IARJSET aligns to the suggestive parameters by the latest University Grants Commission (UGC) for peer-reviewed journals, committed to promoting research excellence, ethical publishing practices, and a global scholarly impact.
FORMULATION ANALYSES, AND ACCEPTABILITY OF SINIGANG-FLAVORED CHIPS
DAIZA VENCH S. FRANCISCO, MAEd-TLE-HE
DOI: 10.17148/IARJSET.2026.13501
Abstract: This study focused on the development and evaluation of sinigang-flavored chips, a novel snack inspired by the Filipino sour soup, sinigang. The research aimed to formulate chips using three protein-0based treatments: Treatment A (Pork-Based), Treatment B (Shrimp-Based), and Treatment C (Fish-Based), and to assess their sensory qualities, overall acceptability, shelf-life stability, microbial safety, and nutritional composition. An experimental-developmental design was employed, integrating natural sinigang flavors with tamarind extract and traditional souring agents into a starch-based chip mixture with the designated protein source. The dough was shaped, dried, and fried to achieve consistent crispness, then evaluated by ten semi-trained panelists for appearance, aroma, taste, and texture using a 9-point hedonic scale. General acceptability was assessed by 100 consumer respondents. Kruskal-Wallis tests and ANOVA were applied to determine significant differences among treatments. Results showed that Treatment C (Fish-Based) achieved the highest ratings for appearance and texture, described as extremely appealing and extremely crunchy, while Treatment A (Pork-Based) led in taste and aroma, rated as extremely delicious and very much pleasant. Overall acceptability favored Treatment A (Pork-Based), indicating the most balanced sensory profile. Shelf-life analysis confirmed product stability, with low moisture and proper drying and frying maintaining crispness and sensory quality over time. Microbial assessment verified safety, with no detection of fecal coliform, E. coli, or Salmonella, and yeast and mold counts within acceptable limits. Proximate analysis indicated high carbohydrate content, moderate fat, modest protein, and low moisture, supporting nutritional value and storage stability. The study concludes that the main protein source significantly influenced sensory perception and consumer preference, with Pork-Based chips providing the most favorable combination of flavor, aroma, appearance, and texture. These findings offer practical insights for the production and commercialization of culturally inspired, ready-to-eat snacks that combine traditional flavors with appealing sensory and nutritional qualities. Keywords: Formulation, Analyses, Acceptability of Sinigang Flavored-Chips
Environmental Challenges and Solutions for Sustainable Development
Sanjeev Kumar Vidyarthi*, Kumari Sushma Saroj, Hari Mohan Prasad Singh
DOI: 10.17148/IARJSET.2026.13502
Abstract: India's rapid industrialization, urban expansion, and resource-intensive economic growth have posed significant challenges to environmental sustainability, including air and water pollution, land degradation, loss of biodiversity, and climate change. These environmental issues have direct consequences on public health, agricultural productivity, and ecosystem stability, thus threatening the long-term success of development goals. This paper provides a comprehensive analysis of the intricate challenges to environmental sustainability in India, addressing institutional, technological, and socio-economic constraints. It critically evaluates various contemporary strategies adopted by the Indian government, civil society, and the private sector-such as the incorporation of renewable energy, waste management systems, green urban planning, and environmental regulations. The consequences of inaction are also examined from the standpoint of ecological decline and social disparities. This research seeks to explore the Environmental Challenges and Solutions for Sustainable Development. Keywords: Environmental issues, socio-economic, public health, sustainable development
Abstract: The process of designing the examination papers is quite lengthy, biased, and time-consuming. The current research introduces a novel solution to create automatic examination papers based on the syllabus documents in the PDF format. The system uses PyMuPDF for extracting information and processing unstructured text using state-of-the-art Natural Language Processing tools. The generator makes use of a transformer neural network model named Flan-T5 which can produce multiple- choice questions (MCQs) along with contextually appropriate distractors and descriptive long-answer questions. The system also incorporates a login module to ensure secure access and provides the option of exporting the question papers in TXT and PDF formats. According to experimental results, the system shows a remarkable improvement in the speed of generating questions and saves almost 85 percent of the time as compared to the conventional technique. The experiments also confirm the quality of the system as far as coherence and grammaticality of the generated questions are concerned. Keywords: Natural Language Processing, Automatic Question Generation, Transformer Models, PDF Text Extraction, Educational Technology
Shabana Khanum, K. Sri Vaishnavi, D. Rajini, K. Madhu Latha
DOI: 10.17148/IARJSET.2026.13504
Abstract: Timely access to plasma donors is critical in medical emergencies, yet traditional coordination methods often rely on phone calls, social media posts, and manual records, which can cause delays. Plasma Connect is a full-stack web platform designed to simplify and improve the plasma donation process by connecting donors, recipients, hospitals, blood banks, and administrators within a single system. The platform allows users to search for plasma donors based on blood group and location, submit donation requests, and track request status in real time. The system is developed using modern web technologies, including React for the frontend, Node.js and Express for backend services, and MySQL for database management. Secure authentication is implemented using JSON Web Tokens and bcrypt encryption. Real-time notifications are provided through Socket.IO, while location-based services are supported using Leaflet maps integrated with OpenStreetMap. The proposed platform improves coordination between donors and recipients, reduces response time in emergency situations, and provides a scalable digital solution for plasma donation management. The results demonstrate that Plasma Connect enhances accessibility, communication, and efficiency in plasma donor coordination. Keywords: Plasma Donation System, Healthcare Web Platform, Donor Matching System, Blood and Plasma Donation Management, React, Node.js, Real-Time Notifications, Location-Based Services, Web Application Development.
Abstract: Healthcare accessibility remains a major challenge, particularly for elderly individuals, rural populations, and users with limited technical proficiency. Most existing healthcare applications rely on text-based interfaces and complex navigation, which can delay timely medical assistance during critical situations. This paper proposes a Voice-Based Intelligent Healthcare Assistant that enables users to interact with healthcare services through natural voice commands. The system integrates speech recognition, multilingual translation, Retrieval-Augmented Generation (RAG), and Large Language Models (LLMs) to perform symptom analysis and provide contextual health guidance. In addition to symptom assistance, the platform supports voice-driven doctor appointment booking and emergency detection with nearby hospital identification using location-based services. The system is implemented using a React Native mobile interface, FastAPI backend, MongoDB database, and Qdrant vector database for semantic retrieval. Experimental evaluation demonstrates that the proposed system provides accurate symptom interpretation, multilingual accessibility, and real-time responses within a few seconds. The solution improves healthcare accessibility and provides an intuitive digital healthcare support system for diverse populations. Keywords: Voice-Based Healthcare System, Artificial Intelligence in Healthcare, Retrieval-Augmented Generation, Multilingual Speech Processing, Medical Symptom Analysis, Digital Health Assistant
The Implementation of IoT enabled, AI Driven Water Quality Monitoring and Controlling System for Aquaculture
G. Krishnaveni, Dr. Ch.Hima Bindu, P. Santhi, Y.Naga jyothi, V. Lakshmi
DOI: 10.17148/IARJSET.2026.13506
Abstract: In this paper, we describe about the design of an IoT-enabled, AI-driven water quality monitoring and control system for aquaculture. Maintaining optimal water conditions is essential for healthy aquatic ecosystems and sustainable food production. The system uses sensors to measure key parameters such as temperature, pH, turbidity, total dissolved solids (TDS), and ammonia levels. These sensors are integrated with Arduino Uno and ESP32 microcontrollers, which collect and transmit real-time data to a cloud platform for storage, visualization, and remote access. An AI model analyzes the data to detect anomalies, identify trends, and predict potential water quality issues. When abnormal conditions occur, the system generates instant alerts and classifies water quality into safe, warning, or critical levels using visual indicators. This enables faster decision-making and reduces the need for manual monitoring. Overall, the system improves efficiency, supports sustainable aquaculture practices, and helps ensure better environmental and production outcomes. Keywords: IoT, Aquaculture, Water Quality Monitoring, Artificial Intelligence, Real-Time Monitoring.
A COMPARATIVE STUDY OF THE DELIVERY APPS BLINKIT AND ZEPTO WITH SPECIAL REFERENCE TO COIMBATORE
Dr. M. K. Palanisamy, Mr. Boobala Krishnan CS
DOI: 10.17148/IARJSET.2026.13508
Abstract: This study focuses on a comparative analysis of two leading quick commerce delivery applications, Blinkit and Zepto, with special reference to Coimbatore. The rapid growth of digital technology and increasing consumer demand for convenience have led to the emergence of ultra-fast delivery platforms. These applications promise delivery within 10-20 minutes, transforming traditional retail practices. The research examines consumer awareness, usage patterns, service quality, delivery speed, pricing, and overall satisfaction. Data was collected from 160 respondents using a structured questionnaire. Statistical tools such as percentage analysis, chi-square test, and ranking analysis were used Keywords: Quick Commerce, Blinkit, Zepto, Consumer Satisfaction, Delivery Speed, Online Grocery, Customer Preference, Coimbatore Market
FACTORS AFFECTING POOR RESULTS IN PAPER AND PENCIL TEST AMONG GRADE 7 STUDENTS IN T.L.E.: BASIS FOR REMEDIAL PROGRAM
ROY D. DIAZ
DOI: 10.17148/IARJSET.2026.13509
Abstract: This study investigated the factors contributing to poor results in paper and pencil tests among Grade 7 students in Technology and Livelihood Education (TLE). A descriptive-correlational research design was employed, with a sample of 155 Grade 7 students from Tuburan National High School. The results showed that teaching methods, student engagement, and curriculum alignment are significant factors affecting poor results in paper and pencil tests. Specifically, the study found that current teaching practices may not effectively cater to the learning needs of students, leading to poor test results. Additionally, student engagement was found to be relatively low, and curriculum alignment was found to be high. The study recommends the implementation of interactive teaching strategies, incorporation of technology, and provision of remedial classes to improve student performance. Furthermore, the study suggests that educators should regularly review and update the curriculum to reflect changes in technology and industry standards. The findings of this study have implications for educators, policymakers, and curriculum developers seeking to improve student outcomes in TLE. Keywords: Curriculum, Teaching Method, Student Engagement
DESIGN AND IMPLEMENTATION OF A PIC-BASED PHASE FAILURE AND THERMAL PROTECTION SYSTEM FOR THREE-PHASE INDUCTION MOTORS
Dr. M. SARITHA, PAWAR ADITHYA, EPPA PRANAYA, DAYYALA BALAJI, NELLUTLA YOGI
DOI: 10.17148/IARJSET.2026.13510
Abstract: Three-phase induction motors are widely used in industrial applications but are highly vulnerable to faults such as phase failure and excessive temperature rise, which can lead to severe damage and reduced operational life. This paper presents the design and implementation of a microcontroller-based protection system using PIC16F72 to detect and prevent such faults. The system continuously monitors phase conditions using voltage sensing circuits and measures motor temperature using an LM35 sensor. Upon detecting abnormal conditions such as phase loss or overheating, the system automatically disconnects the motor supply using a relay mechanism. The proposed system provides real-time monitoring, fast response, and improved reliability, making it suitable for industrial motor protection applications. The microcontroller processes these inputs and triggers a relay mechanism to disconnect the motor supply whenever abnormal conditions are detected. The proposed system ensures fast response, improved reliability, and reduced maintenance costs. The design is simple, cost-effective, and suitable for small and medium-scale industrial applications Keywords: Three-Phase Induction Motor, PIC16F72 Microcontroller, Phase Failure Detection, Temperature Monitoring, Motor Protection System, Relay Control, LM35 Sensor
Automated Harmful Content Control and Blocking System for Social Media
D. Tejaswi, K. Anusha, G. Vanaja, K. Deepa Sri Bhramaramba
DOI: 10.17148/IARJSET.2026.13511
Abstract: The rapid growth of multimedia content on digital platforms has created a need for efficient content moderation systems to prevent the spread of harmful material. This project presents an Automated Harmful Content Control and Blocking System that performs analysis of media before it is uploaded. The system allows users to upload images and videos through a web interface, where the content is processed using a Flask-based backend. For image analysis, OpenCV is used to perform preprocessing techniques such as grayscale conversion and pixel intensity evaluation. A threshold-based method is applied to determine whether the uploaded image contains potentially harmful content. If harmful content is detected, the system blocks the upload and notifies the user; otherwise, the file is stored successfully. The system also supports video uploads and provides a deletion feature for managing uploaded files. This approach ensures real-time moderation, reduces dependency on manual monitoring, and enhances platform safety. Although the current implementation uses basic image processing techniques, it can be extended with advanced machine learning models for improved accuracy in future developments. Keywords: Harmful Content Detection,Image Processing, OpenCV,Flask Content Moderation,Social Media Safety
Abstract: The rapid expansion of educational institutions has led to increasingly complex campus infrastructures, making navigation difficult for students and visitors. Campus Connect is a smart college navigation system designed to provide efficient, user-friendly, and accessible navigation within academic environments. The system enables users to search for classrooms, laboratories, administrative offices, and other facilities through an intuitive interface. Unlike traditional static maps, it provides dynamic search functionality and simulated navigation using structured campus data. The system is implemented using lightweight web technologies and browser-based local storage, ensuring fast performance and offline accessibility. This approach reduces dependency on external servers while improving responsiveness and scalability. The proposed solution enhances user experience, reduces navigation time, and improves accessibility for new users. Keywords: Smart Campus, Navigation System, Web Application, Local Storage, User Interface, Campus Automation
DEVELOPMENT OF AN INTERACTIVE DASHBOARD AND EST-BASED FORECASTING MODEL FOR REAL-TIME DECISION MAKING IN SUPPLY CHAIN MANAGEMENT.
Haritha J
DOI: 10.17148/IARJSET.2026.13513
Abstract: The objective of the project is to develop an Interactive Dashboard for Supplier Weight and Delivery Weight Forecasting using an ETS (Error, Trend, Season) model built using Python to support data-driven decision making within a large engineering manufacturing organisation. The study comprises two integrated components designed to illustrate how visualization and forecasting can facilitate improvements in supply chain monitoring and planning. The organization relies on consistent supplier deliveries and accurate weight forecasting based on historical data for its large-scale manufacturing operations. The manual evaluation of supplier on-time delivery performance and weight of deliveries was often ineffectual and could lead to poor speculation .An Interactive Dashboard was generated in Power BI to visualize supplier performance and delivery timelines. A Forecasting Model was developed in Python with an ETS model predicting the weight of delivery in the future based on historical trends. Combining the analytical power of the Python statistical tool with the data visualization through Microsoft Power BI will help provide actionable insights to improve supply chain management processes and how decisions will be made in the future. The use of an interactive dashboard allowed all users to see how suppliers are performing and the trend regarding weights, and with Python's ETS modelling, a clear prediction of the future delivery weights enables faster decision-making and would lead to increased transparency while using data-informed planning methods to transform supply chain operations.
Keywords: Interactive dashboard, time series forecasting, ETS model, supply chain management, Power BI, Python.
A STUDY ON INFLUENCE OF SOCIAL MEDIA VIDEOS CONTENT TOWARDS CLIENT SATISFACTION
Karthick Sai M
DOI: 10.17148/IARJSET.2026.13514
Abstract: In recent times, social media has been identified as a major medium of communication in the purpose of business and clients. Among all the content available through social media, videos have been identified as gaining prominence as they are considered more engaging, informative, and understandable. Videos are being utilized on social media for the purpose of communication with clients and to promote business.
The main aim of this research is to identify the impact of social media content on client satisfaction. This research aims to find out how social media content is influencing client satisfaction. To attain this, data is being collected in the form of questionnaires from social media users and is being supplemented with literature. From the results of this research, it is identified that videos are playing a vital role in forming opinions in the minds of clients about a brand.
Keywords: Social media, video marketing, customer engagement, client satisfaction, digital communication
Incremental Passivity Control in 7level Cascaded H-Bridge Converters
Dr. N. Kalpana, Gunti Vishal Babu, Kothedigi Naveen Kumar, M Vardhan patel, Laxmi Vara Prasad
DOI: 10.17148/IARJSET.2026.13515
Abstract: This research presents an advanced evolution of the 7-level Cascaded H-Bridge (CHB) inverter by introducing an 11-level topology integrated with Neural Network-Enhanced Incremental Passivity-Based Control (NN-IPBC). The proposed architecture significantly improves power quality metrics, specifically targeting the reduction of Total Harmonic Distortion (THD) and enhancing the precision of capacitor voltage balancing across five seriesconnected modules. By utilizing a high-density modular topology, the system synthesizes a near-sinusoidal 11step waveform. The NN-IPBC algorithm, executed on an Arduino Mega 2560, provides real-time optimization of energy-shaping parameters to ensure global asymptotic stability and rapid transient response. Experimental validation confirms that the 11-level system achieves a THD of less than 4%, making it highly effective for gridtied renewable energy systems.
Keywords: Cascaded H-Bridge (CHB), 11-Level Inverter, Neural Networks, Incremental Passivity-Based Control (IPBC), Power Quality, THD.
FAULT DETECTION IN GEARBOX USING IoT-BASED MONITORING SYSTEM
DR. A. Sethupathy, BE, ME, PHD., Suryakumar C.K
DOI: 10.17148/IARJSET.2026.13516
Abstract: Gearboxes are critical components in industrial machinery, and their unexpected failure can result in significant operational downtime, elevated maintenance costs, and serious safety hazards. Conventional gearbox maintenance relies on periodic manual inspections that are incapable of detecting developing faults in real time. This paper presents the design and implementation of a low-cost, IoT-based fault detection system for gearboxes that integrates DS18B20 digital temperature sensors and an SW-420 vibration sensor with a NodeMCU (ESP8266/ESP32) microcontroller. The system continuously acquires temperature and vibration data, displays readings on a 16×2 LCD screen for on-site visualization, and transmits the data wirelessly to the Blynk IoT cloud platform, enabling remote monitoring via a smartphone application. Fault conditions such as overheating, excessive vibration, gear misalignment, and lubrication failure are automatically detected when sensor readings exceed predefined safety thresholds, triggering instant mobile alerts. Experimental validation confirms that the system reliably identifies abnormal operating conditions within seconds, enabling proactive maintenance intervention. The proposed system achieves a fault detection accuracy exceeding 94%, reduces manual inspection dependency, and provides a scalable and cost-effective solution for predictive maintenance in industrial environments. The total hardware and software implementation cost is estimated at ₹12,000, making it accessible for small and medium-scale industries.
ECO-FRIENDLY INDUSTRIAL AIR PURIFIER WITH SMART MONITORING
DR. A. Sethupathy, BE, ME, PhD, Rahgul Dickson M.R
DOI: 10.17148/IARJSET.2026.13517
Abstract: Industrial environments are major sources of air pollution, emitting hazardous gases, dust, fumes, and toxic chemicals that pose serious risks to worker health and workplace productivity. This paper presents the design and implementation of an Eco-Friendly Industrial Air Purifier with Smart Monitoring — an Internet of Things (IoT)-enabled embedded system capable of continuous real-time air quality monitoring and automated air purification. The system is built around an ESP32 microcontroller interfaced with an MQ135 gas sensor for detecting harmful pollutants and a DHT11 sensor for measuring ambient temperature and humidity. When measured pollution levels exceed predefined safe thresholds, the system autonomously activates a 12V industrial air filter fan via a relay module, ensuring immediate air remediation. Environmental data comprising Air Quality Index (AQI), gas concentration, temperature, and relative humidity are transmitted wirelessly to the Blynk IoT mobile application, enabling remote monitoring and informed decision-making. The entire system is powered by a stable 12V, 2Ah Switch-Mode Power Supply (SMPS). Results demonstrate that the proposed system provides low-cost, scalable, and efficient air-quality management suitable for small to medium-scale industrial settings, including welding workshops, manufacturing units, laboratories, and pharmaceutical environments. The total hardware implementation cost is approximately INR 8,000, making it highly accessible for wide industrial adoption.
Keywords: IoT, ESP32, MQ135, DHT11, Air Quality Monitoring, Industrial Air Purifier, Blynk, Smart Monitoring, Embedded Systems, Relay Module
Design and Implementation of a 10-bit FSM based Digital SAR Logic in 90 nm CMOS Technology
Keerthana K M. E, Dr. M. Santhi M.E, Ph. D
DOI: 10.17148/IARJSET.2026.13518
Abstract: In mixed-signal systems, analog to digital converters (ADCs) are crucial for transforming analog signals into digital data. The Successive Approximation Register (SAR) ADC is one of the most popular ADC architectures because of its low power consumption, moderate resolution, and straightforward hardware design, which make it appropriate for Internet of Things applications, portable electronics, and biomedical devices. The control unit that completes the successive approximation process to produce the final digital output is the digital SAR logic.
In this work, a 10-bit Finite State Machine (FSM) based digital SAR logic using 90 nm CMOS technology is designed and implemented. The binary search conversion from the Most Significant Bit (MSB) to the Least Significant Bit (LSB) is carried out by the suggested architecture.
The Cadence digital design flow, which includes synthesis, timing analysis, power estimation, and physical design processes like floor planning, placement, routing, and GDSII generation, is used to create the design. The results show low power consumption and effective area utilization. For integration in low-power SAR ADCs used in biomedical, wireless sensor, and embedded data acquisition systems, the suggested FSM-based SAR logic provides a small and energy-efficient solution.
Keywords: Successive Approximation Register (SAR), Finite State Machine (FSM), Digital SAR Logic, Verilog HDL, 90 nm CMOS Technology, Low-Power VLSI Design, Analog-to-Digital Converter (ADC).
Abstract: In today’s fast-moving manufacturing and logistics environment, efficient material handling plays a key role in improving productivity and maintaining workplace safety. Corrugated boxes are widely used for packaging and storage because they are lightweight, economical, and recyclable. However, manual handling of these boxes, especially when vertical movement between floors is required, leads to low efficiency, higher labor dependency, and increased risk of injuries , this project presents the design and analysis of a corrugated box elevator system developed to transport corrugated boxes vertically in a safe, reliable, and cost-effective manner. The system is designed by carefully considering industrial requirements such as load capacity, lifting height, speed of operation, space limitations, and safety. Major components including the supporting frame, lifting platform, guide rails, drive mechanism, motor, and transmission system are designed using standard mechanical engineering principles, all components are modeled and assembled using SolidWorks software. Structural analysis is carried out using SolidWorks Simulation to study stresses, displacements, and factor of safety under maximum loading conditions. The analysis results show that the elevator system is structurally safe and suitable for industrial use. The proposed design reduces manual effort, improves material handling efficiency, and enhances operational safety, making it ideal for small and medium scale industries.
Combating Antimicrobial Resistance Through Bacteriophage Therapy: A Targeted Therapeutic Alternative
Shruti Umare
DOI: 10.17148/IARJSET.2026.13520
Abstract: The rapid emergence of antimicrobial resistance (AMR) poses a significant threat to global public health, rendering conventional broad-spectrum antibiotics increasingly ineffective against multidrug-resistant (MDR) bacterial pathogens. This review investigates the potential of bacteriophage therapy as a targeted therapeutic alternative to combat resistant infections. Unlike traditional antibiotics, which often disrupt the host’s commensal microflora, bacteriophages exhibit high specificity toward target bacterial strains. We examine the biological mechanisms of phage-host interactions, including the lytic cycle, and evaluate the advantages of utilizing phage cocktails over monotherapy to prevent the rapid development of phage resistance. Furthermore, the integration of computational tools and bioinformatics platforms— such as PHASTER—in identifying prophage sequences and optimizing therapeutic efficacy is discussed. While bacteriophage therapy presents a promising avenue for personalized medicine, challenges related to pharmacokinetics, bacterial resistance mechanisms, and regulatory frameworks must be addressed. Ultimately, this paper highlights the necessity of continued research and clinical trials to establish bacteriophage therapy as a safe and viable strategy in the modern management of AMR.
COGNITIVE DECLINE PREDICTION: LEVERAGING AI TO DETECT ALZHEIMER’S AT EARLY STAGES
Amulya. S Chandru, Jayalakshmi. M, Sahana. S, Rakshith A.K, Deeksha K.B
DOI: 10.17148/IARJSET.2026.13521
Abstract: Alzheimer’s disease (AD) is a progressive and irreversible neurodegenerative disorder that predominantly affects the brain, resulting in the gradual deterioration of memory, cognitive functions, and behavior. It represents the leading cause of dementia, a clinical condition characterized by a significant decline in cognitive abilities that interferes with daily functioning. Despite extensive research, the precise etiology of Alzheimer’s disease remains unclear; however, it is widely accepted that a combination of genetic predisposition, environmental influences, and lifestyle factors contribute to its onset and progression.
Pathologically, Alzheimer’s disease is marked by the abnormal accumulation of extracellular amyloid-beta plaques and intracellular neurofibrillary tangles composed of hyperphosphorylated tau protein. These abnormalities disrupt neuronal communication, impair synaptic function, and ultimately lead to neuronal degeneration and cell death.
In this context, the present study aims to develop an efficient and accurate automated system for the early detection of Alzheimer’s disease using magnetic resonance imaging (MRI) of the brain. The proposed approach leverages Convolutional Neural Network (CNN) architecture to extract relevant features and perform classification, thereby facilitating improved diagnostic support and early intervention strategies.
Abstract: The discussion regarding music recommendation continues to be mentioned repeatedly since this may strengthen consumer interactions through multiple manners, especially psychic and personal memorable ones. Our project highlights a new development, that includes YOLOv8’s deep learning-driven gesture understanding. Applying a live webcam, an individual’s movements get recorded during actual time and are capable of being grouped into happy, sad, angry, neutral or surprised. While a particular mood becomes apparent, an arrangement of soundtracks will be generated according to the underlying feelings, which at first will be chosen by the database of songs that has been previously established. In light of its faster interpretation rate in addition with characteristic extraction skill, the YOLOv8 prototype has been employed for this research to figure out the movement of face. The result of the observation shows that the implemented system is efficient in identifying facial emotions with high accuracy and recommending music that aligns the current mood of a user. This study emphasizes the potential of integrating cognitive computing and Data Science methods to create intelligent multimedia applications that adapt in real-time to an individual's emotional condition.
Keywords: Emotion classification, OpenCV, music suggestion, Facial Extraction, YOLOv8, Real time capturing images.
Hydro Guard: Strengthening Public Safety Through Advanced Detection And Notification
Mr. L. Anbazhagan, Suriya Prakash M, Ranjith R, Mukesh K
DOI: 10.17148/IARJSET.2026.13523
Abstract: Ensuring public safety has become a major concern with growing urbanization. Traditional manual monitoring of surveillance cameras is inefficient and prone to human error, leading to delayed responses in critical situations. This paper presents an AI-based system for the real-time detection of violence and weapons from video streams. The proposed system utilizes YOLOv8 for accurate weapon detection and a Transformer model to analyze temporal patterns for violence detection. By integrating feature fusion, the system reduces false alarms and automatically generates alerts for immediate security response. This solution offers a scalable and efficient approach to automated public surveillance.
Keywords: Public Safety, AI-Based Surveillance, YOLOv8, Transformer Model, Weapon Detection, Violence Detection, Deep Learning, Real-Time Monitoring.
Abstract: This Forest fires represent a catastrophic threat to global biodiversity and ecological stability, necessitating the development of high-precision, real-time early warning systems. This project introduces a comprehensive monitoring framework that leverages Google’s SigLIP (Sigmoid Loss for Language Image Pre-training), a state-of- the-art Vision Transformer (ViT) architecture, to detect fire and smoke anomalies in various environmental contexts. Unlike traditional Convolutional Neural Networks (CNNs), the implemented SigLIP model utilizes global attention mechanisms to effectively distinguish between subtle visual cues, such as differentiating early-stage smoke from clouds, fog, or thermal haze.The system was fine-tuned on a diverse dataset comprising thousands of images from sources including the FLAME and DeepFire datasets, supplemented by synthetic data for edge-case training. The technical architecture is deployed through a dual-platform approach: a rapid-response Streamlit interface for interactive testing and a full-scale Flask web portal. The Flask-based application provides a production-ready environment featuring secure user authentication, an administrative dashboard for detection logging, and integrated email notification triggers via EmailJS for immediate alert dissemination.Functionally, the application supports both high-resolution static image analysis and sampled video stream processing (MP4/AVI). By utilizing confidence-based thresholding and multi-class probability mapping (Normal, Smoke, and Fire), the system provides actionable intelligence for satellite monitoring, drone surveillance, and fixed CCTV footage. The resulting solution offers a scalable, high-accuracy tool for environmental protection agencies to mitigate devastating impacts of wildfires through rapid, AI-driven detection.
Keywords: Forest Fire Detection, Deep Learning, Computer Vision, Vision Transformer (ViT), SigLIP, Smoke Detection, Fire Detection, Real-time Monitoring, Early Warning System,Transfer Learning, Flask, Streamlit,FLAME Dataset, DeepFire Dataset, Image Classification, Drone Surveillance, Environmental Monitoring.
A Spatial and Socio-Ecological Analysis of Human–Panther Conflicts and Premature Mortality of Panthers in Rajsamand District of Rajasthan
Dr. Devendra Singh Chauhan, Krishna Kanwar
DOI: 10.17148/IARJSET.2026.13525
Abstract: Human–wildlife conflict has emerged as a critical conservation challenge in human-dominated landscapes of India, particularly affecting large carnivores such as the Indian leopard (Panthera pardus). This study examines the patterns, causes, and spatial distribution of premature panther deaths and human–panther conflict in Rajsamand district, Rajasthan. Using a mixed-methods approach, the research integrates geospatial analysis, field surveys, and community- based interviews to identify conflict hotspots and underlying socio-ecological drivers. Secondary data on mortality incidents were analyzed alongside primary data collected from affected villages to assess the role of habitat fragmentation, prey depletion, and anthropogenic pressures.
The findings reveal that a significant proportion of panther deaths are linked to human activities, including retaliatory killings, road accidents, and accidental falls into open wells. Spatial analysis highlights clustering of conflict incidents near forest–agriculture interfaces and rapidly urbanizing zones. Community perceptions indicate a complex relationship characterized by fear, economic loss, and limited awareness of conservation measures.
The study underscores the need for integrated management strategies, including habitat restoration, securing open wells, strengthening compensation mechanisms, and enhancing community participation in conservation programs. By linking ecological patterns with human dimensions, the research contributes to a more nuanced understanding of coexistence challenges and offers practical recommendations for mitigating conflict and reducing premature mortality of panthers in Aravali landscape.
Confidence Level based on Logarithmic Einstein Aggregation Approach for Fuzzy Matrix Games with Multi-Expert Evaluation
PARMPREET KAUR
DOI: 10.17148/IARJSET.2026.13526
Abstract: In real world situations decision-making usually depends on the opinions of multiple experts with different level of knowledge and confidence. Namarta et al. [18] introduced a method for solving intuitionistic fuzzy matrix games using Intuitionistic Fuzzy Einstein Interactive Weighted Averaging (IFEIWA) Operator to aggregate the weightage of multiple experts for payoffs. However, a significant limitation of this method is that they assume all the experts have 100% confidence in their payoffs. But in real world competition scenarios experts have varying degrees of familiarity with specific strategies. For example; they might be very experienced yet still feel unsure about particular strategy. To address this limitation, this paper proposes a significant method for solving intuitionistic fuzzy game problems by introducing CLIFEWA (Confidence Logarithmic Intuitionistic Fuzzy Einstein Weighted Averaging) Operator. To show the superiority, validity and practical applicability of proposed method illustrative example has been given.
Intelligent Indian Sign Language Translator with Real-Time Gesture Recognition and Deep Learning
Dr. B. Aysha Banu, Mrs. A. Asrin Mahmootha, H. Mohamed Fahad Khan, K. Lokesh Krishna, K. Kartheeswaran, M. Mohamed Arshath
DOI: 10.17148/IARJSET.2026.13527
Abstract: Communication barriers between hearing-impaired individuals and the general public represent one of the most persistent challenges in inclusive society design. Indian Sign Language (ISL) serves as the primary expressive modality for approximately 18 million deaf individuals across India, yet its comprehension remains negligible among the general population. This paper presents the design, development, and rigorous evaluation of an Intelligent Indian Sign Language Translator System (ISLTS) that harnesses deep learning and computer vision to recognize hand gestures and translate them into text and synthesized speech in real time. The system employs a Convolutional Neural Network (CNN) trained on 7,500 custom ISL images augmented to 22,500 samples, achieving an overall gesture recognition accuracy of 92.4% and a mean average precision (mAP) of 0.89 across all gesture classes. MediaPipe Hands is integrated for real- time 21-point landmark detection, feeding a CNN classifier that operates at 28 frames per second on standard laptop hardware with latency below 0.5 seconds per prediction. A text-to-speech (TTS) module converts recognized gestures to audible output, enabling bidirectional communication. Comparative evaluation demonstrates that the proposed system outperforms sensor-based and earlier vision-based methods by 18–22 percentage points in accuracy while eliminating the need for specialized hardware. The proposed system offers a scalable, cost-effective, and non-intrusive solution with strong potential for deployment in educational institutions, healthcare settings, and public
Keywords: Indian Sign Language; Deep Learning; Gesture Recognition; Computer Vision; Real-Time Translation; Accessibility; Convolutional Neural Networks; MediaPipe; Text-to-Speech
Comprehensive Factor Analysis and Risk Quantification Study of Fall from Height Accidents
Jayachandran C. V, Dr. N. Dilip Raja, ME, Ph.D.
DOI: 10.17148/IARJSET.2026.13528
Abstract: Falls from height (FFH) remain one of the leading causes of fatal and severe injuries in the construction, industrial, and oil & gas sectors worldwide. Despite regulatory advancements and safety interventions, these incidents continue to pose significant challenges to occupational safety professionals. This study aims to conduct a comprehensive factor analysis and risk quantification of fall-from-height accidents to understand their root causes, contributing conditions, and effective preventive strategies.
The study employs a mixed-method approach combining incident data review, Job Hazard Analysis (JHA), behavioural safety audits, and structured interviews with safety professionals. The Factor Analysis of Incident Data (FAID) is used to identify critical causal elements categorized into human, organizational, environmental, and technical domains. Key findings indicate that over 70% of FFH incidents are linked to a combination of unsafe practices, inadequate supervision, and poor planning during work-at-height activities.
A quantitative risk matrix is developed to assign risk scores based on the frequency and severity of each contributing factor. High-risk activities, such as scaffolding erection, roof work, and temporary platform use, were assessed using Bow-Tie analysis and Failure Modes and Effects Analysis (FMEA) to identify escalation factors and opportunities for risk reduction.
Furthermore, the study integrates Human Factors Engineering (HFE) and Safety Culture Assessments to understand the behavioural patterns associated with non-compliance. The findings suggest that targeted training, competent supervision, and a robust Permit-To-Work (PTW) system significantly reduce FFH risks.
This research offers valuable insights for safety professionals and decision-makers aiming to implement evidence-based controls. It emphasizes the importance of integrating predictive analytics, risk quantification, and human-centric design into fall prevention programs to move beyond compliance and foster a resilient safety culture.
Keywords: Fall From Height, Global Statistics on Fall-Related Injuries and Fatalities, Significance of Risk Quantification for Future Prevention, Contributing Factors to Fall from Height Accidents, Regulatory and Technological Interventions.
Abstract: This paper presents a smart control and protection system for a three-phase generator using an Arduino- based platform. The system continuously monitors voltage, current, and frequency and detects abnormal conditions such as overload, overvoltage, undervoltage, and frequency variations. The system uses sensors and a microcontroller to provide real- time monitoring and protection. Experimental results show improved response time, reliability, and efficiency compared to conventional systems.
Keywords: Arduino, Generator Protection, Voltage Monitoring, Current Sensor, Smart System
Fault-Aware and Predictive Energy Management for Hybrid Energy Storage Systems in Electric Vehicles Using Mamdani Fuzzy Logic
Rakshan Pradeep K, Dr. J. Rangaraj, M.E., Ph.D.
DOI: 10.17148/IARJSET.2026.13530
Abstract: Hybrid Energy Storage Systems (HESS) integrating lithium-ion batteries with supercapacitors are increasingly adopted in electric vehicles (EVs) for dynamic power management. While Fuzzy Logic–based Energy Management Systems (EMS) effectively optimize power-split ratios under nominal operating conditions, they remain insensitive to hardware anomalies including battery overcurrent, thermal excursions, supercapacitor degradation, and converter faults. This paper presents a fault-aware intelligent EMS framework built around a Mamdani Fuzzy Inference System (FIS) that continuously monitors four sensor channels—battery voltage, current, temperature, and state-of-charge (SOC)—and classifies six distinct fault categories in real time via a dedicated Severity Index (SI ∈ [0, 1]). Upon fault detection, the controller adaptively modifies the battery duty cycle k_bat and redistributes transient power demands to the supercapacitor, preserving load continuity and system safety. MATLAB/Simulink simulations incorporating non-ideal component models, thermal dynamics, and converter losses demonstrate a 30% reduction in peak battery current, a 29% decrease in thermal rise (ΔT), and a 20% improvement in SOC retention relative to a conventional HESS without fault awareness. DC bus voltage stability (MAD = 8 V) is fully maintained across all injected fault scenarios. The proposed framework bridges the critical gap between energy optimization and hardware fault management in HESS for EV applications.
Keywords: Hybrid Energy Storage System (HESS), Mamdani Fuzzy Inference System, Fault Detection and Classification, Energy Management System, Battery State-of-Charge, Supercapacitor, Thermal Management, Electric Vehicles, DC-DC Converter, Severity Index
A Comprehensive Review of IoT-Enabled Smart Traffic Management System Using Raspberry Pi
Ms. Dipali Siddheshwar Pawar, Prof. Kavita H. Waghmode
DOI: 10.17148/IARJSET.2026.13531
Abstract: Managing traffic efficiently has become one of the most urgent challenges in today’s rapidly growing cities. Traditional fixed-time traffic signals often fall short because they cannot react to changing road conditions, which leads to unnecessary delays, increased fuel consumption, and higher pollution levels. Over the past few years, the combination of Internet of Things (IoT) technologies, edge computing, cloud platforms, and artificial intelligence has opened new possibilities for creating more adaptive and responsive traffic systems. This review brings together recent research from 2020 to 2025 and examines how sensors, embedded devices, and communication networks are being used to monitor real-time traffic flow, prioritize emergency vehicles, and optimize signal timing. The paper also explores advanced methods such as deep reinforcement learning, computer vision–based vehicle detection, blockchain-secured IoT frameworks, federated learning, and digital twin simulations. By comparing these approaches, the review highlights both their strengths and the remaining challenges that need attention. Overall, the study emphasizes that IoT-enabled smart traffic systems especially those combining edge intelligence with cloud analytics offer a practical and scalable pathway toward safer, cleaner, and more efficient urban mobility.
Keywords: Internet of Things (IoT), Smart Traffic Management, Edge Computing, Adaptive Signal Control, Vehicle Detection, Emergency Vehicle Priority, Cloud IoT Platforms, Traffic Flow Prediction, Intelligent Transportation Systems (ITS), Deep Learning, Reinforcement Learning, Urban Mobility.
An AI-Enabled Crop, Fertilizer, And Yield Recommendation System Using Machine Learning
Nayana M P, Mamatha D S, Madhushri, Harshitha Y, Mona M
DOI: 10.17148/IARJSET.2026.13532
Abstract: Agriculture plays a vital role in global food security, and farmers constantly seek ways to optimize crop selection to maximize yield and profitability. However, identifying the most suitable crop for a specific region is challenging due to factors such as climate conditions, soil fertility, rainfall, temperature, and water availability. Traditional farming practices often rely on experience and assumptions, which may lead to reduced productivity and improper fertilizer usage.The proposed Crop and Fertilizer Recommendation System addresses these challenges by utilizing machine learning techniques to analyze environmental and soil-related parameters for intelligent decision- making. The system collects input data such as soil nutrients, temperature, humidity, rainfall, and pH values, and processes them using machine learning algorithms to recommend the most suitable crop and appropriate fertilizer. By providing accurate and data-driven recommendations, the system helps farmers improve crop yield, reduce resource wastage, and enhance sustainable agricultural practices.The developed model aims to support precision agriculture by assisting farmers in selecting crops that are best suited for their land conditions while also suggesting fertilizers to maintain soil health and productivity. Experimental results demonstrate that machine learning-based recommendations can significantly improve agricultural efficiency and contribute to smarter farming solutions.
A REVIEW ON MACHINE LEARNING MODEL FOR AUTOMATIC HEART DISEASE PREDICTION
Ms. Komal Suryakant Kambale, Prof. Namdev M. Sawant
DOI: 10.17148/IARJSET.2026.13533
Abstract: Heart disease is a leading cause of mortality worldwide, necessitating early detection and prevention strategies. Machine learning (ML) models have emerged as powerful tools for automatic heart disease prediction. This review paper provides an overview of the recent advancements in ML-based approaches for heart disease prediction. We begin by discussing the significance of early detection and the potential of ML in this domain. Next, we conduct a comprehensive literature survey, summarizing the key findings from previous studies. We then present a comparative study of various ML algorithms commonly used for heart disease prediction, highlighting their strengths and limitations. Additionally, we outline the proposed procedure for developing ML models for heart disease prediction and discuss potential future directions. Finally, we conclude by emphasizing the importance of continued research in this area to improve the accuracy and accessibility of automatic heart disease prediction.
DEVELOPMENT OF REMOTE CONTROLLED MOTORCYCLE HELMET WIPER
CHRISTIAN JAY B. MACARIO
DOI: 10.17148/IARJSET.2026.13534
Abstract: This study focused on the design, development, and evaluation of a remote-controlled motorcycle helmet wiper intended to enhance rider visibility and safety during rainy conditions. Specifically, the study aimed to describe the technical features of the device, determine its sensitivity in terms of wiping speed, assess its effectiveness in clearing raindrops under varying rainfall conditions, and evaluate its acceptability in terms of technical features, composition, operating performance, and safety. A developmental research design was employed in the conduct of the study. The developed device consisted of a compact wiper mechanism mounted on the helmet visor and connected to a lightweight 3D-printed housing containing the power supply, rechargeable battery, RF relay receiver, and control circuitry. A flexible spring wire was incorporated to allow smooth visor movement while minimizing strain on electrical connections. The system operated in two modes, slow and continuous, controlled through a handheld wireless remote. The evaluation involved fifty (50) evaluators composed of motorcycle riders, instructors, and individuals with expertise in mechanical, electrical, and safety-related fields. Data were gathered through structured observation, controlled performance testing, and an acceptability evaluation instrument. Results showed that the developed helmet wiper was functionally integrated and suitable for helmet application. In terms of sensitivity, the device recorded an average wiping speed of 30.66 wipes per minute in slow mode and 75 wipes per minute in continuous mode. The device demonstrated effective raindrop removal under light and moderate rainfall in both modes of operation; however, optimal performance during heavy rainfall was observed only when operated in continuous mode. Moreover, the device obtained a “Very Acceptable” rating across all evaluated criteria, indicating positive user perception of its design, durability, operating performance, and safety. Overall, the findings demonstrate that the developed remote-controlled motorcycle helmet wiper is a functional, safe, and user-acceptable safety device with strong potential for practical application in improving rider visibility during adverse weather conditions.
A Study on the Impact of Digital Fatigue and Cognitive Load on Employee Productivity and Work Life Balance with reference to IT Sector in Chennai
Dr G Balamurugan, Joans Balscia V
DOI: 10.17148/IARJSET.2026.13535
Abstract: This study examines the impact of digital fatigue and cognitive load on employee productivity and work–life balance in the IT sector in Chennai. In today’s technology-driven work environment, employees are increasingly exposed to prolonged screen time, multitasking and constant digital connectivity, which often lead to mental exhaustion and reduced efficiency. The study adopts a descriptive research design and uses primary data collected from 113 IT employees through a structured questionnaire using the snowball sampling technique. Secondary data from journals and research articles support the study. Statistical tools such as descriptive statistics, correlation, regression and factor analysis were employed for data analysis. The findings reveal that digital fatigue and cognitive load have a significant negative impact on both employee productivity and work–life balance. Key contributing factors include information overload, continuous virtual meetings and lack of adequate digital breaks. These issues lead to decreased concentration, increased stress and difficulty in managing personal and professional life. The study suggests that practices such as regular digital breaks and better digital management strategies can improve employee well-being, productivity and overall work performance.
Keywords: Digital Fatigue, Cognitive Load, Employee Productivity, Work–Life Balance, IT Sector.
AN EFFICIENT DEEP NEURAL NETWORK APPROACH FOR DIABETES PREDICTION
Dr. Rajendra Prasad Banavathu, Dr.S. Jayaprada, Dr. Kalpana Devi Bai Mudavathu
DOI: 10.17148/IARJSET.2026.13536
Abstract: Millions of people all over the world endure from the incessant condition of Diabetes. Early detection and action can lower the likelihood of problems and assist avoid or delay its development. Diabetes has been predicted using machine learning algorithms using a variety of characteristics, including demographics, clinical data, and lifestyle factors. Using a mix of patient data, including age, body mass index and more we present an approach based on deep learning to predict the chance of acquiring diabetes. K Nearest Neighbor(KNN), Logistic Regression(LR), Support Vector Machine(SVM), Decision Tree(DT) and Random Forest(RF), Deep Neural Networks (DNN) are some of the algorithms used. Each algorithm's accuracy is calculated along with the model's accuracy. The approach with a high accuracy level is used as the model to predict diabetes. This strategy may help medical professionals make knowledgeable judgements and give patients personalized care. A number of metrics, such as accuracy and F1 score, are used to assess the effectiveness of the suggested model. Using deep learning concepts by training the properties of a deep neural network(DNN), we suggest a method for diagnosing diabetes. with 98.49% prediction accuracy, and 93% F1 Score. The experimental findings show that when using Deep learning approach, the suggested system offers good outcomes. This strategy may help medical professionals make knowledgeable judgements and give patients personalized care.
Vehicle Classification and Traffic Density Analysis using RT-DETR
Prajwal.M, Sujayeendra Rao, Dr. Ananth. G. S
DOI: 10.17148/IARJSET.2026.13537
Abstract: Traffic congestion and vehicle monitoring have become major challenges in modern urban transportation systems, requiring efficient real-time traffic analysis and management solutions. This project presents an intelligent Vehicle Classification and Traffic Density Analysis System using the advanced RT-DETR deep learning architecture for fast and accurate vehicle detection in traffic environments. Unlike traditional CNN-based models, RT-DETR uses transformer-based attention mechanisms to improve detection accuracy in crowded and complex road conditions. The system can identify multiple vehicle categories such as cars, buses, trucks, motorcycles, bicycles, and auto-rickshaws from images and videos. A dual-platform deployment using Flask and Streamlit supports both testing and large-scale monitoring applications. The application includes secure user authentication, traffic monitoring dashboards, and automated traffic reporting features. Uploaded traffic videos are processed using frame-based inference techniques to estimate traffic density levels such as Low, Medium, and High. Confidence-based filtering helps reduce false detections caused by occlusion, lighting variations, and dense traffic conditions. The system utilizes OpenCV and PyTorch for efficient image and video processing in real-time environments. Experimental results show that the RT- DETR model provides high detection accuracy, stable performance, and an effective AI-driven solution for smart traffic monitoring and intelligent transportation systems.
Keywords: Vehicle Classification, Traffic Density Analysis, RT-DETR, Deep Learning, Computer Vision, Intelligent Transportation System, Object Detection, Real-Time Monitoring, Flask, Streamlit, OpenCV, PyTorch, Traffic Surveillance, Smart City, Vehicle Detection.
Design and Circuit-Level Analysis of Low-Power Analog Neural Networks in CMOS Technology
Madhubala R, Dr. J. Rangaraj
DOI: 10.17148/IARJSET.2026.13538
Abstract: This paper presents the circuit-level design and performance analysis of a low-power Analog Neural Network (ANN) implemented in standard CMOS technology, targeting energy-constrained biomedical and edge-AI VLSI applications. The proposed architecture realises ANN inference entirely in the continuous-time analog domain using four principal circuit primitives: a Current Correlator (CC), an Adaptive Differential Equaliser (ADEL), a Gaussian Activation Function Circuit, and a Synaptic Function Circuit (SFC). All circuits are designed and characterised in Cadence Virtuoso. Simulation results confirm a peak-to-peak differential voltage gain of 2.928×, a −3 dB bandwidth of 15.89 GHz, and near-unity Gaussian voltage transfer (gain ≈ 1.000) with a current gain of 1.266× under low-supply conditions. Comprehensive transient, DC, and AC analyses validate stable, linear operation across the expected operating range. The work establishes a quantitative performance baseline for future integration of SFC and comparator stages toward a fully functional on-chip analog ANN classifier.
Keywords: Analog Neural Network, CMOS VLSI, Low-Power Design, Current Correlator, ADEL, Gaussian Activation Function, Synaptic Function Circuit, Cadence Simulation, Sub-threshold Operation, Edge AI.
DESIGN OF A MIXED SIGNAL VCO BASED ADC FOR HIGH SPEED APPLICATIONS
Akbar Ali A, Dr. O.Saraniya
DOI: 10.17148/IARJSET.2026.13539
Abstract: This paper demonstrates a power-efficient implementation of a mixed-signal Analog-to-Digital Converter (ADC) based on a Voltage-Controlled Oscillator (VCO). Conventional Flash ADCs face significant limitations in power and area at multi-GS/s speeds as comparator counts grow exponentially. The proposed hybrid Flash-VCO architecture overcomes these challenges by shifting fine quantization into the time domain. A current-starved ring oscillator scheme is employed for the VCO to achieve high efficiency and CMOS scalability. Simulation results confirm that this approach reduces power consumption and mismatch sensitivity, making it suitable for high-speed, low-resolution biomedical and wireless applications.
Implementation of HMI-Based Gesture Recognition and UWB Radar in Autonomous Vehicles
Shrisanjaykumaar K, Dr. O. Saraniya
DOI: 10.17148/IARJSET.2026.13540
Abstract: This paper presents the MATLAB simulation and implementation of a 2–6 GHz CMOS Ultra-Wideband (UWB) radar transceiver front-end designed in 45 nm technology for HMI-based gesture recognition in autonomous vehicles. The transmitter chain employs digital pulse generation (5 ns rectangular pulse), Gaussian pulse shaping for spectral compliance, a Digitally Controlled Oscillator (DCO) providing a 4 GHz carrier, and an up-conversion mixer producing an RF output at 4.5 GHz. The received signal is processed via matched-filter correlation for range estimation, CA-CFAR detection for robust target identification, and a Kalman-filter-based tracker for long-range target following. An 8-gesture recognition vocabulary is implemented, with each gesture mapped to a specific vehicle command. Simulation results confirm FCC Part 15 UWB spectral compliance, accurate range detection at 45.5 m, multi-target resolution of pedestrian-car scenarios at 30–32 m separation, and gesture detection with a sub-5 ns observation window. The system achieves low-power, integrated radar-based HMI suitable for next-generation autonomous vehicles.
THE EFFECT OF ACTIVATION TEMPERATURE ON PORE DIAMETER AND ADSORPTION CAPACITY OF ACTIVATED CARBON DERIVED FROM LIGNOCELLULOSIC BIOMASS: A REVIEW
I Gusti Agung Kade Suriadi, Dewa Ngakan Ketut Putra Negara, I Ketut Adi Atmika, Tjokorda Gde Tirta Nindhia*
DOI: 10.17148/IARJSET.2026.13541
Abstract: Activated carbon derived from lignocellulosic biomass has attracted significant attention due to its potential application in wastewater treatment. This review examines the effect of activation temperature on pore structure development and methylene blue adsorption capacity of activated carbon produced from teak sawdust waste. Activation temperature plays a crucial role in determining pore size distribution, surface area, and adsorption performance. Based on previous studies, increasing activation temperature promotes the transformation of micropores into mesopores due to enhanced devolatilization and gasification reactions. This structural evolution improves the adsorption capacity for methylene blue, a large organic molecule commonly used as a model adsorbate. However, excessively high activation temperatures may lead to pore collapse and reduced adsorption efficiency. This review highlights the importance of optimizing activation temperature to achieve a balance between pore development and structural stability. The findings provide insights into the design of efficient biomass-based adsorbents for environmental applications.
Hybrid Machine Learning For IoT-Driven Heart Health Prediction
Priyanka Vijay Adate, Prof. A. A. Bhise
DOI: 10.17148/IARJSET.2026.13542
Abstract: The rapid growth of intelligent healthcare technologies has enabled continuous monitoring of human health through connected devices and smart sensors. Heart-related illnesses require immediate attention and timely analysis because delayed diagnosis may lead to severe complications. This research presents a smart IoT-based framework designed for early heart risk identification using adaptive learning and sensor-driven analytics. The system acquires real-time physiological information through ECG, pulse oximeter, and temperature sensors connected to an embedded microcontroller platform. Unlike traditional healthcare prediction systems that depend completely on pre-labeled datasets, the proposed framework introduces an adaptive learning strategy capable of dynamically categorizing incoming health data patterns. Processed sensor readings are analyzed using Artificial Neural Network (ANN) and Random Forest techniques to determine the probability of abnormal heart conditions. The proposed approach improves prediction consistency, supports continuous monitoring, and minimizes dependency on cloud-based infrastructure. Experimental evaluation demonstrates that the framework provides reliable prediction performance with improved adaptability for real-time healthcare environments.
Equivalent Circuit Model-Based State of Charge Estimation of Lithium-Ion Batteries Using Kalman Filter Algorithms
Thirumalai V M. E, Dr. A. Anitha M.E, Ph. D
DOI: 10.17148/IARJSET.2026.13543
Abstract: Estimation of accurate SoC of Li-ion batteries is the basic requirement for a safe and efficient operation of the Battery Management System (BMS) of EVs. A systematic method for SoC estimation using a second order ECM (2RC ECM) calibrated from the data of HPPC tests is proposed in this paper. The OCV-SoC relation is extracted from C/20 charge-discharge test and fitted by a seventh order polynomial function. Parameters of ECM passive components (R₀, R₁, C₁, R₂, C₂) are identified from ten discrete SoC values with Levenberg-Marquardt (LM) nonlinear least-square algorithm, then they are represented by seventh order polynomials with regard to the SoC. EKF and UKF, both recursive Bayesian estimators, are implemented and evaluated for Turnigy Graphene 4.6928 Ah lithium cell at 0C with the standard test profiles (C/20 charge-discharge and HPPC). Error estimation (RMSE, MAE and MAX) is used for SoC and terminal voltage estimation respectively. Experiments show that both filters can always converge at the conditions and the UKF is slightly better than EKF at the SoC estimation under most of drive profiles due to the avoid of Jacobian linearisation. The sensitivity of SoC estimation accuracy with regard to parameters' identification accuracy is discussed and future works are summarized as adaptive noise covariance setting and order elevation to 3RC model.
Keywords: State of Charge (SoC), Equivalent Circuit Model (ECM), Extended Kalman Filter (EKF), Unscented Kalman Filter (UKF), HPPC, Levenberg–Marquardt, Battery Management System (BMS), Electric Vehicle (EV).
Abstract: Yoga is an ancient Indian practice that combines physical postures, breathing exercises, meditation, and relaxation techniques to improve overall well-being. In recent years, yoga has gained worldwide recognition as an effective method for improving both mental and physical health. This research paper examines the effects of yoga on mental stability, stress reduction, emotional balance, flexibility, cardiovascular health, and overall physical fitness. The study also highlights the importance of incorporating yoga into daily life for students, working professionals, and elderly individuals. The findings suggest that regular yoga practice significantly contributes to a healthier lifestyle by reducing anxiety, depression, obesity, hypertension, and lifestyle-related diseases.
Abstract: Oral cancer is one of the most common and life-threatening cancers worldwide, particularly in developing regions where tobacco consumption, alcohol usage, and poor oral hygiene are prevalent. Early detection of oral cancer significantly improves survival rates; however, conventional diagnostic methods rely heavily on clinical examination and biopsy procedures, which are invasive, time-consuming, and dependent on expert availability. This project presents an intelligent Oral Cancer Detection System using machine learning and deep learning techniques to assist in early screening. The proposed system employs a Convolutional Neural Network (CNN) for analyzing oral cavity images to classify them as cancerous or non-cancerous. In addition, a machine learning-based clinical prediction model evaluates patient risk factors such as age, tobacco usage, alcohol consumption, and the presence of oral lesions. By integrating image-based analysis with clinical data evaluation, the system enhances diagnostic reliability and decision support. The developed models are deployed using a Streamlit-based web application that allows users to upload oral images and enter clinical details for real-time prediction. Experimental results demonstrate that the image-based model achieves high classification accuracy, while the clinical model effectively supports risk assessment. The proposed system provides a non-invasive, cost effective, and user-friendly solution for preliminary oral cancer screening, aiming to support healthcare professionals and improve early detection outcomes.
Keywords: deep learning, oral cancer detection, oral cancer, risk factors, cancer remedies, hospitals suggested.
An IoT-Based Smart Smoke Alarm System Using Multi-Sensor Fusion and Intelligent Monitoring for Enhanced Fire Safety
K.Venkatesh, V.Ramesh
DOI: 10.17148/IARJSET.2026.13546
Abstract: Early detection of fire hazards is essential to minimize loss of life and property in residential and commercial buildings. Traditional smoke alarm systems tend to use a operating system that is based on a single sensor, resulting in a large false alarm rate and lack of situational awareness. In this paper, a smart smoke detector, implemented using Internet of Things (IoT) and multi-sensor fusion, real-time visual and intelligent alert mechanisms to enhance reliability in detecting fire are introduced. The proposed system has a multiple sensors that detect smoke and fire related parameters, thus minimizing false alarms, an integrated web camera, and operated by a servo motor to do a live monitoring of the environment. The predefined floor plan achieves fire localization of the source of fire, which facilitates the evacuation and fire fight work. The functional reliability and practical feasibility of the system is experimentally proven. The proposed solution can be scaled and applicable to implementation in daily residential and commercial locations and is a large improvement over the traditional fire alarm systems.
Keywords: Internet of Things (IoT), Smart Smoke Detector, Fire Safety System, Multi-Sensor Fusion, Intelligible Monitoring, Smart Buildings.
A Study on Non-Performing Assets in Selected Non-Banking Financial Companies
Dr. B. Asha Daisy, Dinesh Krishna V
DOI: 10.17148/IARJSET.2026.13547
Abstract: Non-Banking Financial Companies are integral to Indian economy as they are in a position to provide loans to areas which might not have access to banks or have not been receiving due attention from the banks. Some risks that NPAs have with regard to smooth operation and long-term survival of NBFCs. The risks originating from NPAs on the economical and efficient functioning of an NBFC come in different forms. In this research paper we would like to discuss the reasons of NPAs and the influence that NPAs have on the operations of the Non-Banking Financial Company. Various reasons for NPAs could include lack of rigorous credit appraisal system, over reliance on select few industries and macroeconomic factors.
Design and Analysis of Digitally Trimmable Sub – Bandgap Reference in CMOS Technology
Harshavardhini R, Dr. M. Santhi
DOI: 10.17148/IARJSET.2026.13548
Abstract: This paper presents the design and implementation of a low-power digitally trimmable CMOS bandgap reference (BGR) circuit intended for high-precision and temperature-stable voltage reference generation in modern VLSI systems. The performance of conventional bandgap reference circuits is significantly affected by process variations, device mismatch, and temperature fluctuations, resulting in reduced output accuracy and long-term stability. To address these limitations, a digitally controlled trimming architecture employing a binary-weighted resistor array integrated with CMOS switching logic is proposed. The trimming network enables fine adjustment of the reference voltage and compensates for fabrication-induced variations, thereby enhancing the overall circuit accuracy and thermal stability. The proposed BGR circuit is designed with emphasis on low power dissipation, compact implementation, and reliable operation over a wide temperature range. Detailed simulation analysis is carried out to evaluate key performance parameters including reference voltage stability, temperature coefficient, line sensitivity, and power consumption under varying operating conditions. Simulation results demonstrate that the proposed digitally trimmable architecture achieves improved voltage precision and reduced temperature dependency when compared with conventional CMOS bandgap reference circuits, making it suitable for low-power analog and mixed- signal integrated circuit applications.
Keywords: Bandgap Reference (BGR), CMOS Technology, Digital Trimming, Low Power VLSI, Voltage Reference Circuit, Binary-Weighted Resistor Array, Temperature Compensation, Analog Integrated Circuits, Process Variation, Low Power Design, CMOS Switches, Mixed-Signal IC Design.
Abstract: Yoga posture correction and recognition systems can help learners practice safely and consistently by providing instant feedback without requiring continuous supervision from an instructor. This project proposes a deep learning–based Yoga Posture Detection system that identifies yoga poses from images and real-time webcam video. The system uses a convolutional neural network (CNN) based classifier trained on a labeled dataset of yoga postures, where each class corresponds to a specific asana. For real-time operation, YOLOv8 is used to detect the person in each frame, the detected region is cropped, and the posture is classified using the trained model. The classification model is trained using transfer learning (MobileNetV2 backbone) to improve accuracy with a limited dataset and reduce training time. The final system is deployed as a web application using Flask with a user-friendly interface built using HTML, CSS, and JavaScript, allowing users to upload images for posture prediction and view top confidence results. Experimental results show that the proposed model achieves around 70% validation accuracy over 41 yoga classes, and performance is analyzed using a confusion matrix, classification report, and Grad-CAM visual explanations. The solution demonstrates an end-to-end pipeline for yoga pose classification and real-time detection, and can be extended further for posture correction and fitness guidance.
