Of Handling Diverse Accents, Background Noise, And Variable Speech Rates, Enabling Robust Performance Across Heterogeneous User Environments. Experimental Evaluation Using A Dataset Of Over 5,000 Voice Samples Demonstrates An Average Intent Recognition Accuracy Of 95.8%, Precision Of 94.9%, And Average Latency Below 220 Milliseconds, Confirming Its Suitability For Real-time Trading Contexts. The System Not Only Enhances Efficiency And Inclusivity In Trading But Also Introduces A Scalable Foundation For AI-assisted Financial Decision-making. By Bridging The Gap Between Voice Interaction And Secure Trading Infrastructure, SVT Contributes To The Democratization Of Fintech Accessibility And Sets The Groundwork For Future Research On Autonomous, Conversational Trading Assistants.

Author: Dr. A. Hemalatha | Anitha Mosses | Anushraj S | Balaji P

Area of research: Information Technology

The Hostel Mess Management System Is A Web-based Application Designed To Automate And Streamline Hostel And Mess-related Operations In Educational Institutions. Traditional Hostel Management Relies On Manual Processes Such As Registers, Paper Forms, And Offline Communication, Which Often Lead To Inefficiencies, Data Errors, And Lack Of Transparency. This System Provides A Centralized Platform Where Students And Administrators Can Interact Efficiently Through A Digital Interface. It Includes Modules For Room Allotment, Attendance Tracking, Mess Menu Voting, Complaint Management, And Night-out Permissions. The System Is Developed Using Modern Technologies Such As Spring Boot For Backend, React.js For Frontend, And MySQL For Database Management. Students Can Mark Attendance, Vote For Preferred Meals, And Submit Complaints Online, While Administrators Can Monitor And Manage All Activities In Real Time. The Implementation Of A Menu Voting System Enhances Student Participation And Satisfaction In Mess Services. Additionally, The Complaint Tracking Feature Ensures Accountability And Timely Resolution Of Issues. Secure Authentication Mechanisms Protect User Data And Ensure Role-based Access Control. The System Reduces Manual Workload, Improves Data Accuracy, And Enhances Operational Efficiency. Overall, The Proposed Solution Provides A Scalable And User-friendly Approach To Modern Hostel Management, Ensuring Transparency, Efficiency, And Improved Communication Between Students And Administration.

Author: Suyash Kakade | Aniket Kandekar | Rushikesh Khetmalis | Sachin Jadhav | Karim Mulani

Area of research: Computer Science And Engineering

Grievance Handling In Most Academic Institutions Still Depends On Paper-based, Informal Processes That Offer Students No Transparency Or Follow-up. The Students Voice And Support System (SVSS) Is A Web-based Platform Built Using Python, Django, MySQL, And JavaScript That Digitizes The Entire Complaint Lifecycle. Students Submit Grievances Online And Instantly Receive A Unique Ticket ID For Tracking. They Can Monitor The Status Of Their Complaints At Any Stage, Attach Supporting Files, Receive Email Updates, And Contribute To A Community Suggestion Board Through Anonymousupvoting. A Dedicated Administrator Interface Provides Real-time Statistics, Category-wise Complaint Breakdowns, And Tools To Update Complaint Status And Post Official Responses. The System Supports Anonymous Filing, Is Fully Responsive On Both Mobile And Desktop Devices, And Incorporates Security Measures, Including CSRF Protection And Password Hashing. Testing Confirmed That SVSS Fulfils All Functional Requirements Reliably And Outperforms Six Comparable Systems In The Literature By Being The Only Solution To Simultaneously Offer Full Lifecycle Management, Low Infrastructure Cost, Mobile Accessibility, And Anonymous Submission Support.

Author: S. Shankar | Monika M | Nihitha B | Santhiya R | Shanjini R S

Area of research: Computer Science

The ERP System Developed For Sanjeevan College Helps Bring All Important Academic And Administrative Work Into One Place. It Manages Student And Faculty Data, Attendance, Exams, And Other Daily Activities In A Simple And Organized Way. By Reducing Manual Work And Paperwork, The System Makes Processes Faster And More Accurate. It Is Designed To Be Easy To Use, Secure, And Useful For Students, Teachers, And Administrators, Helping The College Run More Efficiently.

Author: Vishal Kasture | Akash Pawar | Aditya Gundure | Kapil Patil | Krushna Padgilwar

Area of research: Artificial Intelligence And Data Science

The LLM-Based Carbon Emission Monitoring And Ethical Reporting System Is A Web Application That Helps Organizations Track, Analyze, And Reduce Their Carbon Footprint By Transforming Raw Sales Data Into Actionable Environmental Insights Through An Intuitive Dashboard Interface. Businesses Can Upload CSV Or Excel Files Containing Product Information, And The System Automatically Calculates Carbon Footprints Using Category-specific Emission Factors (such As 8.2 Kg/unit For Electronics Or 1.9 Kg/unit For Dairy). Products Are Classified Into Three Risk Levels - Normal, Critical, Or High-Risk - Based On Total Emissions And Sales Volume, With High-risk Items Triggering Intelligent Recommendations For Eco-friendly Alternatives Like Replacing Plastic Cutlery With Bamboo Or Switching To Plant-based Dairy Options. The System Incorporates An Ethical Reporting Framework Ensuring Transparency, Fairness, And Accountability, While An Interactive Dashboard Visualizes Emissions Through Charts And Summary Cards, And A Government Compliance Report Identifies All High-risk Products With CSV Export Functionality. Built With A Python Flask Backend And Responsive Frontend, This Tool Empowers Organizations To Make Data-driven Sustainability Decisions And Meet Regulatory Requirements.

Author: Maheshwari K | Lokeshwaran S | Ragavan S | JosuvaPM

Area of research: Artificial Intelligence And Data Science

Digital Identity Verification In The Modern Era Frequently Necessitates The Full Disclosure Of Sensitive Personal Data, Which Creates Significant Privacy Vulnerabilities And Security Risks. To Address These Challenges, This Project Proposes A Decentralized, Privacy-preserving Identity Verification System Built On Blockchain Technology. The Framework Utilizes Cryptographic Hashing (SHA-256) To Secure Identity Attributes And Anchor Them On An Immutable, Tamper-proof Ledger.The Core Of The System Is A Selective Disclosure Mechanism, Which Empowers Users To Share Only The Specific Information Required By A Verifier Rather Than Their Entire Identity Profile. By Integrating Concepts From Zero-Knowledge Proofs (ZKP) And Decentralized Identifiers (DID), The Approach Ensures That Authentication Can Occur Without Revealing The Original Sensitive Data To The Verifying Party. This Proposed System Provides A Secure, User-controlled Alternative To Traditional Centralized Databases, Ultimately Enhancing Privacy, Data Integrity, And Trust In Digital Verification Processes.

Author: Maheshwari K | Srinath Sridevan S | Krishna Moorthy G | Antony S

Area of research: Computer Science And Engineering

This Paper Presents A Secure And Efficient Cloud Data Storage Using Deduplication And Blockchain Technology To Reduce Storage Requirements And Enhance Data Security.Cloud Computing Has Become A Fundamental Backbone For Modern Organizations, But The Rapid And Continuous Growth Of Digital Data Has Led To Significant Challenges Such As Increased Storage Costs, Redundant Data Accumulation, And Heightened Security Risks. To Overcome These Limitations, This Project Proposes An Integrated Cloud Storage Framework That Combines Data Deduplication, Semantic Encryption, And Blockchain Technology Into A Unified Architecture. The Deduplication Mechanism Focuses On Identifying And Eliminating Duplicate Data At The Block Level, Ensuring That Only Unique Data Is Stored Within The System. Semantic Encryption Is Employed To Protect Sensitive Information, Ensuring That Encrypted Data Remains Confidential Even After Deduplication And Compression Processes. The Framework Is Designed Such That Unauthorized Users Cannot Derive Meaningful Information From Stored Ciphertexts, Thereby Safeguarding User Privacy. The Blockchain Technology Is Incorporated To Manage Authentication, Access Control, And Metadata In A Decentralized And Tamper-proof Manner. By Maintaining An Immutable Ledger Of Transactions And Access Logs, The System Ensures Transparency, Traceability, And Resistance Against Unauthorized Modifications.

Author: Arunadevi.S | Dharanya.S | Dharanya.V | Gurupriya.R | Ms.S. Gowthami

Area of research: Computer Science And Engineering

Road Accidents Continue To Be A Major Global Concern, Often Caused By Delayed Driver Reaction And Limited Awareness Of Surrounding Conditions. Although Modern Vehicles Include Driver Assistance Technologies, Many Existing Systems Struggle To Provide Reliable Real-time Object Detection Under Challenging Environments Such As Low Light, Heavy Traffic, And Adverse Weather. This Paper Presents VISIONGUARD, An Adaptive YOLO-based Driver Assistance Framework Designed To Deliver Real-time Road Object Intelligence. The Proposed System Utilizes The YOLOv8 Deep Learning Model To Detect Vehicles, Pedestrians, Traffic Signs, And Road Obstacles With High Speed And Accuracy. Unlike Traditional Systems That Are Restricted To Predefined Object Categories, The Framework Incorporates Adaptive Detection Mechanisms To Enhance Flexibility In Dynamic Traffic Environments. The System Processes Live Video Input, Extracts Frames Using OpenCV, And Performs Object Detection With Minimal Latency. Risk Assessment Is Conducted To Generate Immediate Visual And Audio Alerts For Drivers. The Framework Is Optimized For Deployment On Low-power Edge Devices, Ensuring Practical In-vehicle Implementation. Experimental Observations Demonstrate Improved Detection Performance And Real-time Responsiveness. The Proposed Solution Contributes Toward Safer And Smarter Transportation Systemsby Enhancingdriver Awareness And Reducing Accident Risks.

Author: Akalya V | Abinaya S V | Aparna A M | Bhavya Sri T

Area of research: Computer Science And Engineering

Blind And Visually Impaired Individuals Face Continuous Challenges In Their Everyday Lives, Especially When Identifying People, Detecting Obstacles, And Recognizing Currency, Which Significantly Affects Their Independence And Confidence. Traditional Assistive Tools Such As White Canes And Guide Dogs Provide Only Limited Physical Guidance And Lack The Ability To Convey Real-time, Detailed Environmental Information. Existing Digital Solutions Often Focus On A Single Function, Require Multiple Devices, Or Depend Heavily On Internet Connectivity, Which Reduces Their Practicality And Reliability In Dynamic Surroundings.To Address These Limitations, The Proposed System Introduces An Integrated, AI-powered Assistive Technology That Combines Face Recognition, Obstacle Detection, And Currency Identification Into A Single, Compact, And User-friendly Device. The System Captures Real-time Visual Input Through A Wearable Or External Camera And Processes It Using Advanced Deep Learning Algorithms To Ensure High Accuracy And Rapid Response. YOLO Is Utilized For Obstacle Detection, Allowing The System To Identify And Track Nearby Objects, While The Grassmann Algorithm Supports Robust Face Recognition For Identifying Familiar Individuals. Additionally, A CNN-based Model Handles Currency Classification To Help Users Conduct Financial Transactions Independently. All Detected Information Is Converted Into Clear, Context-aware Audio Feedback, Guiding Users Safely And Effectively Through Their Environment. This Unified Approach Minimizes The Need For External Assistance, Enhances Personal Mobility, And Significantly Improves The Overall Quality Of Life For Visually Impaired Individuals. By Offering Affordability, Adaptability, And Precision, The Proposed System Stands As An Innovative Step Toward Intelligent Assistive Technology.

Author: Abina M S | Dharshini R | Janani B | Bhavadharani S

Area of research: MERN STACK DEVELOPMENT

This Paper Presents A Complete Design And Implementation Of A Car Rental Management System Using The MERN Stack (MongoDB Atlas, Express.js, React.js, And Node.js). The System Automates Vehicle Rental Operations, Customer Management, And Booking Workflows. It Ensures Scalability, Real-time Data Handling, And Secure Access Using JWT Authentication And Role-based Access Control. The Frontend Provides An Intuitive Interface For Both Users And Admins, While The Backend Handles All Business Logic Through RESTful APIs. Performance Testing Demonstrates Stable System Behavior Under Concurrent Users With Zero Error Rates. The System Excludes Payment Gateway Integration And Focuses Purely On Booking And Management Functionality.

Author: Abhishek Patel | Naveen Panwar | Dr. Nirupama Tiwari

Area of research: Agricultural Engineering

Crop Yield Estimation Plays A Major Role In Agricultural Planning And Food Security. Traditional Methods Of Crop Yield Estimation Are Time-consuming And Require Extensive Field Surveys. This Study Focuses On Estimating Rice Crop Yield Using Sentinel-1 Synthetic Aperture Radar (SAR) Data Integrated With Land Use/Land Cover (LULC) And Climate Parameters In Orathanadu Taluk, Thanjavur District. Sentinel-1 SAR Data Provides VV And VH Polarization Backscatter Values That Help Analyze Crop Growth And Moisture Conditions. The Collected SAR Data Was Preprocessed Using Radiometric Calibration, Speckle Filtering, Terrain Correction, And Decibel Conversion. Climate Parameters Such As Rainfall, Temperature, And Humidity Were Integrated With SAR Features. A Random Forest Regression Model Was Developed Using Google Colab To Predict Crop Yield. The Model Performance Was Evaluated Using R² And RMSE Values. The Obtained Results Showed High Prediction Accuracy With An R² Value Of 0.9502 And RMSE Value Of 11.24 Kg/ha. Spatial Crop Yield Mapping Was Performed Using GIS Techniques To Identify High And Low Yield Zones. The Study Demonstrates That Integrating Remote Sensing Data With Machine Learning Provides An Efficient And Reliable Method For Crop Yield Estimation.

Author: Prof. G. Durga, M.E. | Keerthana M | Dharani R | Kesavan V | Renuka R

Area of research: Computer Science And Engineering

The Increasing Number Of Vehicle Insurance Claims Has Led To Challenges In Verifying Damages And Detecting Fraudulent Activities. Traditional Claim Verification Methods Are Time-consuming And Prone To Human Error. This Paper Proposes A Deep Learning-based System For Automated Vehicle Damage Detection And Fraud Prevention. The System Utilizes Convolutional Neural Networks (CNN) And Object Detection Models Such As YOLO To Identify And Classify Vehicle Damages From Images. Additionally, Machine Learning Techniques Are Used To Analyze Claim Patterns And Detect Fraudulent Behavior. The Integration Of Artificial Intelligence Improves Accuracy And Reduces Manual Effort. The Proposed System Enhances Transparency And Speeds Up Claim Processing. It Also Minimizes Financial Losses Caused By Fraudulent Claims. Experimental Results Show Improved Performance Compared To Traditional Methods. Overall, The System Provides An Efficient And Reliable Solution For Modern Insurance Industries.

Author: Mr.Mohanasundaram A | Vijay Raghul S | Bharath M | Shanmugarajan P

Area of research: Agriculture

Weed Infestation Is One Of The Major Problems Affecting Crop Growth And Reducing Agricultural Productivity In Groundnut Cultivation. Early Identification And Management Of Weeds Are Essential To Improve Crop Yield. In This Study, An Image-based System Was Developed To Identify Weeds In Groundnut Fields. Images Of Groundnut Crops And Weeds Were Collected Using A Mobile Camera Under Natural Field Conditions. The Collected Images Were Pre-processed Using Resizing And Enhancement Techniques To Improve Image Quality. A Web-based Application Was Developed To Make The System Easy To Use. The Application Allows Users To Upload Field Images And Automatically Analyzes Them To Identify Whether The Image Contains Crop Plants Or Weeds. A Convolutional Neural Network (CNN) Model Was Used To Classify Crop And Weed Images Accurately. Based On The Detection Results, The System Also Provides Suitable Herbicide Recommendations To Control The Identified Weeds Effectively. The Results Are Displayed Through A Simple And User-friendly Interface. This Developed Web System Helps Farmers Quickly Detect Weeds And Select Appropriate Herbicides, Thereby Supporting Better Decision-making, Reducing Manual Labor, Minimizing Excessive Herbicide Usage, And Improving Weed Management Practices In Groundnut Cultivation.

Author: Prof. G. Durga, M.E. | Geethanjali S | Periyannan K | Priyadharshini A | PRIYADHARSHINI A

Area of research: Computer Science And Engineering

Telehealth Systems Have Significantly Improved Access To Medical Consultation, Especially For Patients In Remote Areas. However, Existing Systems Often Lack The Ability To Interpret Patient Emotions And Provide Personalized Doctor Recommendations. This Paper Proposes An Advanced AI-based Telehealth Platform That Enhances Online Medical Consultation By Integrating Multiple Intelligent Techniques. The System Employs Deep Learning-based Facial Expression Analysis To Detect Patient Emotions During Video Consultations. In Addition, A Speech Recognition Module Converts Patient Speech Into Text, Which Is Further Processed Using Natural Language Processing (NLP) Techniques To Extract Symptoms And Identify Possible Medical Conditions. Based On Both Emotional And Clinical Insights, An Intelligent Recommendation Model Suggests The Most Suitable Doctor For The Patient. The Platform Also Supports Secure Video Consultations, Digital Prescription Generation, And Integrated Pharmacy Delivery Services. The Proposed System Aims To Improve Diagnostic Accuracy, Enhance Patient-doctor Interaction, And Provide A More Personalized And Efficient Healthcare Experience.

Author: Mrs. N. Savitha | M.Bharatha selvan | R.Surya | P.Ayus | S.Gnanashivan

Area of research: Computer Engineering

Automatic Image Captioning Is A Challenging Task At The Intersection Of Computer Vision And Natural Language Processing That Involves Generating Semantically Meaningful Textual Descriptions From Visual Input. This Paper Presents A Deep Learning-based Image Captioning System That Integrates Convolutional Neural Networks (CNNs) For Visual Feature Extraction With Long Short-Term Memory (LSTM) Networks For Sequential Language Generation. The Proposed Architecture Employs A Pre-trained VGG16 Model As The Visual Encoder To Extract High-level Feature Representations, Which Are Subsequently Fed Into A Word-embedding-enhanced LSTM Decoder To Generate Context-aware, Grammatically Coherent Captions. The System Is Trained And Evaluated On The Flickr8k Dataset Comprising 8,000 Images With Five Human-annotated Captions Each, Supplemented By Custom Real-world Images To Assess Generalization Capability. Experimental Evaluations Using BLEU-1 Through BLEU-4 Metrics Demonstrate Competitive Captioning Performance With BLEU-1 Of 0.587 And BLEU-4 Of 0.142. Beam Search Decoding Further Improves Caption Quality Over Greedy Search. Results Confirm The Effectiveness Of The CNN-LSTM Pipeline For Automated Image Description, With Applications In Accessibility Tools, Content Indexing, And Human-computer Interaction.

Author: Sahil Nandkumar Pawar | Rohit Kishor Pawar | Bhushan Prabhakar Zade | Aditya Nagesh Sagar

Area of research: Computer Science And Engineering Specializing In Cyber Security

SQL Injection Remains One Of The Most Critical Security Threats To Web Applications, Enabling Attackers To Manipulate Database Queries And Gain Unauthorized Access To Sensitive Information. Traditional Detection Methods Often Fail To Identify Complex And Evolving Attack Patterns. This Paper Proposes An Intelligent SQL Injection Detection System Using A Cost-sensitive Stacked Generalization Learning (CSSGL)-based Hybrid Approach That Integrates Machine Learning And Deep Learning Techniques. The Proposed Approach Analyzes Query Structures, Performs Feature Extraction, And Classifies Queries As Normal Or Malicious With High Accuracy. Experimental Results Demonstrate That The Model Achieves An Accuracy Of 96.8% With Reduced False Positive Rates, Outperforming Conventional Detection Methods. The System Is Capable Of Detecting Both Known And Unknown Attacks Efficiently, Making It Suitable For Real-time Web Application Security.

Author: Ajanya Arunkumar | G.Rajalakshmi | Sedna Sebastian | Mrs. R. Devika

Area of research: Civil Engineering

The Construction Industry In India, Particularly In Maharashtra, Relies Heavily On Small And Medium Enterprises (SMEs) For Project Execution, Yet These Firms Face Significant Vulnerabilities Due To Fragmented And Complex Supply Chains. This Study Investigates Supply Chain Risk Management (SCRM) Practices Tailored For Construction SMEs, Addressing A Critical Gap In Sector-specific Risk Mitigation Strategies. Through A Mixed-methods Approach Involving Surveys, Semi-structured Interviews With SME Owners, Project Managers, And Suppliers, Along With Case Studies Of Ongoing Construction Projects In Maharashtra, The Research Identified Key Supply Chain Risks Such As Material Price Volatility, Supplier Delays, Transportation Disruptions, Quality Inconsistencies, Regulatory Uncertainties, And Financial Liquidity Issues.The Assessment Revealed That Risks Related To Raw Material Procurement And Logistics Have Particularly High Likelihood And Severity, Yet They Are Often Inadequately Managed Due To Limited Resources, Lack Of Formal Risk Protocols, And Low Technological Adoption Among SMEs. The Study Further Highlights The Predominantly Reactive Nature Of Current Risk Management Practices And The Associated Operational Challenges. To Address These Issues, The Research Proposes A Simplified, Practical Four-phase SCRM Framework Encompassing Risk Identification, Assessment, Mitigation, And Continuous Monitoring. This Framework Integrates Low-cost Tools, Stakeholder Collaboration Mechanisms, And Localized Strategies Suitable For Resource-constrained SMEs.

Author: Somnath Tekale

Area of research: Civil Engineering

The Construction Industry Faces Persistent Challenges Due To Fragmentation, Poor Coordination, And Complex Stakeholder Interactions, Leading To Significant Organizational Interface Risks. These Risks, Occurring At Physical, Informational, Contractual, And Relational Boundaries Between Project Parties, Are Major Contributors To Delays, Cost Overruns, Rework, Disputes, And Safety Issues. This Empirical Study Investigates The Impact Of Construction 4.0 Technologies On Mitigating Such Organizational Interface Risks, With A Specific Focus On Medium-to-large Scale Building And Infrastructure Projects In The Mumbai Metropolitan Region, India. A Sequential Explanatory Mixed-methods Research Design Was Adopted. Quantitative Data Were Collected Through A Structured Questionnaire From 372 Construction Professionals, While Qualitative Data Were Gathered Via 28 Semi-structured Interviews And Five Embedded Case Studies. Structural Equation Modeling (PLS-SEM) And Thematic Analysis Were Used For Data Analysis. The Findings Reveal That Soft (organizational) Interface Risks Are Significantly More Severe Than Hard (physical) Risks In The Mumbai Context. Results Indicate A Strong Negative Relationship Between Construction 4.0 Adoption Maturity And Interface Risk Severity (β = -0.689, P < 0.001), With An R² Value Of 0.582. Digital Twin Technology, Integrated With IoT And BIM, Emerged As The Most Effective Solution For Both Hard And Soft Interface Mitigation. Six Key Mechanisms — Real-time Visibility, Predictive Analytics, Automated Clash Detection, Enhanced Collaboration, Immersive Coordination, And Off-site Standardization — Were Identified. Based On The Integrated Findings, The Study Proposes The C4OIRM Framework (Construction 4.0 Organizational Interface Risk Management Framework), A Four-layer Practical Model For Systematic Implementation. The Research Contributes To Both Theory And Practice By Providing Empirical Evidence From A Developing Country Context And Offering Actionable Recommendations For Project Stakeholders, Technology Providers, And Policymakers Aiming To Improve Project Performance Through Digital Transformation.

Author: Swaraj Gharad

Area of research: Computer Science And Engineering

The Maritime Supply Chain Is One Of The Most Complex And Globally Distributed Networks In The World, Involving Multiple Stakeholders Such As Shipping Companies, Port Authorities, Freight Forwarders, Customs Agencies, Insurance Providers, And End Customers. Traditional Container Management Systems Are Burdened With Significant Inefficiencies Including Data Silos, Lack Of Real-time Transparency, Vulnerability To Document Fraud, High Administrative Overhead, And Excessive Delays Caused By Manual Paper-based Processes. This Paper Proposes A Comprehensive Blockchain-based Framework Designed To Deliver Secure, Transparent, And Highly Efficient Management Of Ship Containers Throughout The Entire Maritime Supply Chain Lifecycle. The Proposed System Leverages The Ethereum Blockchain Platform And Solidity Smart Contracts To Automate Container Registration, Real-time Cargo Tracking, Document Hash Verification, Multi-party Authorization, And Automated Regulatory Compliance Enforcement. A Fully Functional Decentralized Application (DApp) Built On React.js For The Frontend And Node.js For The Backend Interfaces With The Ethereum Network Via Web3.js And MetaMask To Deliver Real-time Container Visibility And Tamper-proof Digital Audit Trails. Off-chain Bulk Data Is Stored Using The InterPlanetary File System (IPFS) To Avoid Blockchain Bloat While Preserving Cryptographic Integrity. The System Significantly Reduces Document Processing Time By Up To 84.6%, Eliminates Document Fraud, Enhances Inter-stakeholder Trust, And Lowers Overall Operational Costs. Performance Evaluations Conducted On The Ethereum Sepoliatestnet Using Simulations Of 500 Container Shipments Demonstrate High Transaction Throughput, Sub-30-second Confirmation Latency For Clearance Operations, And Robust Resistance To Common Smart Contract Attack Vectors. The Results Confirm That The Proposed Framework Is Technically Viable And Economically Competitive For Real-world Deployment In Modern Maritime Logistics Infrastructure.

Author: Mr. A. Mohanasundaram | Akash S | Indhirajith R | Roshanth S | Sivaprasanth I

Area of research: Electronics And Communication Engineering

This Project Presents The Design And Implementation Of A UART (Universal Asynchronous Receiver/Transmitter) Transmitter Using A Finite State Machine (FSM) In Verilog HDL. UART Is A Widely Used Serial Communication Protocol That Enables Data Transfer Between Devices Without The Need For A Shared Clock Signal. The Proposed System Converts Parallel Input Data Into A Serial Data Stream By Following The Standard UART Frame Format, Which Includes A Start Bit, Data Bits, And A Stop Bit.The Design Utilizes An FSM To Control The Sequential Transmission Process Through Different States Such As IDLE, START, DATA, And STOP. A Shift Register And Bit Counter Are Used To Ensure Correct Data Sequencing And Timing. The Implementation Is Verified Using A Testbench And Simulated To Observe Correct Waveform Behavior.This Project Provides A Simple And Efficient Approach For Implementing UART Communication In Digital Systems And Can Be Extended To Include Features Such As Parity Checking, Baud Rate Generation, And Full-duplex Communication For Real-time Embedded Applications.

Author: Ahalya S | Alagulakshmi P | Jayasri P | Maha lakshmi S | Dr.R.Sudha

Area of research: Agricultural Engineering

Agricultural Sustainability Is Challenged By Nearly 40% Post-harvest Losses Due To Inefficient Manual And Destructive Ripeness Testing. Bio-Sentry Is An Advanced, Non-destructive Automated System Designed To Bridge This Gap Through A Multi-dimensional Sensing Approach. The Core System Is Built On The ESP32 Microcontroller, Which Orchestrates A Sophisticated Multi-Sensor Fusion Strategy. It Utilizes Acoustic Resonance Analysis Via A MAX9814 Sensor To Measure Internal Density Through Mechanical Tapping By A Servo Motor. An MQ-135 Gas Sensor Is Integrated To Monitor Ethylene Gas Concentrations, A Primary Biological Indicator Of The Fruit's Ripening Stage. To Ensure A Comprehensive Evaluation, The ESP32-CAM Module Performs Visual Inspection, Capturing Surface Colour And Texture Data. A Unified Classification Algorithm Processes These Heterogeneous Inputs To Grade Produce Into Unripe, Ripe, Or Spoiled Categories. The Results Are Transmitted To A Customized IoT Dashboard (Blynk), Allowing Stakeholders To Access Real-time Analytics Remotely. By Incorporating GIS (Geographic Information System) Principles, The System Enables Geospatial Tracking For Optimized Supply Chain Logistics. Bio-Sentry Provides A Scalable, Low-cost Solution That Reduces Waste And Promotes Data-driven Practices In Modern Smart Agriculture.

Author: Abinaya S | Tholkappiyan M | Harishma R | Thirisha R | Ms. S. Vinoba Jenifer

Area of research: FINANCE

This Study Focuses On Trend Analysis Of The Financial Performance Of TI Cycles Of India Limited. It Examines Changes In Key Factors Like Sales, Profit, Expenses, Assets, And Liabilities Over Time To Understand The Company’s Growth And Stability. The Study Uses Secondary Data From Annual Reports And Applies Tools Such As Percentage And Comparative Analysis. The Results Show That The Company Has Experienced Both Growth And Fluctuations Due To Market Conditions And Cost Factors. It Highlights The Importance Of Cost Control And Efficient Resource Use. Overall, Trend Analysis Helps In Better Decision-making And Future Planning For Sustainable Growth.

Author: Subha.R | Ms.B.Nivetha

Area of research: Information Technology

Artificial Intelligence (AI) Has Become A Powerful Tool In Developing Assistive Technologies That Improve Accessibility For Individuals With Disabilities. Among These, Hearing And Speech-impaired Individuals Face Significant Challenges In Communication Due To The Lack Of Widespread Understanding Of Indian Sign Language (ISL). ISL Is A Visual Language That Relies On Gestures, Facial Expressions, And Body Movements. However, Most People Are Not Familiar With It, Leading To Communication Barriers In Everyday Life. Existing Systems Mainly Focus On Recognizing Individual Alphabets Or Static Gestures, Which Limits Their Ability To Provide Real-time, Meaningful Communication. This Paper Proposes An AI-based Two-way Communication System Designed To Bridge The Gap Between Hearing-impaired Individuals And Others. The System Converts ISL Gestures Into Text And Speech While Also Translating Spoken Language Into Text, Enabling Bidirectional Communication. The Proposed Approach Integrates Computer Vision, Machine Learning, And Natural Language Processing Techniques. MediaPipe Is Used For Capturing Real-time Hand Landmarks, While Long Short-Term Memory (LSTM) Networks Are Employed For Recognizing Dynamic Gesture Sequences. Recognized Gestures Are Mapped Into Gloss Representations, Which Are Further Processed Into Meaningful Sentences Using NLP Models Such As BART. Additionally, Multilingual Translation Is Achieved Using IndicTrans2, And Speech Output Is Generated Using Indic Text-to-Speech Systems. The System Is Designed To Be Efficient, Cost-effective, And User-friendly, Making It Suitable For Real-world Applications. The Results Demonstrate Improved Accuracy, Real-time Performance, And Better Contextual Understanding Compared To Existing Methods. This Research Contributes To Enhancing Accessibility And Promoting Inclusivity By Enabling Effective Communication Between Hearing And Non-hearing Communities.

Author: Nivetha S M | Obuli Dharani Dharan O | Akash S | Arunprasath S | Mouleeshwaran G

Area of research: Computer Science And Engineering

The Rapid Expansion Of Digital Banking Has Improved The Accessibility And Efficiency Of Financial Transactions But Has Also Introduced Significant Security Challenges. Traditional Authentication Methods Such As Passwords And PINs Are Vulnerable To Attacks Like Phishing, Brute-force, And Shoulder Surfing. This Research Proposes A Multi-layered Authentication Framework To Enhance Net Banking Security. The First Layer Uses An Illusion-oriented PIN Mechanism That Conceals User Input Through A Dynamic And Deceptive Keypad. The Second Layer Integrates Real-time Facial Biometric Authentication For Accurate User Verification. Advanced Techniques Are Applied To Extract And Analyze Unique Facial Features. This Combination Ensures That Only Authorized Users Can Access The System, Even If Credentials Are Compromised. The Proposed Approach Effectively Reduces Unauthorized Access And Financial Fraud. It Also Maintains A Balance Between Security And Usability Without Requiring Additional Hardware.

Author: Mrs.Banuppriya P | Moulieswaran J | Mythiswaran M | Kavinkumar R | Somnath C

Area of research: Computer Science And Engineering

Due To The Fast-paced Development Of Cities And The Rise In The Number Of Cars, Parking Has Become Quite Challenging To Manage. Conventional Parking Systems Depend Mainly On Human Efforts, Tickets, Or Just Basic Sensors. Such Approaches Prove To Be Ineffective, Expensive, And Error-prone. Besides, There Is No Assurance Of Any Sort Of Safety, Making It Possible For Unauthorized Individuals And Car Thefts. The System Proposed Is Aimed At Solving Such Inefficiencies. The System Utilizes The Concept Of Deep Learning Smart Parking System. The System Utilizes Facial Recognition And License Plate Detection, Where The System Can Automatically Identify Who Is Accessing A Particular Parking Bay. High-definition Cameras Are Used To Take Images Of Both The Driver's Face And The License Plate Of The Vehicle. The Vehicle's Identity Is Determined Using The YOLO (You Only Look Once) Algorithm That Helps Detect And Recognize License Plates. Grassmann Algorithm Is Used For Authenticating Drivers In Order To Ensure That Only Authorized Personnel Will Be Allowed To Enter The Parking Facility. It Involves Two Methods Of Authentication, Making The Process Much More Secure While Reducing Dependency On Tangible Tokens Such As RFID Cards. Such An Automated Process Minimizes The Need For Human Labor And Therefore Results In Improved Management Of Parking Facility As Well As Traffic Flow. Moreover, The Cost Of Hardware Is Reduced Since There Is No Requirement Of Any Special IoT Sensor. Lastly, The Process Helps In Collecting Useful Data Regarding Traffic Movements.

Author: Mrs. Mohanasundaram A | Manikandan P | Naveen P | Sudharsanam S | Kaleeswaran M

Area of research: Computer Engineering

Early Detection Of Pomegranate Diseases Matters Because Catching Them Late Means Lost Yield And Real Financial Damage. This Paper Describes A Disease Detection System Built On YOLOv11, Which Identifies Bacterial Blight And Fungal Infections Directly From Fruit Images. The Model Achieves 99.2% Precision, 99.1% Recall, And 99.5% MAP@50 On Our Validation Set, While Running At Over 220 FPS On GPU Hardware—demonstrating Strong Suitability For Real-world Agricultural Deployment.

Author: Chintamani Adak | Anish Chaudhari | Dinesh Sabale | Vikram Mugale

Area of research: Civil Engineering (Concrete Technology / Nanomaterials)

This Publication Summarizes An Experimental Investigation On Improving The Mechanical Properties Of Concrete Using Nanoparticles Such As Nano-silica, Nano-alumina, And Nano-titania. The Study Focused On Enhancing Compressive Strength, Split Tensile Strength, Flexural Strength, And Durability Of M30 Grade Concrete. Different Percentages Of Nanoparticles Including 1%, 2%, And 3% Were Incorporated Into Concrete Mixes And Tested After 7, 14, And 28 Days Of Curing. The Results Demonstrated That Nanoparticles Significantly Improve Concrete Performance, Particularly At 2% Dosage Where The Maximum Strength Enhancement Was Observed. The Research Confirms That Nanoparticle-based Concrete Can Be Used In Modern Infrastructure For Improved Strength, Durability, And Sustainability.

Author: Mr. Chavda Sunny Surajsinh | Pankaj Kumar Yadav | Shiddhant Basnet | Kumkum Bhattacharjee

Area of research: Computer Science And Engineering

The Rapid Growth Of Internet Usage And The Widespread Adoption Of Virtual Private Networks (VPNs) Have Significantly Enhanced Secure Communication, But Have Also Introduced New Challenges In Detecting Cyber Threats Hidden Within Encrypted Traffic. Traditional Intrusion Detection Systems Often Fail To Identify Such Threats Due To Their Reliance On Signature-based Methods And Inability To Analyze Encrypted Payloads Effectively. This Project Presents A Hybrid Deep Learning-based VPN Traffic Detection System That Integrates Convolutional Neural Networks (CNN) And Long Short-Term Memory (LSTM) Networks To Accurately Identify Anomalous Network Behavior. The System Utilizes Structured Network Traffic Data With Features Such As Protocol Type, Service, Flag Status, Traffic Rates, And Byte Counts, Which Are Preprocessed Using Label Encoding And Feature Scaling Techniques. In Addition, Conventional Machine Learning Models Such As Random Forest And Support Vector Machine (SVM) Are Implemented For Performance Comparison. The Hybrid CNN–LSTM Model Captures Both Spatial And Temporal Patterns In Network Traffic, Resulting In Improved Detection Accuracy. A Real-time Web-based Dashboard Developed Using Streamlit Enables Users To Input Parameters And Visualize Prediction Results, Including Classification Outcomes And Confidence Scores. An Automated Alert Mechanism Is Also Incorporated To Notify Users Of Suspicious Activities, Facilitating Timely Response To Potential Threats. Experimental Results Demonstrate That The Proposed System Outperforms Traditional Methods In Terms Of Accuracy, Precision, And Recall, Providing A Scalable And Efficient Solution For Real-time VPN Traffic Monitoring And Cyber-attack Detection In Enterprise Environments.

Author: Dr. K. Sangeetha | Mrs.N.Poornima

Area of research: Computer Science And Engineering

Nowadays, Social Media Platforms Have Emerged As One Of The Primary Modes Of Communication Where Memes And Comments Are Used Extensively For Spreading Ideas, Humor, And Thoughts. This Increasing Trend, However, Has Been Accompanied By The Speedy Proliferation Of Offensive Language, Hateful Rhetoric, And Cyberbullying, All Of Which Harm Individuals And Their Communities On Social Media Platforms. The Current Solutions For Moderating Such Content Rely Mainly On Manual Reporting And Simplistic Techniques Like Keyword Filters, Which Are Inefficient Because They Do Not Comprehend Context And Are Ineffective Because They Lack Speed. Memes, Especially, Have Text Along With Graphics And Images, Which Complicates Efforts To Automatically Detect Any Kind Of Harmful Content Contained In Them. As A Solution To This Problem, The Proposed Project Will Design A Smart System For Classifying Memes. In The Proposed System, Optical Character Recognition (OCR) Technology Is Applied To Recognize Text From Memes, Which Is Then Analyzed With NLP Technologies Like Tokenization, Stemming, And Stop Word Removal. The Sentiment Of The Posts Is Classified Using The VADER Algorithm Into Three Categories: Positive, Negative, And Neutral. At The Same Time, Deep Learning Image Classifier Is Used To Determine If There Are Any Unsuitable Pictures In Posts. Based On The Output, The System Can Automatically Remove Harmful Content, Issue Notifications And Warnings, As Well As Keep Track Of User Activity, Thus Blocking Users Who Frequently Violate The Community Standards.

Author: BHAVATHARANI M | Dr.Nilabar Nisha U | Thirisa S | Deepika G | Keerthanasri S

Area of research: Computer Science And Engineering

The Internet Of Things Framework Has Seen Swift Transformations Regarding Its Applications And Global User Base. The Necessity For Reliability In Ensuring Excellent Service Quality Is Paramount, Considering The Characteristics Of Data Transmission Via Wireless Media. The Emergence Of High-performance Chips With Compact Sizes And Low Power Consumption, Capable Of Executing Reasonably Sophisticated Algorithms, Has Become Essential For Internet Of Things Applications. This Research Study Concentrates On The Design And Implementation Of Turbo Code Blocks Utilizing The BCJR Method To Couple The Bits Within The Composite Transport Block. The Information And Parity Bits Are To Be Integrated To Enhance Information Exchange Within The Transport Block, Hence Significantly Decreasing The Mistake Rate In The Error Waterfall Phase. A Comparative Analysis With Respect To The Error Rate Has Been Done So As To Evaluate The Quality Of Service Of The Proposed Work. The Lower Error Rate Of The Proposed Work Ensures The High Quality Of Service And Trustworthiness Of The IoT System.

Author: Chaitanya Dewangan | Mr. Deepesh Dewangan

Area of research: Electronics Engineering

Conventional Computer Networks Have Undergone A Paradigm Shift In Terms Of The Advent Of Wireless Pervasive Networks Such As IoT And Fog Networks. The Ease Of Mobility And Adaptive Configuration Enables Significant Ease Of Deployment And Use Of Wireless Networks Over Wired Networks. However, The Associated Challenge Remains The Fact That Wireless Software Defined Networks (SDNs) Are More Prone To Attacks From Adversaries Due To The Absence Of A Secured Communication Medium. The SDN Framework Allows For A Completely Software Based Control Plane Of The Network, Which When Coupled With Stochastic Computing Can Be Leveraged To Analyze Network Data. The Analysis Of Data Passing Through An Adversarial Channel Can Be Used For Identifying Potential Attacks On The Network. This Paper Presents A Deep Learning Model For Analyzing Channel Attributes To Estimate Potential Adversarial Activity And Secure Data Transmission. The Performance Metrics Of The System Has Been Chosen As The Error Rate And Sum Secrecy Rates. Comparing The Performance Of The Proposed System With Benchmark Models Indicates Improved Performance Of The Proposed System.

Author: Pooja Jhanjhot | Amit Sharma | Neelam Sharma

Area of research: ECE

Present Day Communication Systems Are Facing Some Critical Issues Which Are Increased Number Of Users, The Amount Of Bandwidth Availability To Be Used By The Users And The Need For Ever Increasing Data Rates. The Major Concern Regarding All The Problems Is The High Capacity Expectation From Wireless Channels. However, Wireless Channels Are Often Random In Nature With Frequency Selective Nature At The Basest. The Limitation In The Bandwidth Support By Any Channel Makes The Data Rate Support To Be Limited. In This Paper, A Deep Neural Network Assisted Massive MIMO System Has Been Designed And Has Been Employed To Commonly Existing Diverse Channel Conditions. To Increase The Spectral Efficiency And Simultaneously Reduce The BER Of The System, The Maximum Ratio Comining (MRC) Approach Has Been Used Along With MMSE And ZFE Equalization Techniques. The Proposed System Has Been Simulated On Matlab. The Performance Of The System Has Been Evaluated In Terms Of The Bit Error Rate And Spectral Efficiency Of The System.

Author: Harsh Patidar | Neelam Sharma | Amit Sharma

Area of research: ELT (for Science Students )

Student-centered Learning Plays A Major Role With The Development Of The Education System, Especially In Learning English As Second Language (ESL) Contexts. Here, The Learner Is Also Entrusted With Much Responsibility In The Teaching/learning Process. In Such Circumstances, The Part Played By Peer Assessment As A Prominent Alternative Assessment Tool Is Significant. The Significance Of Peer Assessment Is Highlighted In Different Learning Contexts And Research. However, Studies On Peer Assessment Are Still Limited In Sri Lankan Context. Hence, This Study Aims To Investigate The Extent To Which Peer Feedback Is Valid In Assessing Presentation Skills In The ESL Context. Peer Assessment Is Still A New Concept In The Context Where This Study Has Been Conducted, And It Needs More Exploration And Practice. The Participants Of This Quantitative Study Were Three Teachers, 20 Students As Assessors And 30 Students As Assesses, Who Were Studying Technology In A Higher Educational Institute. The Assessors Gave Peer Feedback For The Oral Presentations Of Their Peers. The Feedback Forms Were The Research Instruments Of This Study And Independent Samples T-test Was Used To Analyze Data. The Results Indicate That Student Assessors Have Given Higher Marks Than Teachers. This Study Underscores The Need To Approach Peer Assessment From A Systemic Perspective, Considering Its Role Within The Course As A Whole And Its Interconnections With Other Learning Activities.

Author: Sevwandi Ganesha Pathmaperuma | Sevwandi Pathmaperuma

Area of research: Computer Science And Engineering

In Today’s Highly Digitized World, It Has Become Crucial To Have The Ability To Share Multimedia Content In A Safe And Reliable Manner Owing To Increasing Threats From Cyber Security Experts. With That In Mind, This Project Will Focus On Designing An Innovative Yet Highly Efficient System For Safe Content Sharing With The Use Of Hybrid Encryption Algorithm Based On AES, Which Is Ideal For Data Encryption Along With ECC, Which Will Help Generate And Distribute Keys. Key Components In This Design Include Content Provider, Server, Content Requester, And Access Control, Which Facilitate Safe Uploading, Storage, And Retrieval Of The Content In An Encrypted Format. Further Adding To The Security Aspect, The System Uses Discrete Wavelet Transform Based Steganography Whereby The Audio File Is Hidden Inside A Carrier Image, Thus Offering An Extra Layer Of Security By Hiding The Information. It Can Be Noted That The Presence Of Confidential Information Will Remain Secret As No One Would Even Realize Its Presence Since The File Itself Would Not Reveal Anything To Begin With. Moreover, The System Includes Access Control As Well As A Notification System, Which Will Alert The Content Provider If Any Unauthorized Request Is Made For Accessing The Multimedia Content.

Author: Dr.U.Nilabar Nisha | Ajaykumar N | Dinesh V | Veeramani V | Vishwa V

Area of research: Computer Science And Engineering

Cloud Server Infrastructures Form The Foundation For Technical Services, Where The Services Will Be Highly Available, Critical For Business Continuity And User Satisfaction. Traditional Systems Notify The User After The Failures Had Occurred. It Is Completely Based On Threshold-based Alerts And Results In Performance-degradation, Downtime, Inefficient Resource Utilization And Delayed Recovery. To Address This Problem, Our Proposed Framework Twins The Cloud Server Digitally To Predict The Anomalies And Self-heal The Cloud Server Environments. The Framework Continuously Mirrors Telemetry Data Such As CPU Utilization, Memory Usage, Disk I/O And Network Throughput Into A Virtual Replica Of The Physical Infrastructure. A Hybrid Anomaly Detection Model That Combines Long Short-Term Memory (LSTM) Networks And Isolation Forest Is Used To Identify Early-stage Performance Degradation. The Proposed System Results In Reduced Downtime, Efficient Resource Utilization When Compared To Traditional Monitoring Techniques. The Framework Demonstrates The Feasibility Of Autonomous And Intelligent Cloud Infrastructure Management.

Author: Karthik. M | Sanjay Jerene M | Ranjith R | Shafee Hamath H | Vigneshkumar. R

Area of research: Environmental Science

Repetitive Use Of Soil For Farming Losses Its Nutritious Values.As A Result, To Increase Soil Fertility And Plant Nutrition Chemical And Biological Fertilizers Have Been Added Tremendously Nowadays. Eggshell Which Is A Municipal Waste In Our Society But It Is A Natural Source Of Calcium Carbonate And It Gained Attention For Its Potential Benefit On Fertility. It’s Help To Neutralize Acidic Soils And Improve The Soil PH Level. Eggshell Powder Also Contains Other Essential Nutrients Such As Magnesium, Potassium And Phosphorus Etc. Which Are Important For Plant Growth And Development. The Aim Of This Study Is To Investigate The Potential Of Eggshell As A Fertilizer Which Can Be A Solution Of A Waste Problem. Carbon, Nitrogen, Calcium, Zinc And Manganesecontent Of Standard Soil Containing Eggshell And Without Eggshell Have Been Analyzed.Study Showed That The Soil With Eggshells Powder Has Higher Percentage Of Carbon Than Soil Without Eggshell Powder.This Experimental Study Proves That Eggshell Powder Can Be Used As Fertilizer Which Is A Cost-effective Solution Of Waste Problem.

Author: Monami Haque