Advancing Healthcare through Artificial Intelligence: The Role of Association Rule Mining in Clinical Decision Support and Healthcare Analytics

Authors: Geofrey Nyabuto, Charles Kibet Ng'etich, Edwin Seno, George Kihara Mburu, Marion Jeptoo, Joyadams Munene, Muriithi Alex Karani, John kimani Muragu, Nyairo Charles Magati

Abstract: Association rule mining (ARM) is a data mining approach used to discover frequent co-occurrence patterns and conditional relationships in large datasets. In healthcare, ARM has been applied to electronic health records, claims databases, laboratory data, prescription data, disease registries, and public health datasets to reveal clinically meaningful patterns that may support diagnosis, medication safety, risk stratification, and service planning. Objective: This review synthesizes how ARM has been applied in healthcare, focusing on methods, clinical application areas, implementation challenges, and future research directions. A systematic review design guided by PRISMA 2020 was used to structure the manuscript. Literature was organized around peer-reviewed ARM studies in healthcare, including clinical decision support, diagnostic test ordering, disease-medication association mining, adverse drug reaction signal detection, risk factor discovery, hospital readmission analysis, privacy-preserving mining, and emerging causal or hybrid ARM approaches. The literature shows that Apriori remains the most frequently used ARM algorithm, although FP-Growth, weighted Apriori, class association rules, negative association mining, privacy-preserving ARM, and causal irredundant ARM are increasingly used to address computational, interpretability, privacy, and clinical validity limitations. ARM is valuable because it produces transparent IF-THEN rules that clinicians can inspect, but uncontrolled rule generation, weak validation, data quality limitations, and spurious associations remain major barriers. ARM has clear potential in healthcare knowledge discovery and decision support, particularly where interpretability is required. Future research should prioritize external validation, clinician-centered rule evaluation, integration with electronic medical records, explainable hybrid models, privacy-preserving analytics, and evidence from low- and middle-income healthcare settings.

DOI: https://doi.org/10.5281/zenodo.20199703

Design And Performance Analysis Of An Automated Guidance System For Tractors For Precision Farming

Authors: Mr. Sangharsh Dongare, Dr. Vijayashri Mohobiya

Abstract: The advancement of precision agriculture has led to the adoption of automated systems that enhance operational accuracy and reduce resource wastage. Among these technologies, automated tractor guidance systems play a crucial role in improving field efficiency and minimizing human intervention. This research focuses on the design and development of an integrated automated steering system for agricultural tractors using Global Navigation Satellite System (GNSS) technology. The proposed system combines an RTK-enabled GNSS receiver, inertial sensors, an electronic control unit, and an electro-hydraulic steering actuator to enable accurate trajectory tracking. The design methodology emphasizes system integration, component selection, and control architecture development. The performance of the system is evaluated through design validation parameters such as steering response, path deviation, and system stability. The study demonstrates that the designed system can achieve high precision in field navigation while maintaining robustness under varying operating conditions. The findings highlight the importance of system-level design in developing efficient and reliable automated agricultural machinery.

EV Charging Locator System: A Web-Based Real-Time Solution For Smart Electric Vehicle Charging Infrastructure

Authors: Swapnil Pradip Jadhav

Abstract: – The rapid proliferation of electric vehicles (EVs) has created an urgent need for scalable, accessible, and real-time charging infrastructure discovery systems. This paper presents the design, development, and evaluation of a web-based EV Charging Locator System that enables EV users to identify, navigate to, and interact with nearby charging stations using geolocation services, interactive mapping, and a responsive user interface. The proposed system addresses the critical gap of fragmented charging network information by consolidating multi-network station data on a single unified platform. The system is built using modern web technologies including HTML5, CSS3, JavaScript (ES6+), and integrates third-party APIs such as Google Maps and Open Charge Map. A user-friendly dashboard, smart filtering, and station management module empower both EV users and charging station administrators. Experimental evaluation demonstrates reduced search time, improved navigation accuracy, and enhanced user experience compared to existing solutions. The system contributes significantly to smart city goals and the green energy transition.

DOI: https://doi.org/10.5281/zenodo.20205640

Streamlining The Code Review Process Using Artificial Intelligence: A Practical Framework For Enhancing Software Quality And Development Efficiency

Authors: Neha Asthana

Abstract: The rapid evolution of modern software engineering practices has intensified the demand for faster development cycles, higher code quality, and improved operational efficiency. Traditional code review processes, while essential for maintaining software reliability and security, frequently create development bottlenecks due to the extensive manual effort required to validate syntax, formatting, and compliance with coding standards. This article presents a practical implementation framework for streamlining code review operations through the adoption of artificial intelligence (AI)-assisted analysis integrated within DevOps workflows. The initiative focused on incorporating AI-assisted review capabilities into existing Git-based pull-request and continuous integration pipelines to automate repetitive review activities and accelerate validation processes. By enabling automated identification of syntax inconsistencies, formatting deviations, boilerplate inefficiencies, and commonly recurring coding issues, the framework allowed software reviewers to prioritize higher-value technical concerns including architectural integrity, security vulnerabilities, scalability, and business logic validation. The implementation demonstrated measurable operational improvements, including an approximate 30% reduction in pull-request review time and a 35% decrease in post-review rework across development teams. The study further examines the technical and organizational challenges associated with integrating AI into enterprise software development practices. One major challenge involved the generation of inconsistent or contextually irrelevant AI recommendations that occasionally conflicted with project-specific coding patterns and domain-specific business requirements. This limitation was addressed through iterative prompt refinement, enforcement of internal engineering standards, and selective application of AI to repetitive development tasks such as validation routines and boilerplate generation. Security and code quality concerns also emerged due to the potential introduction of insecure coding patterns or software anti-patterns through AI-generated suggestions. To mitigate these risks, the framework incorporated layered governance mechanisms including static code analysis, automated security scanning, peer validation procedures, and mandatory human review for sensitive components. An additional barrier to adoption stemmed from initial developer skepticism regarding the reliability and contextual awareness of AI-generated outputs. Adoption rates improved significantly after repositioning AI as an assistive pre-review mechanism rather than a replacement for human expertise. Continuous peer validation, governance-based coding standards, and collaborative review practices contributed to increased organizational trust and broader engineering acceptance.

Consumer Buying Behaviour: A Study Of Factors Influencing Purchase Decisions

Authors: Taranveer Singh, Manisha Kalra

Abstract: Consumer buying behaviour refers to the process through which individuals select, purchase, use, and dispose of goods and services to satisfy their needs and wants. Understanding consumer behaviour is essential for businesses to design effective marketing strategies and achieve customer satisfaction. This research paper examines the major factors influencing consumer buying behaviour, including cultural, social, personal, and psychological factors. The study also highlights the stages involved in the consumer decision-making process and the impact of digital marketing on purchasing decisions. The paper concludes that changing lifestyles, technological advancements, and increased access to information have significantly transformed consumer buying patterns in modern markets.

Restaurant Sales Intelligence Report In Tableau

Authors: K.Ravindhar, P.KamalaKar, Dr.Diana Moses

Abstract: This analysis examines a restaurant chain sales dataset covering transaction recorded across November and December 2022, spanning five major European cities — London, Lisbon, Madrid, Berlin, and Paris. The dataset capture nine key attributes including Order ID, Date, Product, Price, Quantity, Purchase Type, Payment Method, Manager, and City, forming a structured foundation for evaluating multi-dimensional sales performance. Burger emerge as the highest revenue-generating product, followed by Fries and Chicken Sandwiche, while Sides & Other contributes the least to overall revenue. Geographically, Lisbon leads in both total revenue and quantity sold, closely followed by London, whereas Berlin and Paris record comparatively lower performance. Purchase behavior analysis reveals that Online transaction are the most frequently used channel, followed by Instore and Drive thru. In terms of payment preferences, Credit Card dominates as the most widely adopted method, ahead of Cash and Gift Card usage. These finding highlight key revenue regional performance disparities, and customer purchasing preferences, offering a data-driven foundation for improving inventory planning, targeted marketing strategies, and overall operational decision-making.

Mobile Marketing & Consumer Engagement: Evidence From Indias Digital Economy

Authors: Jyotish Kumar, Dr. Meentu Grover

Abstract: India's mobile economy is no longer a future promise — it is today's commercial reality. This study examines the impact of mobile marketing on consumer engagement with empirical evidence drawn from 150 respondents across Punjab, India. Adopting a descriptive-analytical research design and administering structured questionnaires, the study investigates how SMS campaigns, social media advertisements, push notifications, mobile applications, and influencer marketing shape consumer purchase intentions, brand loyalty, and overall satisfaction. Statistical techniques including descriptive analysis, correlation, regression, and chi-square testing were employed to interrogate the data. Results confirm that mobile marketing exerts a statistically significant positive effect on consumer engagement (β = 0.61, p < 0.001), with social media marketing and personalized notifications emerging as the strongest predictors. Brand trust and perceived usefulness were identified as critical mediating constructs. These findings carry decisive implications for digital marketers, policy designers, and scholars navigating India's fast-evolving mobile-first marketplace.

Microfinance And Poverty Reduction: An Empirical Study Of Financial Inclusion And Rural Development.

Authors: Jatinder Singh, Manisha Kalra

Abstract: – Microfinance has emerged as an important financial instrument for reducing poverty, promoting self-employment, and improving the socio-economic conditions of low-income households. The concept of microfinance involves providing small loans, savings facilities, insurance, and other financial services to economically weaker sections who are generally excluded from the formal banking sector. This research paper examines the role of microfinance in poverty reduction by analyzing its impact on income generation, employment opportunities, women empowerment, and rural development. The study also highlights the challenges faced by microfinance institutions (MFIs) and suggests measures for improving their effectiveness. The paper concludes that microfinance plays a significant role in poverty alleviation when combined with proper training, financial literacy, and government support.

Role Of YouTube Marketing In Influencing Consumer Buying Behaviour

Authors: Mandeep Kaur, Dr joe Christy N

Abstract: The rapid growth of social media platforms has transformed consumer buying behavior, with YouTube emerging as one of the most influential digital marketing channels. This research paper examines the impact of YouTube on consumer purchase decisions through a data-oriented approach. The study analyzes how YouTube advertisements, influencer marketing, product reviews, unboxing videos, and user-generated content affect consumer awareness, purchase intention, and actual buying behavior. Secondary data from recent empirical studies and industry reports are used to evaluate consumer responses toward YouTube-based marketing strategies. The findings reveal that YouTube significantly influences consumer decision-making, particularly among younger audiences such as Generation Z and Millennials. Factors such as content credibility, emotional engagement, influencer trustworthiness, and video quality positively affect purchase intentions. The study concludes that YouTube has become a powerful tool for marketers in shaping consumer attitudes and driving purchasing decisions.

Role of Training and Development Policies in Employee Competence in Organizations

Authors: Prachi Barnwal, Dr Navneet Seth

Abstract: Training and development policies play a crucial role in improving employee competence, organizational productivity, and overall business performance. In the modern competitive environment, organizations increasingly invest in employee training programs to enhance technical skills, communication abilities, leadership qualities, and job efficiency. The present study examines the impact of training and development policies on employee competence using a data-oriented approach. The study is based on secondary data collected from research journals, HR reports, and organizational studies. The findings reveal that effective training policies significantly improve employee skills, motivation, productivity, and job satisfaction. The study concludes that organizations with strong training and development practices achieve higher employee performance and organizational effectiveness.

A Study On The Impact Of Social Media Marketing On Consumer Buying Behaviour

Authors: Salama Juma Shehe, Dr. Sahil Nazir

Abstract: This study examined the impact of social media marketing on consumer buying behavior. The rapid growth of digital technology and social networking platforms has transformed the way businesses communicate with customers and promote their products and services. The main objective of the study was to analyze how social media marketing influences consumer purchasing decisions. Specifically, the study examined the role of social media advertisements, online reviews, influencer marketing, and promotional content in shaping consumer buying behavior. The study employed a quantitative research design using a survey method. Data were collected through an online questionnaire created using Google Forms and distributed through social media platforms. A total of sample size 100 respondents participated in the study. The collected data were analyzed using descriptive statistics, including frequencies and percentages, and presented using tables and charts. The findings indicated that social media marketing significantly influences consumer buying behavior. The results showed that most consumers rely on social media platforms to obtain information about products, read reviews, and compare alternatives before making purchase decisions. The study concludes that social media marketing plays a critical role in influencing modern consumer purchasing behavior. Businesses should therefore invest in effective social media marketing strategies to improve brand visibility, customer engagement, and sales performance. The study recommends that companies should enhance their social media presence, collaborate with influencers, and provide reliable and engaging content to attract and retain customers.

Design and Optimization of Solar Thermal Collector with Integrated Phase Change Material (PCM)

Authors: Mr. Uddesh Dhanraj Dongre, Prof. Mithlesh Pandey

Abstract: Solar thermal collectors are widely used for converting solar energy into useful thermal energy for domestic and industrial applications. Conventional collectors suffer from energy loss during cloudy weather and nighttime due to the absence of efficient thermal storage systems. To overcome this limitation, Phase Change Materials (PCM) are integrated into solar thermal collectors. PCM absorbs excess heat during sunshine hours and releases stored thermal energy during low solar radiation conditions. This research focuses on the design and optimization of a solar thermal collector integrated with PCM. Paraffin wax is selected as PCM because of its high latent heat capacity, thermal stability, chemical inertness, non-corrosive nature, and suitable melting temperature range. The performance of the collector is evaluated based on thermal storage capability, charging and discharging characteristics, outlet water temperature, heat retention, and efficiency improvement. The study shows that PCM integration significantly improves thermal efficiency and maintains outlet temperature for longer duration compared to conventional collectors. The optimized collector demonstrates enhanced energy utilization, reduced temperature fluctuation, and better thermal stability. The proposed system is suitable for domestic water heating, industrial thermal applications, agricultural drying systems, and renewable energy storage applications.

Design Method For Online Totally Self-Checking Comparators Implementable On FPGAs

Authors: Harishankar T, Dr.T.R.Ganesh Babu

Abstract: In the context of their growing use in critical fields of application, like aviation electronics, automotive control systems, and industrial automation, FPGA circuits' operation must be guaranteed against both soft errors and any other defects that may arise during run-time. This paper analyzes in depth an approach for implementing Totally Self-Checking (TSC) comparators for online diagnostics in FPGAs in a way which maximizes its effectiveness in terms of test pattern complexity and hardware overhead. In particular, the presented technique utilizes the circuitry features of Look-Up Tables (LUTs) to provide comprehensive online testing with a number of test vectors proportional to O(n), while guaranteeing complete fault coverage and regardless of the specific LUTs configuration. The results of a comparison among recent techniques for implementing TSC, both BIST-based and Dual Modular Redundancy (DMR), show that the described solution offers an outstandingly effective performance with regard to SER (0.055 FIT).

DOI: https://doi.org/10.5281/zenodo.20233433

AI Driven Intrusion Detection System Using Hybrid Deep Learning In Cloud Environment

Authors: Dr Vijayalakshmi V, Ms.Sneha R. V. Kumbhar

Abstract: However, the rise in cloud computing usage has resulted in increased complexity and vulnerability of organizations' IT infrastructure. In addition, cloud services have created new vulnerabilities that can easily be targeted by sophisticated attacks since traditional intrusion detection methods lack the ability to cope with the dynamically changing nature of cloud environments. This paper offers a novel, AI-powered hybrid deep learning framework for intrusion detection in cloud environments. The hybrid IDS is based on a combination of Triplet Attention-based Residual CNN for spatial feature extraction of network traffic, Bi-LSTM with attention mechanism for temporal dependency modeling, and Particle Swarm Optimization for hyperparameter optimization. Based on the evaluation results performed on the CSE-CIC-IDS2018 and UNSW-NB15 dataset, the suggested hybrid architecture attains an impressive accuracy of 99.12%, precision of 98.9%, and recall of 99.0%, outperforming the performance of individual CNN (96.4%) and Bi-LSTM (95.8%). In terms of efficiency, the PSO-based architecture has a latency less than 50 ms with minimal false positive rate of only 1.2%.

DOI: https://doi.org/10.5281/zenodo.20233688

Farmers’ Perceptions of Marketing Functions Rendered by Cooperative Marketing Societies: A Five-Factor Model for Understanding Multi-Dimensional Service Quality

Authors: Associate Professor Dr. S.Sureshbabu, Research Scholar Mr. A.kannan

Abstract: This study examines farmers' perceptions of marketing functions and services rendered by Cooperative Marketing Societies (CMS) through comprehensive exploratory factor analysis of data from 620 farm-members. Principal Components Analysis with Varimax rotation identifies five distinct dimensions of CMS marketing functions: Market Operations and Transaction Efficiency, Pricing and Bargaining Effectiveness, Market Access and Infrastructure Support, Post-Harvest and Quality Support Services, and Information and Financial Support Services. The findings reveal that farmers rate infrastructure support (mean = 4.30) and storage facilities (mean = 4.23) most favorably, while expressing moderate satisfaction with pricing transparency (mean = 2.69) and income impact (mean = 3.28). Cluster analysis segments farmers into three groups: 62.4% highly satisfied, 23.2% moderately satisfied, and 14.4% less satisfied with CMS functions. The five-factor model explains 64.183% of cumulative variance, establishing a robust framework for understanding CMS service quality and performance. The study provides evidence-based insights for strengthening cooperative marketing functions and designing targeted interventions to enhance farmer satisfaction across service dimensions.

DOI: https://doi.org/10.5281/zenodo.20242694

MindMeld: A Tiered Orchestration Framework For Automated Synthesis And Deployment Of Production-Grade Multi-Agent Systems From Natural Language Specifications

Authors: Prof. Chetan Kumar V, Hardik Jain, Pranathi B H, Vikas R P, Srinidhi Prabhu M U

Abstract: The deployment of production-grade Multi-Agent Systems (MAS) from natural language specifications remains a significant challenge in software engineering, requiring so- phisticated role decomposition, reliable tool integration, exe- cutable code synthesis, and robust packaging with dependency management. This paper presents MindMeld, a novel tiered LLM orchestration framework that transforms natural language requirements into deployable, containerized multi-agent systems through a three-phase pipeline architecture. MindMeld introduces several key innovations: (1) a formal planning phase that generates machine-verifiable JSON agent specifications with explicit dependency graphs and interface con- tracts; (2) a closed-loop validation tier combining static analysis, dynamic runtime testing in isolated sandboxes, and iterative self-refinement based on structured error feedback; and (3) an automated integration phase that synthesizes orchestration logic, manages inter-agent communication, and produces containerized artifacts with complete dependency resolution. We conduct comprehensive evaluation on 47 diverse natural- language build requests spanning 8 task categories (data pro- cessing, API integration, document analysis, notification systems, workflow automation, content generation, monitoring, and multi- modal processing). Our results demonstrate that MindMeld achieves 78.7% end-to-end build success compared to 34.0% for single-pass generation baselines, with an average of 1.8 validation iterations per sub-agent. Ablation studies reveal that the planning phase contributes 23.4% improvement and the val- idation loop adds 21.3% improvement to overall success rates. A controlled user study with 24 participants shows 3.2× reduction in deployment time and 4.1/5.0 satisfaction scores. These results establish MindMeld as a practical framework for bridging the gap between natural language intent and production-ready multi- agent systems.

Integrated Groundwater Quality Assessment And Machine Learning Prediction In Central Uttar Pradesh, India

Authors: Nitin Mishra

Abstract: Groundwater is the principal source of drinking and irrigation water in the Indo-Gangetic alluvial plains of Uttar Pradesh, India. Rapid urbanization, agricultural intensification, excessive groundwater abstraction, and geogenic contamination have significantly affected groundwater quality in the region. The present study evaluates groundwater quality in Central Uttar Pradesh using hydrogeochemical assessment, entropy-weighted water quality index (EWQI), and machine learning (ML) prediction techniques. A total of 178 groundwater samples were analyzed for major physicochemical parameters including pH, EC, TDS, TH, Ca2+, Mg2+, Na+, K+, HCO3−, Cl−, SO42−, NO3−, F−, SiO2, and CO32−. The entropy weight method was employed to minimize subjectivity in water quality assessment, while hydrogeochemical interpretations were carried out using Piper and Gibbs diagrams. Three machine learning models, namely Classification and Regression Tree (CART), Random Forest (RF), and Extreme Gradient Boosting (XGBoost), were implemented to predict groundwater quality conditions. The results revealed that groundwater chemistry is predominantly controlled by rock–water interaction and ion exchange processes, with Ca–HCO3 and mixed hydrochemical facies dominating the study area. The EWQI values indicated that most groundwater samples fall within good to medium drinking water quality categories, although localized fluoride enrichment was observed in several locations. Among the applied models, XGBoost demonstrated superior predictive capability with R2 = 0.9597, RMSE = 2.2376, and MAE = 1.7690, outperforming RF and CART models. The findings highlight the effectiveness of integrating GIS-based hydrogeochemical analysis with machine learning approaches for groundwater quality prediction and sustainable groundwater management in Central Uttar Pradesh.

DOI: http://doi.org/10.5281/zenodo.20268302

Solar Powered Grass Cutter With Mobile Remote Control For Medium Outdoor Space

Authors: Kamble Abhishek B, Nimbalkar Saurabh S, Patil Abhayrajsinh M.

Abstract: The increasing demand for eco-friendly and automated landscaping solutions has led to the development of a Solar-Powered Grass Cutter with Mobile Remote Control for medium outdoor spaces such as gardens, parks, and institutional campuses. The system harnesses solar energy via photovoltaic panels, which charge a 12V, 7Ah lead-acid battery through an MPPT charge controller. An ESP32 microcontroller receives Bluetooth commands from a mobile application to drive four Johnson DC gear motors (12V, 10 RPM) for mobility and a PMDC motor (12V, 5000 RPM) for blade actuation, interfaced via an L298N motor driver. Field testing demonstrated a cutting efficiency of 98.5% for grass heights of 15–25 mm and 78% for overgrown grass exceeding 60 mm, with operational runtime of 1.5–2 hours per full charge. The total fabrication cost is approximately Rs. 10,955, making it a cost-effective and environmentally sustainable alternative to conventional fuel-powered grass cutters.

Heavy Metal Accumulation In River Sediments: A Case Study Of The Gomti River

Authors: Nisha Gautam

Abstract: Heavy metal contamination in river sediments is a major environmental concern due to rapid urbanization and industrialization. The present study assessed heavy metal contamination in sediment samples collected from selected sites of the Gomti River in Lucknow, Uttar Pradesh, India. Sediment samples were collected from Gaughat, Kudiya Ghat, Daliganj Bridge, and Hanuman Setu during November 2025 and analysed for chromium (Cr), nickel (Ni), arsenic (As), cadmium (Cd), and iron (Fe). Pollution assessment indices including Enrichment Factor (EF), Contamination Factor (CF), and Pollution Load Index (PLI) were used to evaluate contamination levels and anthropogenic influence. The results revealed significant spatial variation in heavy metal concentrations. Iron showed the highest concentration, while cadmium exhibited extremely high concentrations compared to its background value, indicating severe contamination. The concentration pattern followed the order: Fe > Cr > Ni > Cd > As. EF analysis indicated extremely severe enrichment of Cd, whereas Cr showed moderate enrichment. CF results also confirmed very high contamination by Cd. PLI values at all sampling sites were greater than 1, indicating polluted sediment conditions. The study concludes that anthropogenic activities such as sewage discharge, urban runoff, and industrial effluents are major contributors to heavy metal accumulation in the Gomti River sediments.

DOI: http://doi.org/10.5281/zenodo.20268759

GIS-Based Mapping Of Groundwater Contamination In Lucknow District, Uttar Pradesh

Authors: Zaira Siddiqui

Abstract: Groundwater is an essential source of drinking water in urban regions; however, rapid urbanization, industrial growth, and anthropogenic activities have significantly deteriorated groundwater quality in many Indian cities, including Lucknow. The present study aims to evaluate the spatial variability of groundwater quality in Lucknow district using Geographic Information System (GIS)-based techniques and Water Quality Index (WQI) approaches. Major physicochemical parameters including pH, electrical conductivity (EC), total hardness, calcium (Ca²⁺), magnesium (Mg²⁺), chloride (Cl⁻), fluoride (F⁻), nitrate (NO₃⁻), and sulphate (SO₄²⁻) were analyzed for groundwater quality assessment. Spatial interpolation of groundwater parameters was performed using the Inverse Distance Weighting (IDW) method in GIS to generate thematic distribution maps and identify contamination hotspots. Two groundwater quality assessment approaches, namely Arithmetic Water Quality Index (AWQI) and Weighted Water Quality Index (WWQI), were applied to evaluate overall groundwater suitability for drinking purposes. The results revealed significant spatial variability in groundwater quality across Lucknow district. Elevated concentrations of hardness, EC, nitrate, chloride, and sulphate were observed in several urbanized and densely populated regions, indicating strong anthropogenic influence on groundwater systems. The AWQI and WWQI hotspot maps indicated that eastern and southeastern parts of Lucknow district exhibited comparatively poor groundwater quality, while northern and western regions showed relatively better water quality conditions. Comparative analysis demonstrated that WWQI provided a more realistic and reliable assessment because parameter-specific weighting improved sensitivity toward critical contaminants. GIS-based hotspot mapping successfully delineated vulnerable groundwater zones and highlighted areas requiring immediate monitoring and management intervention. The study demonstrates that integration of GIS and WQI techniques is highly effective for groundwater quality assessment, contamination hotspot identification, and sustainable groundwater resource management. The findings of this study can support policymakers and environmental planners in developing targeted groundwater protection and remediation strategies for rapidly urbanizing regions.

DOI: http://doi.org/10.5281/zenodo.20268893

Analysis Of Leachate From The Municipal Solid Waste Disposal Site And Its Impact On Groundwater Quality At Lucknow

Authors: Shivanshi Verma

Abstract: This study evaluates leachate quality from a municipal solid waste disposal site in Lucknow and examines its impact on nearby groundwater. The analytical framework, sampling design and index-based interpretation were prepared in line with the uploaded thesis and sample journal paper, while the numerical results were derived from the uploaded laboratory workbook. One leachate sample and seven groundwater samples were assessed for physicochemical and heavy metal parameters using APHA-based methods. The leachate showed acidic to near-neutral reaction (pH 6.1), very high electrical conductivity (83,892 µS/cm), total dissolved solids (38,180 mg/L), chemical oxygen demand (16,800 mg/L), biochemical oxygen demand (2,000 mg/L), hardness (1,620 mg/L), chloride (980 mg/L), sulphate (678.5 mg/L), nitrate (103.44 mg/L), fluoride (8.8 mg/L) and substantial heavy metal burden, indicating strong contaminant potential. The Leachate Pollution Index was 25, confirming significant pollution load. Groundwater quality varied spatially: Sample-7 recorded a WQI of 75.13 and fell in the good category, whereas Samples 2–4 were poor and Samples 1, 5 and 6 were very poor. Elevated TDS, alkalinity, hardness, iron, manganese, nickel, copper and zinc were the major causes of groundwater deterioration. The data indicate that leachate migration has affected groundwater quality in the vicinity of the disposal site, although the effect is not controlled by distance alone. The study recommends leachate containment, regular groundwater surveillance, and priority treatment for metal and salinity-related contamination.

DOI: https://doi.org/10.5281/zenodo.20269275

A Systematic Review on Hybrid Transformer Framework for Temporal Representation Learning and Longitudinal Risk Prediction In Clinical Time-Series

Authors: Abdullahi Idris, Aminu A. Abdullahi, Jamilu Awwalu, Abdullahi Uwaisu Muhammad

Abstract: The increasing availability of Electronic Health Records (EHRs), ICU monitoring systems and clinical sensor technologies has generated large volumes of temporal healthcare data that require advanced analytical approaches for effective interpretation and prediction. Traditional machine learning and statistical models often face challenges in handling complex temporal dependencies, irregular sampling, missing values and censored survival outcomes in clinical time-series data. This study employed a Hybrid Transformer Framework for Temporal Representation and Longitudinal Risk Prediction in Clinical Time Series synthesizing the relevant studies and clinical decision-making. The framework integrates the Transformer-LSTM architecture with Cox Proportional Hazards (Cox PH), Survival Random Forest (SRF) and XGBoost algorithms. The Transformer component captures long-range temporal dependencies using self-attention mechanisms, while the LSTM network models short-term sequential clinical patterns. Cox PH is applied for interpretable survival analysis, SRF for nonlinear ensemble survival prediction and XGBoost for high-performance risk classification and prediction. The review study utilizes healthcare datasets such as MIMIC-III, MIMIC-IV, elCU and PhysioNet as well as providing suitable comparative approaches against baseline models.

DOI: https://doi.org/10.5281/zenodo.20269570

Scalable Database Systems for Big Data Analytics: Challenges and Solutions

Authors: Shah Md. Tanzimul Kabir, Zahid Hassan Ome

Abstract: This paper provides a comprehensive analysis of scalable database systems, specifically designed to support big data analytics, and examines their evolution, challenges, and emerging technologies in the exascale data processing era. By examining recent research studies from 2021 to 2026, the current paper seeks to investigate how distributed database architectures, including NewSQL, cloud-native, and data lakehouse, address the fundamental scalability challenge known as the "scalability trilemma" consisting of consistency, availability, and partition tolerance. The current research introduces the Adaptive Scalability Evaluation Framework (ASEF), which integrates horizontal scaling, elastic resources, query optimization, and storage efficiency. The analysis shows that recent scalable database architectures are based on disaggregated storage and compute architectures, enabling near-linear scaling to thousands of nodes with query latencies under 100ms for petabyte-scale data sets. Cloud-native database architectures are shown to be highly elastic, with variations in query latency at the 95th percentile below 15% during scaling events. Newly emerging architectures for lakehouses, which bring the flexibility of data lakes and the performance of data warehouses, provide query performance that is 3 to 5 times better than traditional data lakes and reduce the total cost of ownership by 30 to 50 percent. Evaluation in five dimensions for analytical workloads, such as scaling behavior, consistency model, query performance, storage efficiency, and operational complexity, shows that systems with workload awareness and adaptivity perform much better than static configurations. Continuous optimization provides an improvement in throughput performance that is between 2 to 4 times.

DOI: https://doi.org/10.5281/zenodo.20270564

A Study On The Relationship Between Leadership Styles And Team Performance In Startups

Authors: Anshu Kumar Mishra, Sohail Verma

Abstract: This paper investigates the relationship between leadership styles and team performance in startup organisations, using survey-based data collected from 120 respondents comprising founders, co-founders, team leads and early-stage employees across multiple sectors. The study identifies transformational leadership as the dominant style in the sample and finds strong positive associations between vision-driven leadership, team trust, communication frequency and performance outcomes. Transactional leadership shows moderate relevance in goal-setting and accountability, while laissez-faire approaches correlate with lower performance consistency. Exploratory chi-square testing reveals significant concentration in leadership style distribution, a meaningful link between startup stage and performance rating, and a strong association between trust levels and team output. The paper concludes that startup performance is not driven by a single leadership template but by the leader's ability to adapt style to team maturity, organisational stage and the demands of rapid growth. A hybrid leadership model combining transformational inspiration with transactional clarity emerges as the most effective pattern for high-performing startup teams.

Visualization And Analysis Of Pro Kabaddi League Data Across All Seasons Using Tableau

Authors: Myana Ramesh, Kanchapogu Prasanth, Mr. T. Srinivas

Abstract: Every PKL match across multiple seasons outcomes, dates, venues, scores, teams in one place. That's what this dataset is. What you can actually do with it is more interesting than the description suggests. Win/loss trends show which teams hold up across a full season and which ones are inconsistent. Scoring patterns reveal whether a team plays the same way regardless of opponent or adjusts. Venue data is underrated — some teams genuinely perform differently away from home, and the numbers show it. Zoom out across seasons and the league's own growth becomes visible too. More cities, more matches, more structure. PKL didn't stay the same sport it was in its first season, and this data captures that shift better than any summary could.

Machine Learning Applications In Network Security

Authors: Mazlan Othman

Abstract: Machine learning (ML) has emerged as a powerful approach for enhancing network security by enabling intelligent detection, prevention, and response to cyber threats. With the increasing complexity and scale of modern networks, traditional rule-based security systems are often insufficient to identify sophisticated attacks such as zero-day exploits, phishing, and advanced persistent threats (APTs). This paper explores the application of machine learning techniques in network security, focusing on how supervised, unsupervised, and reinforcement learning models can analyze network traffic patterns to detect anomalies and malicious activities. It also examines the role of ML in intrusion detection systems (IDS), intrusion prevention systems (IPS), malware detection, and behavioral analysis. Cloud-based and real-time security monitoring systems are discussed as key enablers for scalable ML deployment in distributed network environments. Additionally, the study highlights challenges such as adversarial attacks, data imbalance, privacy concerns, and model interpretability. Emerging solutions including federated learning, explainable AI, and edge-based security analytics are also reviewed. The findings emphasize that machine learning significantly strengthens network security frameworks by enabling proactive, adaptive, and intelligent threat detection mechanisms.

DOI: https://doi.org/10.5281/zenodo.20281072

PLC And SCADA Design Of Dairy Processes

Authors: Professor Mayur Patil, Sayyad Ayaj Riyaj, Sayyad Sameer Shahabuddin, Wadgaonkar Hrushikesh Kanifnath

Abstract: Dairy processing, including pasteurization, storage, and packaging, demands precise control to ensure product safety, quality, and efficiency. This report presents the design of an automated dairy processing system using PLC (Programmable Logic Controllers) and SCADA (Supervisory Control and Data Acquisition) technology. The proposed system integrates sensors (temperature, level, flow, and pH) and actuators (valves, pumps, motors) with PLCs to execute control logic, and a SCADA HMI for real-time monitoring, data logging, and operator interaction. Automation is essential in large- scale dairy plants to reduce manpower, prevent contamination, and optimize processes. The system aims to automate milk pasteurization, Clean-In-Place (CIP) cleaning cycles, and packaging lines, resulting in consistent product quality, improved throughput, and traceability. Technical specifications, software details, and implementation methodology are discussed, and advantages and limitations of the PLC/ SCADA solution are highlighted.

Towards Fine-Grained Depressive Symptom Recognition In Memes Via Multimodal Transformer-CNN Fusion

Authors: Mrs. J. Annie Jennifer, Dr. R. Gunasundari

Abstract: The mental health indicators can be found in memes, and it is quite complex since memes consist of both text and images, and one must analyze both elements to understand their meaning. This research proposes a novel deep learning technique named Multi-CNN. Its aim is to detect depression-related signs by analyzing their linguistic and visual content simultaneously in memes . The technology uses both the BERTweet model for natural language processing and ResNet18 features for images from a neural network. It was assessed using a dataset of internet memes annotated according to eight depression indicators. Early stopping, data augmentation, and others helped improve its performance, while results were estimated by means of a weighted F1 score. As the study shows, it is more effective to use linguistic and visual components simultaneously than to employ the model based only on language or solely on image analysis for identifying the presence of depressive signs in memes. The multimodal approach resulted in a weighted F1 score of 0.6846, while the language-based model received 0.6716. Using just the picture is ineffective when it comes to recognizing depression-related memes. The study's findings indicate that visual information and text together create strong cues for investigating mental health issues. Besides, the results point to fresh techniques and technologies that can handle the intricate heterogeneous datasets found in social media.

A Hybrid Deep Learning Framework For Real-Time Yield Prediction And Process Monitoring In Biomanufacturing

Authors: Hadiza Ibrahim Aminu, Abdullahi Mohammed Ibrahim, Buhari Aliyu, Zainab Ibrahim Aminu, Abubakar Safiyanu

Abstract: Bioprocessing plays an essential role in the large-scale production of biological products, where accurate monitoring and control are key for both yield and quality. This work aims to develop and assess a predictive framework based on Artificial Neural Networks (ANN) for estimating product yield in bioprocess operations. A multi-phase approach was implemented, beginning with data collection from online sensors and laboratory analyses, followed by preprocessing steps that included normalization, outlier removal, noise filtering, and feature engineering, utilizing dimensionality reduction through Principal Component Analysis. A hybrid ANN model was created, integrating Feed-Forward Neural Networks (FNN) for steady-state predictions, Long Short-Term Memory (LSTM) networks for learning temporal sequences, and Convolutional Neural Networks (CNN) for interpreting spectroscopic data.The model, trained using supervised learning and cross-validation, achieved strong predictive performance with a Mean Squared Error (MSE) of 1.0139 and a coefficient of determination (R²) of 0.9756, capturing 97.6% of yield variance. Predicted versus actual values showed high consistency, confirming robustness for real-time monitoring. Minor overfitting was observed at extreme values, highlighting the need for dataset expansion and regularization. Overall, the results demonstrate that ANN-based modeling effectively captures nonlinear dynamics in bioprocessing, supporting proactive optimization, disturbance detection, and integration into industrial-scale monitoring systems.

DOI: http://doi.org/10.5281/zenodo.20282959

Automated Classification Of Reptiles And Amphibians Using MobileNetV2 And Transfer Learning

Authors: Bandaru Jyothi, M.Radhika

Abstract: This article presents a new approach to automated amphibian and reptile categorization that makes use of deep Convolutional neural networks (CNNs) and transfer learning. By developing a reliable and precise MobileNetV2 model for species identification using deep learning, we tackle the limitations of traditional classification methods while also acknowledging the ecological importance of these two vertebrate groups. Using a transfer learning approach on a massive collection of amphibian and reptile images, we train a pre-trained Convolutional neural network (CNN) to overcome the issue of small dataset size. The model is able to generalize well across several species due to its high extraction efficiency. Additionally, the article delves into the significance of image augmentation techniques for enhancing model performance, particularly in cases when labeled data is scarce. Results are favorable when the proposed method is used to overcome challenges caused by changes in size, posture, and environmental factors. Ecological monitoring, conservation efforts, and biodiversity surveys might benefit from the model's classification accuracy, which we prove by comparing it to a large dataset of amphibians and reptiles. With an accuracy rate of 82%, the proposed MobileNetV2 model cans correctly categories amphibians and reptiles. The growing field of computer vision as it pertains to animal ecology and biology has a scalable and successful approach to automated species identification, which this work adds to it. The results show that deep learning techniques particularly transfer learning, have the potential to address the issues with animal categorization. Additional investigation on the connection between AI and biodiversity protection might result from this.

DOI: http://doi.org/10.5281/zenodo.20285882

ContractSphere AI: A Smart Contract Management System Using Artificial Intelligence And Blockchain

Authors: Harshada Magar, Yash Bhalekar, Sarthak Belvalkar, Om Jadhav

Abstract: This paper presents ContractSphere AI, a system designed to help organizations manage their contracts more easily and securely. Managing contracts in companies involves many steps such as writing, reviewing, checking legal rules, and storing the final signed document. Doing all these steps manually takes a lot of time and often leads to mistakes. ContractSphere AI uses artificial intelligence to automate these steps and uses blockchain technology to make sure that signed contracts cannot be changed or faked. The system can understand contract language in multiple languages and can handle contracts from different countries with different legal rules. It uses a language model trained on legal documents together with a search system that finds relevant rules and contract examples. The final signed contract is stored securely by saving its unique hash on the blockchain, which proves the contract is genuine. This paper describes how the system works, explains the main processing steps, and discusses how well the system performs in terms of speed, security, and cost.

DOI: https://doi.org/10.5281/zenodo.20286629

Deep Learning And Image Processing-Based Bank Check Verification System

Authors: Marella Maheswari, P ASHOKA

Abstract: Revolutionizing the verification of bank checks, this innovative technology simplifies the process by integrating deep learning, image processing, and an intuitive Django-based web interface. It streamlines the process with little human participation, making it easier than ever before. Our Convolutional neural network (CNN) trained on the IDRBT check dataset and executed in PyTorch has a 99.14% success rate in recognizing handwritten digits, as shown in the introductory article. Adaptive thresholding and Gaussian blurring are implemented in the source code to enhance the picture preparation. The optical character recognition (OCR) in MATLAB can recover machine-printed text with 97.7 percent accuracy, including IFSC codes and account numbers, when Pytesseract is used in the code for region-based text extraction. The approach uses SVM classification and SIFT feature extraction for real-time authenticity checks, allowing signature verification powered by SIFT and SVM to reach 98.1% accuracy. The web-based interface allows more users to upload photos of checks, train models, see datasets, and get immediate categorization results ("Genuine" or "Not Genuine"). The system complies with CTS-2010 standards for Indian banks and the extraction of critical details such as signatures, amounts, and check numbers is possible even if it supports formats from other countries. In order to automate the verification process and decrease processing time, operational expenditures, and fraud risks, it makes use of contour detection and region-based analysis. This scalable solution sets a new standard for secure, efficient financial transactions by combining the rigors approach from the paper with the actual code implementation. Future versions may support more than one language and format.

DOI: http://doi.org/10.5281/zenodo.20286653

Intelligent Prediction Of Smartphone Addiction Through Machine Learning Algorithms

Authors: Singareddy Saritha, M. Sivaparavathi

Abstract: A rising number of individuals are displaying signs such as excessive phone usage, loss of productivity, and even physical and psychological health concerns, making Smartphone addiction a major worry in recent years. The development of reliable instruments for the prediction of Smartphone addiction and the identification of those at risk is, hence, necessary. Using survey data on Smartphone use, we constructed a machine learning model to forecast Smartphone addiction in this research. There was a wide variety of mental health concerns addressed in the survey, including demographics, phone use patterns, and anxiety, despairs, and stress. The model was constructed using a well-liked and efficient machine learning technique. In this work, numerical variables are normalized and categorical variables are encoded as part of the data preprocessing to make sure the model can train properly. Also, we used measures like accuracy to measure the model's performance on the remaining data after training it on a subset of the data. The algorithm has successfully predicted Smartphone addiction with a high degree of accuracy, according to the findings. Use habits of mobile phones, including how often notifications were checked, how many hours spent on the phone daily, and the applications used most often, were the most critical variables for predicting addiction. Age, gender, and stress levels were other important factors. The constructed model has a number of possible uses. Healthcare providers might use it to identify patients at risk of Smartphone addiction and intervene accordingly. Also, app makers may utilize it to make their applications less addicting and more conducive to healthy phone habits. In a nutshell, the results show that machine learning algorithms can effectively predict Smartphone addiction. We need to conduct further studies to confirm our results on bigger and more varied datasets and to investigate other possible uses for this approach.

DOI: http://doi.org/10.5281/zenodo.20286754

Eco-Rupees: Plastic to Pride

Authors: Eeshritha

Abstract: India generates millions of tonnes of plastic waste annually, much of which is non-biodegradable. This paper explores the feasibility of using recycled polypropylene (PP) to produce polymer currency notes. Drawing on Australia’s pioneering adoption of polymer banknotes, the study eval-uates technical, economic, policy, and social challenges, and proposes a phased roadmap for India to transition towards sustainable currency pro-duction. The findings suggest that recycled PP notes could simultaneously address waste management, enhance currency durability, and position India as a global leader in sustainable finance innovation.

Analyzing Amazon Sales Dataset with Tableau: A Visualization Approach

Authors: Malgaram Punith Teja, Srishti Singh, Baddula Sreeya Yadav, Mrs.Sumayya Samreen

Abstract: This research paper examines what makes an e-commerce business succeed. They use a data set on fifty thousand Amazon transactions to see the effect of all sorts of different variables on sales. What factors we considered was the sale price, discount percentage, ratings and amount of reviews. We examined the mechanics of pricing ,examined the level of consumer confidence in a product , we studied preferences we considered payment mode It shows how offs, and customer ratings influence earned money of the businessThis results from suggesting discounts might increase sales over a specific term but influence long term revenues via performing well on customer ratings and consumer trust. We find that both markets of North America and Middle-East have the revenue. The findings of this research may be accessed by merchants and marketers who wish to set their prices to helps consumers in their purchase decisions and at the same time, increase their revenues. Based on these results, they are able decide about their business.

DOI: https://doi.org/10.5281/zenodo.20302852

Data Visualization Of Nfl Offensive Player Stats,1999-2013 Dataset

Authors: Gajabheenkar Roshini, Dr.Lavanya Pamulaparty

Abstract: This research focuses on analyzing NFL offensive player statistics from 1999–2013 using Tableau visualization techniques. The dataset contains player demographics combine performance, draft information, and offensive statistics. Various visualizations such as bar charts, heat maps, dashboards, and highlight tables are used to identify player performance trends and statistical patterns. The analysis helps understand how player attributes and performance metrics contribute to offensive success in the NFL.

Ai Powered Ats Resume Screeing And Job Recommendation System

Authors: Ragula Rajesh, Potti Rakesh, Poojari Jayakrishna, Mrs.V. Elavenil

Abstract: This study describes the deployment of a cloud-native AI system designed to assist HR departments by offering automated resume screening and candidate-job matching. To provide precise and contextually aware responses, the system employs a Retrieval-Augmented Generation (RAG) technique, which combines a language model with a local knowledge store of job descriptions. We developed this totally with free and open-source technologies like AWS Lambda, BERT, and ChromaDB to make the solution more accessible for startups and SMEs. Open-source approaches, such as Tesseract for OCR, are used to add scanned resume capabilities. To ensure accuracy and validity, the data is also gathered from reputable sources like ESCO skill ontology and LinkedIn datasets. We used PDFs from these sources for RAG and stored them in vector databases for efficient document retrieval, as this system aims to bridge the information gap in high-volume hiring.

DOI: https://doi.org/10.5281/zenodo.20304132

Bangladeshs Journey From Economic Basket Case To Middle-Income Complexities.

Authors: Dr. Mohammad Shah Alam Chowdhury

Abstract: Since its independence in 1971, Bangladesh has transformed from one of the poorest nations in the world famously mischaracterized as a "basket case" to one of the fastest-growing economies in South Asia. This paper examines the trajectory of Bangladesh's economic growth, driven primarily by a booming ready-made garment (RMG) industry, robust remittance inflows, and significant advancements in social development indicators such as female labor force participation. However, despite reaching lower-middle-income status in 2015, the nation currently faces severe macroeconomic headwinds. Elevated inflation, banking sector vulnerabilities, low tax revenue mobilization, and external shocks have slowed recent GDP growth. This paper analyzes the historical drivers of growth, structural bottlenecks, and the urgent policy reforms required to ensure sustainable and inclusive economic development.

DOI: https://doi.org/10.5281/zenodo.20305441

India’s Corporate Titans: Strategic, Financial, and Integrative Dimensions of the Top 10 Mergers and Acquisitions

Authors: Bikku Kumar

Abstract: Mergers and acquisitions (M&A) have emerged as the most consequential instruments of corporate strategy in India's post-liberalisation growth narrative. This paper undertakes a rigorous multi-dimensional examination of India's ten largest M&A transactions — spanning banking, e-commerce, steel, telecommunications, cement, aluminium, pharmaceuticals, and automotive sectors — executed between 2007 and 2023 with a cumulative deal value exceeding USD 113 billion. Employing a descriptive-analytical framework grounded in secondary financial data, the study evaluates pre- and post-merger performance across key metrics including Return on Equity (ROE), Earnings Per Share (EPS), and Debt-to-Equity (D/E) ratio. A comparative matrix is deployed to assess strategic intent realisation, post-merger integration efficacy, and sector-specific determinants of M&A success or failure. Statistical analysis of financial ratios reveals a statistically significant divergence between successful and failed deals when measured against pre-merger benchmarks, with successful integrations yielding a mean ROE improvement of 2.1 percentage points. The findings unequivocally establish that strategic alignment, due diligence rigour, and cultural integration capacity are the decisive success factors — not deal size alone. The research contributes an empirically grounded, sector-comparative understanding of M&A dynamics in an emerging market context.

DOI: https://doi.org/10.5281/zenodo.20305623

Social Realism And Class Conflict In The Works Of Aravind Adiga And Rohinton Mistry.

Authors: Dr. Mohammad Shah Alam Chowdhury

Abstract: The present study explores the representation of socialism and social realism in the selected works of Aravind Adiga and Rohinton Mistry. Socialism, as a socio-economic ideology, advocates equality, collective welfare, and the reduction of class divisions; however, its practical implementation in India has often revealed contradictions marked by poverty, corruption, exploitation, and social inequality. Contemporary Indian English literature reflects these realities by portraying the struggles of marginalized communities and critiquing the failures of political and economic systems. The study examines how Adiga and Mistry depict the harsh realities of Indian society through themes such as class oppression, poverty, labor exploitation, corruption, identity crisis, and social injustice. Aravind Adiga’s novels, particularly The White Tiger and Last Man in Tower, expose the inequalities of post-liberalization India where capitalist ambitions overshadow socialist ideals. His narratives present a satirical and confrontational critique of economic disparity, moral decay, and institutional corruption. In contrast, Rohinton Mistry’s works, including A Fine Balance and Such a Long Journey, offer a compassionate and humanistic portrayal of ordinary individuals struggling against political oppression, social discrimination, and economic hardships. The research further compares the narrative techniques, ideological perspectives, and social concerns reflected in the writings of both authors. While Adiga emphasizes rebellion, survival, and individual ambition within a corrupt socio-economic structure, Mistry focuses on resilience, human dignity, empathy, and collective suffering. Despite their differing approaches, both writers critically examine the failures of governance and the widening gap between the privileged and marginalized sections of society. The study concludes that the works of Adiga and Mistry serve as powerful critiques of socio-political realities in contemporary India. Their fiction not only reflects the complexities of socialism and capitalism but also highlights the role of literature as a medium for social awareness, resistance, and the representation of marginalized voices.

DOI: https://doi.org/10.5281/zenodo.20305758

Telecom Network Intelligence System Using AI

Authors: Shubham sahu, Dr. Dharmbir Yadav

Abstract: The rapid growth of modern telecommunication technologies such as 2G, 4G LTE, and 5G has significantly increased the complexity of telecom network management. Telecom operators continuously generate massive amounts of network data related to traffic usage, bandwidth utilization, latency, throughput, and user activity. Traditional telecom monitoring systems mainly rely on manual analysis and threshold-based alert mechanisms, which are often unable to predict future network congestion, performance degradation, or operational failures effectively. To overcome these limitations, intelligent and automated monitoring solutions are required for efficient telecom network management. This research work, titled “Telecom Network Intelligence System using AI,” proposes an Artificial Intelligence (AI) and Machine Learning (ML) based framework for intelligent telecom network monitoring, traffic prediction, congestion detection, and performance analysis. The proposed system integrates telecom KPI analytics, predictive Machine Learning models, and real-time dashboard visualization to support proactive telecom operations and data-driven decision-making. The system analyzes important telecom Key Performance Indicators (KPIs) such as Call Setup Success Rate (CSSR), Call Drop Rate, LTE Throughput, PRB Utilization, Network Latency, and User Throughput collected from 2G, 4G LTE, and 5G networks. Machine Learning algorithms including Linear Regression, Random Forest, and Long Short-Term Memory (LSTM) are utilized for traffic forecasting, anomaly detection, and congestion prediction. The proposed framework also integrates Python, SQL, Advanced Excel, and Microsoft Power BI for telecom data preprocessing, predictive analytics, and interactive dashboard development. The AI-driven dashboards provide real-time KPI monitoring, network health visualization, congestion alerts, and technology-wise performance comparison. Experimental analysis demonstrates that the proposed system improves prediction accuracy, reduces operational complexity, supports proactive fault management, and enhances telecom network efficiency. The research contributes toward the development of intelligent telecom monitoring systems capable of supporting future AI-driven telecom operations, AIOps integration, self-healing networks, and next-generation 5G/6G communication infrastructures.

A Study On Consumer Buying Behaviour Towards Flipkart Mobile Application

Authors: Ms. K. Divarshini, Dr. P. Poornima

Abstract: The rapid growth of e-commerce and smartphone usage has significantly transformed consumer buying behaviour. Online shopping platforms provide convenience, variety, and easy access to products, influencing consumer decisions. This study aims to analyse the buying behaviour of consumers towards the Flipkart mobile application. Primary data was collected through a structured questionnaire from 100 respondents. The findings reveal that factors such as discounts, product variety, user-friendly interface, and delivery services play a major role in influencing consumer decisions. The study concludes that Flipkart is widely preferred due to its convenience and attractive offers, but improvements in delivery speed and product quality can further enhance customer satisfaction.

DOI: http://doi.org/10.5281/zenodo.20309209

IoT and Machine Learning-Based Framework for Real-Time Methane Gas Detection and Bovine Health Monitoring in Dairy Farms

Authors: Dr. Deepika, Abhinav K G, Adithya Verma M A, Chiranth S Shetty, G Suhas Kartik

Abstract: Dairy farms generate substantial quantities of methane gas through enteric fermentation and manure decomposition. Elevated methane concentrations in enclosed or poorly ventilated cowsheds adversely affect cattle health, reduce milk productivity, and pose safety hazards to farm workers. Conventional gas-monitoring systems are reactive and threshold-based, generating alerts only after dangerous concentrations have already been reached. This paper presents an IoT and Machine Learning (ML)-based framework for real-time methane detection and bovine health risk classification. MQ-4 (methane), MQ-135 (air quality/ammonia), and DHT22 (temperature and humidity) sensors interface with an ESP32 microcontroller to collect continuous environmental readings that are transmitted to Firebase cloud storage via Wi-Fi using MQTT/HTTP protocols. Five supervised ML classifiers — Random Forest, Decision Tree, Support Vector Machine (SVM), Logistic Regression, and K-Nearest Neighbors (KNN) — are trained and evaluated for three-class bovine health risk classification (Low, Moderate, High). Random Forest achieved the highest performance with 96.8% accuracy, 96.5% precision, 96.8% recall, and an F1-score of 96.6% at the 90-10 train-test split, outperforming SVM (91.3%), Decision Tree (84.1%), KNN (79.6%), and Logistic Regression (76.9%). Automated alerts are delivered to farmers via a real-time Arduino IoT Cloud dashboard, email, and mobile push notifications. The proposed system is scalable and cost-effective.

DOI: https://doi.org/10.5281/zenodo.20321084

Design And Development Of Iot Based Agribot Solar Tracker

Authors: Ullas M K, Vishal Kumar B N, Yashwanth A S, Dr. S V. Anil Kumar, Dr. S V. Anil Kumar

Abstract: The Agri Bot Solar Tracker is a smart agricultural robot designed to support modern farming through automation and renewable energy. It performs agricultural tasks such as seed sowing and field monitoring using sensors to measure soil moisture, humidity, and temperature. The system uses Wi-Fi communication for remote monitoring and control through a monitoring and control station. Powered by solar energy with an integrated solar tracking system, the robot maximizes energy efficiency by adjusting the solar panel according to sunlight direction. This sustainable system reduces manual labor, improves operational efficiency, and promotes eco-friendly farming practices.

Sociosphere: A Social Network Platform For Empowering Real-World Social Change

Authors: Purushotam Naidu k, R.Srilatha, S.Gayathri, U.N. Harshitha, P. Siri Chandana

Abstract: Urban population growth creates new challenges related to civic infrastructure, and there is a need for efficient and smart complaint management systems. This paper describes the SocioSphere, which is an AI-based civic issue management platform that uses Natural Language Processing (NLP), machine learning, and high-performance web technologies to automatically process and route complaints. A report verification module (Fake/Real) built with Logistic Regression and engineered textual features can filter out spam and low-quality complaints. Valid complaints use a transformer model (RoBERTa) to identify the multi-class categories to which the complaint belongs. Furthermore, we have added a method of estimating the urgency of a complaint through the use of VADER-based sentiment analysis and heuristics for engagement, thus allowing for priority-based decisions. FastAPI is used to develop the backend API layer, offering high-speed (asynchronous/low latency) performance for model inferences and data processing. Complaints will be stored in the system's database and dynamically routed to appropriate authorities for final resolution. The experimental results demonstrate that the approach is effective for both classification and validation, as well as improving transparency and reducing manual work through the use of data-driven governance within smart city systems.

DOI: https://doi.org/10.5281/zenodo.20321890

Development of Nonconventional First-Class Fly Ash Bricks Using Silica Fume and Alkali Activators

Authors: Sanju, Rahul Kumar Jha, Shivam Kumar, Sumit kumar, Ashish Juneja

Abstract: This study focuses on developing eco-friendly fly ash bricks using silica fume and alkali activators (NaOH and Na₂SiO₃) as a sustainable alternative to traditional clay bricks. The objective is to utilize industrial waste effectively while reducing environmental degradation caused by clay brick production. Fly ash was used as the primary material, with silica fume added to enhance mechanical strength and alkaline chemicals to initiate geopolymerization. Bricks were prepared by mixing materials, molding, and proper curing. Tests including compressive strength, water absorption, and visual inspection were conducted. Results indicate that alkali-activated fly ash bricks with silica fume exhibit superior strength and durability compared to conventional clay bricks. This approach promotes waste utilization, low pollution, and energy-efficient construction, offering a promising solution for sustainable building practices.

Fabricartion Of Portable Noodle Making Machine

Authors: Jayanth L, Kaviraj M Kumkumgar, Kuldeep Raj M S, Madhuraj H R, Dr. Mohammad Rafi. H. Kerur

Abstract: This project (Phase 2) presents the fabrication of an innovative, portable noodle making machine aimed at providing a cost-effective and user-friendly solution for small-scale and home-based noodle production. Traditional noodle machines tend to be expensive, and require considerable expertise, limiting their accessibility especially for households and micro-entrepreneurs. To address this gap, the proposed machine utilizes a lightweight frame built from available materials, with food-grade stainless steel components for all parts in contact with the dough. The core mechanism involves a threaded extrusion system powered by a small electric motor, which efficiently transforms freshly prepared dough into uniformly shaped noodles. The process comprised conceptual sketches, noodle quality, and portability. Test results demonstrate consistent noodle extrusion with ease of operation and quick cleaning, making it suitable for diverse environments including homes, street vendors, and small eateries. The modular construction further enhances maintainability and transport convenience. This project not only offers a practical fabrication approach but also supports entrepreneurial activities by enabling affordable fresh noodle production. Overall, the project contributes an innovative, accessible, and sustainable noodle-making solution that promotes food variety and small business empowerment.

Effect Of Spent Mushroom Substrate-Based Compost Enriched With Micronutrients On The Productivity Of Maize (Zea Mays L.) And Soil Health

Authors: Pratyush Ranjan Sahu, Nishith Das

Abstract: The incorporation of agro-industrial residues like spent mushroom substrate (SMS) into nutrient management strategies provides a sustainable pathway for intensive agriculture. A field experiment was carried out during the Kharif of 2025 at GIET University, Odisha, to assess the impact of SMS-based compost supplemented with zinc (Zn), boron (B), and neem cake on the physiological, yield, and economic indices of maize (Zea mays L., var. VNR 4226). Utilizing a Randomized Block Design (RBD) with eight treatments and three replications, the study revealed that integrating SMS with micronutrients and the Recommended Dose of Fertilizers (RDF) significantly augmented crop performance. Treatment 8 (T8) (SMS@7t/ha + dried plant debris@2t/ha + cow dung@1t/ha + 5% Zn + B + RDF) delivered the highest plant stature (217.47 cm), maximum dry matter accumulation (237.63 g/plant), and superior yield attributes. This resulted in an exceptional kernel yield of 8.17 t/ha, a 175% increase over the FYM control. Soil chemical properties, notably available phosphorus, improved considerably under SMS regimes. Economically, T8 yielded the highest net monetary returns (₹1,18,642/ha), whereas T6 (RDF + 5% neem cake) optimized the Benefit-Cost ratio (2.38). These findings advocate for the integrated use of fortified SMS compost to enhance maize productivity and soil health.

Real-Time Sign Language Detection Using Computer Vision And Machine Learning

Authors: Assistant Professor. Sukanya H N, Adithya N, Akash H S, Farazulla Khan, G P Chinmayaradhya

Abstract: Sign language is the primary communication medium for deaf and hard-of-hearing individuals, yet it remains largely inaccessible to the general public, creating a persistent commu-nication barrier. This paper presents a real-time sign language detection system that leverages computer vision and machine learning to recognise hand gestures and convert them into readable text or speech with minimal latency. The proposed framework follows a structured processing pipeline comprising data acquisition, key-frame extraction, skin-colour-based hand segmentation, face-region elimination, morphological filtering, and noise reduction. Discriminative spatial features are derived using fuzzy triangular membership functions, and gesture recognition is performed by a K-Nearest Neighbour (Mediapipe) classifier trained on a self-collected dataset of two-handed dynamic signs. For real-time operation, the system employs the MediaPipe library for hand-landmark detection and a Convolutional Neural Network (CNN) trained with TensorFlow/Keras for gesture classification. Experimental evaluation demonstrates an overall gesture recognition accuracy of approximately 92%, with a high-confidence detection of 99.6% for the “Peace” gesture and an average detection-plus-translation latency of approximately 150 ms per frame. The system requires no specialised sensors or gloves, making it cost-effective and practically deployable in educational institutions, healthcare facilities, and public service environments. Results confirm the feasibility and effectiveness of the proposed approach as an assistive communication solution for hearing-impaired individuals.

Job Satisfaction Among Employees And Its Impact On Domestic Life

Authors: Priya kumari, Sohail Verma

Abstract: Job satisfaction is one of the most important aspects influencing employee performance, mental well-being, and overall quality of life. In the modern competitive work environment, employees often face workload pressure, stress, long working hours, and work-life imbalance, which directly affect their domestic and family life. This study examines the relationship between job satisfaction and employees’ domestic life and analyses how workplace conditions influence family relationships, personal happiness, and social well-being. The paper highlights factors such as salary, working conditions, job security, organizational support, work-life balance, and employee recognition in determining job satisfaction levels. The study also discusses how satisfied employees maintain healthier family relationships, lower stress levels, and improved domestic harmony, whereas job dissatisfaction may lead to emotional stress, conflicts, and reduced quality of life at home. The findings suggest that organizations should focus on employee welfare, flexible work policies, and supportive work environments to improve both job satisfaction and domestic well-being.

Impact Of Digital Payments On Daily Life. A New Setup

Authors: Sweta Pandey, Meentu Grover

Abstract: Digital payment systems have transformed the way people conduct financial transactions in their daily lives. The rapid growth of internet technology, smartphones, and financial technology has increased the adoption of digital payments across the world. In India, digital payment methods such as Unified Payments Interface (UPI), mobile wallets, internet banking, debit cards, and credit cards have become highly popular due to convenience, speed, and security. This paper examines the impact of digital payments on daily life and analyses how cashless transactions have influenced consumer behaviour, business activities, and economic growth. The study highlights the advantages of digital payments, including faster transactions, financial inclusion, transparency, reduced dependency on cash, and improved online shopping experiences. It also discusses challenges such as cyber fraud, privacy concerns, internet dependency, and lack of digital literacy among certain sections of society. The paper concludes that digital payments have significantly improved the efficiency and convenience of daily financial activities and will continue to play an important role in the future digital economy.

Analysis Of Risk Management In Construction Project.

Authors: Mahmud Danladi, Salihu Sarki Ubayi, Mahmud Danladi

Abstract: The construction industry is highly susceptible to uncertainties and risks that significantly influence project delivery in terms of cost, time, quality, safety, and sustainability. This study examined the analysis of risk management in project construction within the Nigerian construction industry. Specifically, the study identified the types of risks associated with project construction, examined the factors affecting risk management, and evaluated the effects of risk management on construction project performance. A descriptive quantitative research design was adopted. Data were collected through structured questionnaires administered to 80 construction engineers involved in risk management practices, out of which 74 valid responses were retrieved, representing a response rate of 92.5%. Descriptive statistical tools including frequency distribution, percentage analysis, mean item score, and standard deviation were used for data analysis. The findings revealed that inadequate site investigation, inadequate specification, contractor’s experience, weather implications, natural disasters, new technology, and shortage of resources were among the most significant risks affecting construction projects. Resource availability, project complexity, and time compression were identified as the major factors affecting risk management implementation. Furthermore, the study established that risk management strongly affects project cost, completion time, productivity, project quality, health and safety, and environmental sustainability. The study concluded that effective risk management is essential for successful construction project delivery and recommended proper site investigation, adequate resource allocation, experienced workforce engagement, and proactive risk management strategies to improve project outcomes in Nigeria.

DOI: https://doi.org/10.5281/zenodo.20324799

Smart Vendor AI: An AI-Driven Smart Vendor Management System For Real-Time Freshness Detection And Dynamic Retail Intelligence

Authors: Sudarshan K, Sushmitha H Y, Varshanth Gowda M L, Vinay C N

Abstract: Street vendors selling fruits and vegetables across India face a persistent challenge: perishable stock loses value as the day progresses, yet pricing remains static. This paper presents Smart Vendor AI, a complete end-to-end system that combines inventory management, point-of-sale operations, analytics dashboards, sales forecasting, and AI-assisted product quality assessment within a unified web-based platform.. The pipeline consists of six sequential layers: a fine-tuned YOLOv8s model for ripeness classification, a signal engine that converts raw predictions into weighted freshness scores, a deterministic market con- text module, an XGBoost pricing model trained on 5,000 realistic scenarios, a rule-based decision engine, and a FAISS-backed retrieval-augmented generation module powered by LLaMA 3.3 70B. Experiments on banana and tomato datasets show classifi- cation accuracy of 99.3% and 98.6% respectively. The system delivers specific, actionable vendor instructions—including an exact discount percentage and an inventory action string—without requiring any technical knowledge from the user. Results indi- cate meaningful potential to reduce the 30–40% annual revenue loss that vendors typically incur through spoilage and mispricing.

AGRONEXUS: An IoT-Based Real-Time Environmental Monitoring And Public Display Framework For Smart Campuses

Authors: Mrs. Pragati Sharma, Aman Chandel, Harsh Sharma, Priya Upadhyay, Safiya Naaz, Sunny Kumar, Tanu Saini, Vashu Dhiman

Abstract: The escalating degradation of environmental quality in educational institutions and public spaces demands cost -effective, real-time monitoring solutions. Conventional systems rely on centralised infrastructure or mobile applications that fail to deliver localised, immediate feedback. This paper presents AgroNexus, an IoT-driven environmental monitoring and public display platform that integrates the ESP32 microcontroller with four sensing modules—DHT22 (temperature and humidity), MQ135 (air quality), a rain sensor (precipitation detection), and DS3231 (real-time clock)—to deliver continuous data acquisition, threshold-based alerting, and live display via a six-panel P10 LED matrix. Experiments conducted in a simulated campus environment demonstrate that AgroNexus achieves high sensor accuracy, low false-alert rates, and sub-three-second display refresh cycles, outperforming single-sensor baselines across all evaluation metrics. The framework is economical, scalable, and readily deployable in smart campuses, industrial zones, and public spaces, establishing a transparent and auditable pipeline for environmental awareness.

DOI: http://doi.org/10.5281/zenodo.20325235

IOT Based Environment Monitoring System Using STM32

Authors: Mrs. Parul Gupta, Safiya Naaz, Priya Upadhyay, Mohd. Arshad, Tanu

Abstract: The rapid degradation of environmental quality driven by industrialization and urbanization demands continuous, real-time monitoring of key atmospheric and ecological parameters. This paper presents the design and implementation of a low-power, solar-powered IoT-based environmental monitoring system built around the STM32 microcontroller. The proposed system integrates a suite of sensors to measure temperature, humidity, atmospheric pressure, air quality, UV radiation, and soil moisture. Data is transmitted wirelessly over Wi-Fi and LoRa protocols to a cloud-based dashboard for real-time visualization and historical analysis. The system is entirely powered by a solar photovoltaic panel coupled with a lithium-ion battery and a power management unit, ensuring uninterrupted autonomous operation in remote locations without access to the electrical grid. Experimental results demonstrate reliable data acquisition with a sampling accuracy exceeding 97%, an end-to-end data transmission latency of less than 2 seconds, and continuous operation exceeding 72 hours on battery backup under cloudy conditions. The proposed system offers a cost-effective, scalable, and energy-autonomous alternative to conventional environmental monitoring stations.

DOI: https://doi.org/10.5281/zenodo.20325332

Real-Time AI-Based PPE Compliance And Safety Intelligence For Construction Sites

Authors: S. Santhosh Kumar, Dr. R. Senthamil Selvi

Abstract: The construction site is considered a risky place for employees, and the risks are associated with falling objects, machines, and exposure to harmful substances. Monitoring the implementation of Personal Protective Equipment (PPE) standards, including helmets, vests, gloves, boots, and masks, is of critical importance in preventing accidents and injuries. The conventional approach to monitoring the implementation of these standards is through manual observation, which is associated with time delays and human error. This study proposes an intelligent framework for the implementation of PPE standards and safety monitoring using an improved YOLOv11 deep learning model for the detection and classification of different types of PPE in real-time construction site video feeds. The model is trained on a diverse dataset to cater to complex backgrounds, lighting, occlusion, and multiple PPE pose angles, ensuring the model performs well in diverse site environments. The framework helps improve workplace safety by ensuring compliance, reducing the probability of accidents caused by negligence, and promoting regulatory compliance, thereby creating a culture of consistent PPE usage and safe work practices across the construction industry.

DOI: http://doi.org/10.5281/zenodo.20325549

Ai Image Fraud Detector

Authors: Shreya Shashikant Patil, Shital Nivrutti Sutar, Prachi Prasad Patil, Mrs . Meghana Khare

Abstract: Artificial intelligence has made it possible to generate highly realistic images, which can be mis used for misinformation, fraud and identity theft. Detecting such AI- generated images manually is difficult and time consuming. Detecting such AI-generated images has become very important to maintain the authenticity of digital content. This paper presents an AI Image Fraud Detector such that uses deep learning techniques to classify as real or fake. The system integrates YOLO (You Only Look Once) model with a web-based applications developed using Flask and JavaScript. Users can upload images through a user-friendly interface, and the system provides prediction result along with confidence scores. The model processes images in real time and ensures fast detection. Experimental results show that the system performs efficiently with good accuracy depending on the dataset quality. This research contributes to improving digital security by providing an automated solution for detecting AI-generated images. In this research, we developed an AI image fraud detection system using deep learning models such as VGG16, ResNet, and InceptionV3.Thesemodels are trained on a dataset containing both real and AI generated images. The system compares the performance of all three model to find which one give better accuracy. The model is trained on a dataset from Kaggle that contain both real and fake images of Aadhar- id photo and other documents. Image preprocessing techniques are used to improve performance of the model. The result show that deep learning models can effectively detect fake images, with one model performing better based on accuracy and efficiency. The study highlights that using multiple models improve reliability and provides a strong solution for detecting AI-generated images in real world applications. We also tested different settings of the model to understand what works best. Our study shows that it is a strong and reliable method for detecting AI-generated images and can be useful in real-world applications. Model is addressing the increasing challenge of AI-generated image detection, laying a foundation for future research in critical area.

Review Paper On Advance Robotic Arm Hand With Object Detection Vehicle

Authors: Prof. V. U. Bansude, A. S. Yadav, D. D. Pawal, A. S. Yadav

Abstract: The robotic arm is one of the most significant innovations in the field of automation and robotics, capable of replicating human arm movements with high precision, accuracy, and repeatability. Over the past decades, researchers have developed robotic arms for various applications such as industrial manufacturing, medical surgery, agriculture, space exploration, and defense operations. Early robotic arm systems were limited to simple wired control and basic pick-and-place operations. However, recent advancements have integrated modern technologies including artificial intelligence (AI), computer vision, machine learning, and Internet of Things (IoT) to achieve intelligent and autonomous functionality. This paper presents a comprehensive survey of existing robotic arm systems with emphasis on their design methodologies, actuation techniques, control mechanisms, and practical applications. A comparative analysis of various research works has been conducted to understand the technological evolution and identify limitations in current robotic arm systems. The study also highlights future opportunities for developing intelligent robotic arms capable of performing complex real-world tasks with improved efficiency and reliability.

Laro-based Wearable Women Safety Alert System

Authors: Amrutha H, Chaithra HM, Chandana BM, Chethana GH, Mr. Santhosh Babu KC Assistant Professor

Abstract: Women's safety remains a critical global concern, with increasing incidents of harassment, assault, and emergencies requiring immediate intervention. Traditional safety devices such as panic buttons and mobile applications have limitations: they rely on cellular connectivity, which may be unavailable in remote areas, and they lack automatic fall detection for situations where the user cannot manually trigger an alert. This project presents a comprehensive LoRa based women safety device that combines manual panic activation, automatic fall detection, and dual communication channels for maximum reliability. The system consists of two units: a portable transmitter unit carried by the user and a stationary receiver unit placed at a trusted location such as home, workplace, or police station. The transmitter unit uses an ESP32 microcontroller with a panic button for manual emergency activation and an MPU6050 sensor for automatic fall detection. When an emergency is detected, the transmitter sends an alert via LoRa wireless communication (operating at 433MHz) over long distances (several kilometers). Simultaneously, a GSM800L module sends an SMS alert directly to authorities or emergency contacts. The receiver unit, comprising another ESP32 with a LoRa module, buzzer, and LCD display, receives the LoRa transmission, displays the alert message on the LCD, and activates an audible buzzer to notify personnel at the receiving location. This dual-path communication ensures that even if one channel fails (GSM network down or LoRa interference), the other channel may still deliver the alert. The system is designed to be wearable, low-power, and effective in both urban and remote areas where cellular coverage may be unreliable.

</p

Natural Space As A Transformative Environment For Childrens Well-being

Authors: Murmylo Yulia

Abstract: Children's happiness” is the principal vector of any society and the foundation on which the Sustainable Development Goals rest today, tomorrow and for the next generation. This multidimensional term encompasses complex components, without each of which it remains incomplete. We examine the interrelation between the phenomenon of children's happiness and nature-based practices (in the context of an environment in which, through neurobiological, sensory and interpersonal mechanisms, qualitative changes take place in the child's personality, emotional repertoire, cognitive strategies and immune profile). We review existing methodologies, the international studies that have been conducted on this topic, and their results, and draw a conclusion about the most effective practices contributing to the enhancement of children's happiness. This article is unique in that it identifies a set of aspects of children's well-being, presents concrete methodologies for analysing this multifaceted concept, lays out natural factors of influence, summarises a research base on the impact of nature on the younger generation across individual components, and describes working programmes that demonstrate the action of the natural environment on children as transformative. The author argues that, from the standpoint of sustainable development, nature-oriented programmes possess a unique property: they are simultaneously a tool for achieving goals (improving children's health and well-being) and a means of forming agents of sustainable development in the next generation. Adapting the principles of the Stanford course “Interpersonal Dynamics” to nature-based programmes for children opens up the possibility of creating a new class of pedagogical products.

DOI: http://doi.org/10.5281/zenodo.20341762

SkillLink: A Web-Based Peer-to-Peer Skill Exchange And Mentoring Platform With AI-Assisted Session Management

Authors: Manoj S, Chaitra B P, Nandan J M, Nehal Eldho Binu

Abstract: SkillLink is a web-based peer-to-peer mentoring platform designed to enable real-time skill exchange between learners and teachers. The system is developed using the MERN stack and integrates WebRTC for browser- based video conferencing, Socket.IO for real-time communication, and the Gemini API for AI-assisted interaction. Teachers publish skills and availability through a drag-and-drop calendar interface, while learners can browse and book sessions directly. The platform includes session lifecycle management, subscription-based access control, a credit-based reward system, and a five-star rating mechanism. Experimental evaluation demonstrates low-latency communication, reliable session tracking, and efficient mentor matching, making SkillLink a scalable alternative to conventional e-learning systems.

Smart Attendance System Using Face Recognition

Authors: Shital Vishwanath Ban, Shankar Sanjay Rathod, Prerana Prakash Malgave, Mrs. M.R, Raste

Abstract: Traditional attendance systems are time-consuming and prone to errors such as proxy attendance. This paper presents a Smart Attendance System using Face Recognition technology. The system automatically detects and recognizes faces to mark attendance. It uses machine learning and image processing tech-niques for accurate identification. The system captures real-time images through a camera, processes them, and updates attendance records. It reduces manual effort and improves accuracy. The system is implemented using Python, OpenCV, and a database for storing attendance data.

MedLens: An AI-Powered Radiology Report Simplification System for Improved Patient Accessibility

Authors: B. M. Promod Kumar, Bhavana N. S., C. Chinmayi, Deepthi C. Shekar, Deenadayal B. K.

Abstract: Radiology reports generated from imaging modalities such as X-rays, Computed Tomography (CT), Magnetic Resonance Imaging (MRI), and ultrasound scans are critical clinical documents. However, these reports are authored in complex medical terminology intended for radiologists and specialist physicians, rendering them largely inaccessible to patients and non-medical users. This communication gap results in confusion, anxiety, and increased dependency on healthcare professionals for basic explanations. This paper presents MedLens, an AI- powered radiology report simplification system that bridges this gap by leveraging Natural Language Processing (NLP) and Retrieval-Augmented Generation (RAG). The system extracts text from uploaded PDF reports using PyMuPDF, processes clinical content using Google Gemini AI models, and generates accurate, context-aware patient-friendly summaries. It further classifies the urgency of findings into levels (Low, Moderate, High, Critical), and integrates multilingual translation, text-to-speech functionality, and an AI-powered contextual chatbot. The platform is deployed using FastAPI on the backend and React.js with Tailwind CSS on the frontend. Experimental results demonstrate that MedLens successfully simplifies complex medical terminology, detects critical conditions, provides multilingual support, and enables interactive report-based queries, thereby empowering patients with better health awareness and facilitating informed discussions with healthcare providers.

DOI: https://doi.org/10.5281/zenodo.20351715

Class-Balanced Knowledge Distillation for Imbalanced Urban Vehicle Detection on CAVI-14

Authors: Parag Hossain

Abstract: Urban vehicle detection systems face a fundamental challenge that is often overlooked in benchmark datasets: severe class imbalance. In real-world traffic scenes, common vehicles such as cars appear thousands of times more frequently than critical but rare categories including ambulances, e-bikes, and motorcycles. This imbalance causes standard detectors to become biased toward majority classes, leading to unacceptable failure rates for minority class detection in safety-critical applications. In this paper, we propose a novel Class-Balanced Knowledge Distillation (CBKD) framework specifically designed to address this challenge on the challenging CAVI-14 dataset, which contains fourteen urban vehicle categories with up to fifteen-fold class imbalance. Our method integrates three key components: class-balanced sampling to ensure equal exposure to all classes during training, focal loss with class-specific weights to down-weight easy majority examples, and knowledge distillation from a teacher model pretrained on a synthetically balanced dataset. Extensive experiments demonstrate that CBKD achieves perfect mean average precision at 0.50 intersection-over-union threshold (mAP50) of 1.000 and near-perfect mAP50-95 of 1.000 after one thousand training epochs. Per-class F1 scores consistently exceed 0.97 across all fourteen categories, including the rarest classes. Qualitative results on validation images show accurate detection even under heavy occlusion and challenging lighting conditions. Our approach establishes a new state-of-the-art on the CAVI-14 dataset and provides a practical, reproducible solution for imbalanced object detection in intelligent transportation systems.

DOI: https://doi.org/10.5281/zenodo.20351972

AgriHub: An AI-Powered End-to-End Agricultural Decision Support Platform

Authors: Mohammed Munyim Hussain V, Poorvaj K P, Prashanth S R, Preetham M, Mr P Prasanna

Abstract: Agriculture remains a cornerstone of economic activity across developing nations, yet smallholder farmers routinely face yield gaps caused by uninformed decisions on crop selection, soil nutrition, and disease management. This paper presents AGRI HUB, a web-based Crop and Soil Management System that unifies several machine-learning and deep-learning services behind a single Flask-driven interface. Four core modules are delivered: (i) smart crop recommendation using a Random Forest classifier trained on seven agro-climatic parameters, achieving 99.55% accuracy across 22 crop classes; (ii) soil nutrient analysis and fertilizer recommendation through NPK deficit computation against crop-specific thresholds; (iii) plant disease detection using a ResNet-9 convolutional neural network capable of classifying 38 disease categories from leaf photographs; and (iv) real-time, weather-driven activity planning by consuming OpenWeatherMap API data to generate seven-day farming calendars. An AI chatbot powered by the Google Gemini large language model supplements the analytical modules with conversational agronomic guidance. A crop profitability comparison dashboard rounds out the system, enabling evidence-based economic decisions. Experimental evaluation confirms that the integrated platform consistently outperforms single-module alternatives in both accuracy and decision breadth, offering a scalable, cost-effective tool for precision agriculture.

AgriHub: An AI-Powered End-to-End Agricultural Decision Support Platform

Authors: Govardhan Jadhav, Anand Ahire, Hitesh Kalal, Rishi Mishra, Prof.S.R.Agrwal

Abstract: Social media usage has increased significantly in recent years, leading to concerns about addictive behavior and its impact on users’ productivity and mental well-being. This paper presents a Social Media Addiction Tracker system designed to monitor, analyze, and manage user engagement across various platforms. The proposed system collects data such as screen time, frequency of usage, and interaction patterns, and applies data analytics and machine learning techniques to identify signs of excessive usage and potential addiction. Based on the analysis, the system provides real-time feedback, usage reports, and personalized alerts to help users regulate their social media habits. Experimental evaluation demonstrates that the system effectively raises user awareness and supports behavior modification. The proposed solution aims to promote healthier digital habits and improve overall well-being.

Incentive-Driven Social Media Usage Regulation System

Authors: Govardhan Jadhav, Anand Ahire, Hitesh Kalal, Rishi Mishra, Prof.S.R.Agrwal

Abstract: Social media usage has increased significantly in recent years, leading to concerns about addictive behavior and its impact on users’ productivity and mental well-being. This paper presents a Social Media Addiction Tracker system designed to monitor, analyze, and manage user engagement across various platforms. The proposed system collects data such as screen time, frequency of usage, and interaction patterns, and applies data analytics and machine learning techniques to identify signs of excessive usage and potential addiction. Based on the analysis, the system provides real-time feedback, usage reports, and personalized alerts to help users regulate their social media habits. Experimental evaluation demonstrates that the system effectively raises user awareness and supports behavior modification. The proposed solution aims to promote healthier digital habits and improve overall well-being.

Hybrid Generative Artificial Intelligence and Quantum-Mechanical Screening for Accelerated Drug Lead Optimization

Authors: Prof. R. Raveendhra

Abstract: Artificial intelligence (AI) is transforming pharmaceutical research by enabling rapid molecular prediction, virtual screening, and biological data integration. However, many current AI systems lack energetic realism and mechanistic interpretability. This manuscript presents a conceptual framework termed Adaptive Quantum-Generative Optimization (AQGO), integrating generative AI, molecular transformers, quantum-mechanical screening, molecular docking, and expert pharmacological validation. The framework is designed to improve lead optimization by combining data-driven molecular generation with physics-based energetic evaluation. The article reviews current advances in AI-driven drug discovery, the role of quantum chemistry in molecular simulation, translational challenges, and future directions for hybrid AI–quantum systems. Emphasis is placed on explainability, reproducibility, ethical deployment, and scientific transparency. The proposed architecture highlights the potential of combining generative intelligence with quantum-mechanical validation to support more efficient and reliable pharmaceutical discovery pipelines.

Smart Classroom and Digital Learning

Authors: Lalita Sonawane

Abstract: Smart classrooms and digital learning are changing education with the help of technology. Tools like smart boards, projectors, online classes, artificial intelligence, and virtual classrooms help students learn in an easy and interesting way. During the COVID-19 pandemic, online learning became very important because schools and colleges were closed. This research paper explains the meaning, benefits, challenges, and future of smart classrooms and digital learning. The information for this paper was collected from books, journals, websites, and research articles. The study shows that smart classrooms improve communication between teachers and students, increase student participation, and provide flexible learning opportunities. Students can study anytime and anywhere through digital platforms. However, there are also some problems like poor internet connection, high technology cost, lack of digital skills, and cyber security risks. The paper concludes that smart classrooms and digital learning are important for the future of education and need proper support, training, and infrastructure.

A Proximal Adaptive Momentum Algorithm with Variance Reduction for Nonconvex Composite Optimization: Convergence Analysis and Complexity Bounds

Authors: Dr.K.Srinivasan, Dr. M. K. Vediappan

Abstract: We propose and analyze the Proximal Adaptive Momentum with Variance Reduction (PAMVR) algorithm, a novel first-order method for solving nonconvex composite optimization problems of the form min F(x) = f(x) + g(x), where f is a smooth nonconvex function and g is a proper convex, lower-semicontinuous regularizer. PAMVR integrates three complementary mechanisms: (i) a momentum-corrected gradient estimator with adaptive step sizes, (ii) a periodic variance-reduction snapshot strategy inspired by SVRG, and (iii) a proximal operator for handling the nonsmooth component. Under standard Lipschitz-gradient and bounded-variance assumptions, we establish global convergence to an epsilon-approximate stationary point with a sample complexity of O(n + n^{2/3}/epsilon^2) stochastic gradient evaluations, matching the best-known bounds for this problem class while requiring weaker algorithmic assumptions than existing momentum-based methods. We further prove almost-sure convergence of the iterate sequence under a Kurdyka-Lojasiewicz (KL) regularity condition, obtaining explicit convergence rates depending on the KL exponent. The theoretical findings are validated on benchmark nonconvex problems including sparse logistic regression, matrix completion, and neural network training, demonstrating consistent improvements of 15–32% in convergence speed over PROX-SVRG, ProxGD-M, and Spider-Boost baselines. These results establish PAMVR as both a theoretically sound and practically competitive method for large-scale nonconvex optimization.

DOI: https://doi.org/10.5281/zenodo.20354108

A Study On The Impact Of Financial Literacy On Financial Decision-Making Among College Students With Special Reference To Coimbatore District

Authors: Ms. Nandhini R, Mr. Mohan Kumar

Abstract: Financial literacy plays a significant role in enabling individuals to make informed and effective financial decisions in their daily lives. In today’s rapidly changing financial environment, college students are increasingly required to manage personal finances, including budgeting, saving, investing, and controlling expenses. However, many students lack adequate financial knowledge and awareness, which may lead to poor financial behaviour and long-term financial instability. This study aims to assess the level of financial literacy among college students and examine its influence on their financial decision-making behaviour. The research also focuses on identifying the major sources of financial information used by students and analysing the relationship between financial literacy and saving habits. The study highlights the importance of financial education in developing responsible financial behaviour among young adults. By identifying gaps in financial awareness, the research provides useful insights and recommendations for improving financial literacy programs and promoting better financial management practices among students for a financially secure future.

DOI: https://doi.org/10.5281/zenodo.20354178

A Study On Work From Home And Its Impact On Employee Productivity, Work-Life Balance And Job Satisfaction With Special Reference To It And Bpos Employees In Coimbatore City

Authors: Dr. Kowsalya G, Ms. Kowsalya

Abstract: The COVID-19 pandemic catalysed the most rapid and large-scale transition to remote work in human history, transforming work from home from a marginal flexibility benefit into the dominant mode of employment for millions of knowledge workers globally and in India. While the immediate crisis has subsided, hybrid and fully remote work arrangements have become a permanent feature of the employment landscape, particularly in the Information Technology and Business Process Outsourcing sectors that constitute two of Coimbatore's most significant and fastest-growing industries. This study examines the impact of work from home arrangements on employee productivity, work-life balance, and job satisfaction among IT and BPO employees in Coimbatore city. Primary data were collected through a structured questionnaire administered to 120 IT and BPO professionals currently working from home or in hybrid arrangements. Secondary data were gathered from academic journals, NASSCOM reports, SHRM publications, and government employment surveys. Statistical tools including simple percentage analysis, weighted average method, and chi-square test were employed. Findings reveal that while work from home significantly improves perceived productivity and time flexibility for a majority of respondents, challenges in work-life boundary maintenance, social isolation, and home infrastructure quality create significant well-being risks that require proactive organisational and policy intervention.

DOI: https://doi.org/10.5281/zenodo.20354225

A Study On The Socio-Economic Impact Of The Tamil Pudhalvan Scheme On Low-Income Families

Authors: Ms. Dr.B. Geethpriya, Mr. M. Balakumar

Abstract: This study examines the socio-economic impact of the Tamil Pudhalvan Scheme a state-funded educational welfare initiative introduced by the Government of Tamil Nadu on low-income families residing in the Coimbatore district. A structured questionnaire was administered to 151 beneficiary respondents selected using simple random sampling. Data were analyzed using percentage analysis and one-way Analysis of Variance (ANOVA) to test whether statistically significant differences exist in perceived financial relief, educational motivation, and dropout-prevention effectiveness across educational level groups (school, diploma, undergraduate, and postgraduate). ANOVA results indicate significant between-group differences in financial impact (F(3, 147) = 4.82, p = .003) and educational motivation (F(3, 147) = 3.61, p = .015), while dropout prevention did not reveal statistically significant variation (F(3, 147) = 2.14, p = .097). The findings suggest that the scheme delivers differentiated benefits depending on the student's level of education. Post hoc Turkey HSD tests revealed that postgraduate students perceived significantly greater financial relief compared to school-level beneficiaries. Policy recommendations include increasing the monthly stipend, expanding digital outreach, and integrating the scheme with vocational skill programmes.

DOI: https://doi.org/10.5281/zenodo.20354253

A Study On Ai-Driven Consumer Segmentation And Social Marketing Strategies For Sustainable Water Purification Businesses In Coimbatore City

Authors: Ms. Revathi G, Mr. Ashwath R

Abstract: The study adopts a descriptive research design and is based on both primary and secondary data. Primary data was collected from 150 respondents using a structured questionnaire, while secondary data was gathered from journals, articles, and online sources. The research focuses on identifying consumer segments, analyzing the impact of AI in understanding customer preferences, and evaluating the effectiveness of social marketing strategies in influencing consumer awareness and purchasing behavior. The study adopts a descriptive research design and is based on both primary and secondary data. Primary data was collected from 150 respondents using a structured questionnaire, while secondary data was gathered from journals, articles, and online sources.

DOI: https://doi.org/10.5281/zenodo.20354287

Exploratory And Visual Analytics Of Mtcars Dataset Using Tableau Tool

Authors: Cherukupalli Harshitha, Darapu Saivenkat, Muvvapati Koushik, Mrs.K.Sireesha

Abstract: The mtcars11 dataset provides complete data about vehicle performance and their corresponding features. The dataset includes essential features which measure fuel efficiency through miles per gallon and provide engine specifications and horsepower and vehicle weight and transmission type and driving conditions. The information assists in examining trends associated with vehicle effectiveness, performance, and operational conduct. Through the use of data visualisation methods on this dataset, we seek to comprehend how elements such as weight, engine power, and transmission type affect fuel efficiency and overall performance. It also aids in recognising patterns under various driving circumstances like traffic and weather. The knowledge acquired can enhance decision-making in automotive evaluation, vehicle development, and performance improvement.

DOI: http://doi.org/10.5281/zenodo.20355470

Deep Shield: Protecting Against Deepfakes

Authors: Dr. M. C. Padma, Bhoomika M, Faika Mehvish, Praveen Kumar R

Abstract: The rapid proliferation of deepfake videos—synthesised using Generative Adversarial Networks (GANs) and allied deep-learning techniques—poses grave risks to societal trust, democratic processes, and personal privacy. Existing detection approaches predominantly rely on frame-level spatial analysis and consequently fail to capture temporal inconsistencies that arise in manipulated sequences. This paper presents Deep Shield, a hybrid deep-learning framework that couples a ResNeXt convolutional neural network (CNN) for spatial feature extraction with a Long Short-Term Memory (LSTM) recurrent network for temporal sequence modelling. Each video frame is first preprocessed via face detection and alignment, after which ResNeXt encodes per-frame spatial embeddings that are subsequently fed into the LSTM to capture inter-frame inconsistencies. A fully connected classifier then labels the video as Real or Fake alongside a confidence score. The system is validated on three benchmark datasets—FaceForensics++, DFDC, and Celeb-DF—achieving detection accuracy exceeding 99 % together with precision, recall, and F1-score values above 99 %. The framework is wrapped in a Django-based web interface that allows nontechnical users to upload videos and obtain results in near real time. Robustness testing under compression artefacts, low-light conditions, and adversarial inputs confirms the generalisability of the approach.

DOI: http://doi.org/10.5281/zenodo.20375755

Biometric Smart Attendance System

Authors: Prof. Dr H R Divakar, Kavana S, K P Renuka Prasad, Manoj Kumar S R, Lipika K

Abstract: The traditional attendance system used in educational institutions is often time-consuming, error-prone, and vulnerable to proxy attendance. Existing hardware-based attendance solutions such as RFID systems require additional infrastructure, maintenance cost, and dedicated devices, making them less flexible and expensive for large-scale deployment. To overcome these limitations, this paper presents a Biometric Smart Attendance System that combines biometric verification, AI-based face verification, and GPS location validation to ensure secure, accurate, and reliable attendance management without using dedicated hardware components. The proposed system uses multi-factor authentication to verify student identity before marking attendance. Face verification technology identifies and authenticates students, while biometric authentication provides an additional layer of security. GPS-based location verification ensures that attendance can only be marked when the student is physically present within the authorized classroom premises. The system is developed using React, TypeScript, FastAPI, Python, DeepFace, OpenCV, and Supabase. Test results show face verification accuracy of 96% under normal lighting and overall attendance accuracy of 95%.

DOI: https://doi.org/10.5281/zenodo.20374000

Earthquake-Induced Behaviour Study Of Multi-Storey Irregular Buildings Using ETABS

Authors: Prof. Shyam Prasad H R, Rakshitha V, Rohan K M, Naveen S, Supriya R S

Abstract: The increasing development of urban areas and architectural requirements often lead to the construction of multi- storey reinforced concrete (RC) structures that do not comply with the symmetric and uniform structure that was assumed in classical seismic design theory.Multi-storey reinforced concrete (RC) structures are commonly built with an eccentric and non- uniform structure due to urban developments and architectural requirements, which do not match the symmetric and uniform structure used in classic seismic design theory. In this study, the behaviour of a G+7 RC building with an L-shaped plan irregularity and soft-storey vertical irregularity is studied under three structural configurations: Bare Frame, Infill Wall Frame and Shear Wall Frame and modeled in ETABS. The response spectrum analysis has been done according to IS 1893 (Part 1):2016, gravity loads as per IS 875 (Parts 1-3):1987, RC design as per IS 456:2000, ductile detailing as per IS 13920:2016. These values of storey drift ratio, lateral stiffness, storey shear and storey displacement were extracted and compared. The Shear Wall structure reduced the roof level lateral displacement by ~99.9% and storey drift by ~99.8–99.9% compared to the Bare Frame, and provided around five times the stiffness. The displacement reduction for the Infill Wall model ranged from intermediate (62–78%). A clear performance hierarchy was generated: Shear Wall > Infill Wall > Bare Frame, with shear wall being the critical member in an irregular building in high seismic zone construction, to ensure safety.

DOI: http://doi.org/10.5281/zenodo.20375845

Traffic Sign Recognition

Authors: Prof. P.S. Togrikar, A.N.Jamdade, P.R.Shirke, H.J.Phadtare

Abstract: Traffic Sign Recognition (TSR) is an essential component of Advanced Driver Assistance Systems (ADAS) and intelligent transportation. This paper presents a cost-effective IoT-based TSR system using an ESP32-CAM for image acquisition and a backend server for processing. Due to limited edge-device capability, images are transmitted via a Telegram Bot for remote inference using the YOLOv3 deep learning model trained on the GTSRB dataset. To enhance robustness under real-world conditions such as occlusion and varying illumination, preprocessing techniques like CLAHE and data augmentation are applied. The system returns annotated results through a Telegram interface and a local GUI. Experimental results demonstrate high accuracy and reliable performance, validating the effectiveness of the proposed approach. The system also shows strong performance under partially occluded conditions, improving real-world applicability. Furthermore, the proposed architecture ensures low-cost deployment and scalability for smart transportation systems. This work highlights the potential of integrating IoT with deep learning for practical and accessible traffic monitoring solutions.

Advanced Experimental Techniques (growth & fabrication) of semiconductor nanostructures: From morphology to electronic states

Authors: Pragati Sharma, Bhomik Nahariya, Aryan Rajput, Vansh

Abstract: In view of their size-dependent properties, both physical and chemical, semiconductor nanostructures have emerged as an essential component within modern nanotechnology. Novel device functionalities and adaptable electronic states are being established as possible by having the ability to accurately tune morphology, from zero-dimensional quantum dots to one-dimensional nanowires and two-dimensional thin films. The link between structural morphology and electronic characterization is demonstrated in this paper's assessment of sophisticated experimental methods for the growth and manufacturing of semiconductor nanostructures. Alongside top-down techniques such as lithography and etching, molecular beam epitaxy (MBE), chemical vapor deposition (CVD), atomic layer deposition (ALD), and laser ablation are also presented. In addition, it focuses on the ways in which defects, interfaces, and quantum confinement influence electronic states.

DOI: https://doi.org/10.5281/zenodo.20375584

Application Of Nature-Based Solutions For Climate Change: A Comprehensive Review And Feasibility Study

Authors: R. Swetha

Abstract: Climate change poses one of the most formidable challenges to global ecological and socioeconomic stability in the twenty-first century. As atmospheric concentrations of greenhouse gases continue to rise, the scientific community has increasingly turned to Nature-Based Solutions (NbS) as a viable and cost-effective complementary strategy to conventional technological mitigation approaches. This report provides a systematic analysis of Nature-Based Solutions, examining their mechanisms, classifications, documented effectiveness, and real-world implementation challenges. NbS encompass a spectrum of ecosystem-centred interventions — including reforestation, wetland restoration, urban greening, and sustainable agricultural practices — that simultaneously deliver climate mitigation benefits while enhancing biodiversity and community resilience. Key findings of this report indicate that NbS possess the theoretical capacity to contribute between 10 and 12 gigatons of CO₂ equivalent reductions annually by 2030, representing approximately 30% of the mitigation required to limit global warming to 1.5°C. However, this potential is contingent upon significant upscaling of political commitment, financial investment, and cross-sector governance frameworks.

Application Of Nanobubble Technology In Wastewater Treatment For Enhanced Pollutant Removal: A Comprehensive Review

Authors: Manuela Christy Dany S, Dr. Nithyalakshmi B

Abstract: Keeping global water resources clean is becoming harder every year. Industrial growth has pushed wastewater systems into a corner, and the usual treatment methods are starting to look worn out. They demand a lot of energy and still struggle with stubborn pollutants that refuse to break down. This review takes a close look at nanobubble (NB) technology as a more sustainable option. Nanobubbles are tiny, sub-micron gas cavities with an unusually long life in water, and that alone makes them interesting. They also bring unusual physicochemical properties, including high internal pressure and the formation of reactive oxygen species (ROS). The paper covers the basic mechanisms behind NBs, their contribution to aeration and flotation, and their strong performance in removing organic dyes, nutrients, heavy metals, and pathogens. Reported studies show that NB-based systems can push Chemical Oxygen Demand (COD) removal above 90% while using much less energy than conventional activated sludge treatment. The aim here is straightforward: give researchers and practitioners a clear view of the technical value and economic promise of NB technology in modern water purification.

EcoSort: An AI-Powered Garbage Segregation System Using MobileNetV3 And Deep Transfer Learning

Authors: Sukanya H N, Assistant Professor, Pavan Kumar T S, Prajwal S Shetty, Pranay Ekunde, Sanjay M

Abstract: Improper waste disposal remains one of the most pressing environmental challenges in both urban and rural settings, contributing to pollution, health hazards, and reduced recycling efficiency. Traditional manual waste segregation is error-prone, labour-intensive, and cannot scale to the volumes of waste generated daily. This paper presents EcoSort, an AI-powered full-stack web application that automates waste classification using a fine-tuned MobileNetV3 Large deep learning model trained via transfer learning. The system classifies waste images into three categories—Recyclable, Non-Recyclable, and Hazardous—achieving approximately 94 % overall accuracy with precision values of 0.95, 0.94, and 0.94 respectively. EcoSort integrates real-time webcam-based detection, a microservices architecture (React/Vite front-end, Node.js/Express back-end, Flask AI service, MongoDB Atlas), Role-Based Access Control (RBAC), JWT authentication, and perceptual hashing (pHash) for duplicate-image detection. A gamification layer comprising reward tiers (Bronze to Platinum), a coupon marketplace, and a community leaderboard motivates responsible waste disposal. Load testing confirmed stable operation under 100 concurrent users with average response times below 3.5 seconds. The platform aligns with UN Sustainable Development Goals SDG 3, SDG 11, SDG 12, and SDG 13, offering a scalable, intelligent pathway toward smarter waste management.

Sponge City Concept For Sustainable Stormwater Management: A Comprehensive Review

Authors: Darshana N V, Nithyalakshmi.B

Abstract: The fast pace of urbanization and worsening climate-driven stressors have disrupted the natural cycles of urban hydrological processes, making existing linear infrastructures increasingly susceptible to extreme pluvial flooding events. The Sponge City concept can be seen as an essential paradigm shift towards a decentralized nature-based method for urban areas to manage rainwater in terms of its absorption, storage, infiltration, and purification.In this review paper, we synthesize empirical data, policies, and hydrological models of ten key studies to examine the effectiveness of the Sponge City paradigm at various scales. This paper will analyze the development trends of LID-based structural controls, quantitative limitations for peak flows, life-cycle maintenance challenges, and multiple ecological benefits. The synthesized literature reveals that although green infrastructure produces impressive hydrological and economic benefits when dealing with conventional rainfall, its performance suffers considerably when confronted with an extreme cloudburst. Therefore, this paper sets up a robust research agenda for future urban planners, namely that a mandatory paradigm must be embraced in the form of a "green-gray hybrid infrastructure" system with institutional and technological arrangements for real-time monitoring.

Irrigation System (Kuhl) in Himachal Pradesh

Authors: Ritik Rana

Abstract: Sustainable agricultural production largely depends on the proper development, conservation, and efficient utilization of irrigation resources at the micro level. In Himachal Pradesh, diverse geographical conditions have led to the adoption of traditional irrigation systems known as Kuhls, which play a vital role in supporting agriculture and rural livelihoods. Kuhls are narrow, manually constructed surface channels that divert water from natural streams and ravines through gravity flow to irrigate terraced fields. These systems, built mainly with local materials such as river boulders and soil, represent an eco-friendly and community-managed method of irrigation. Despite their historical and agricultural significance, Kuhls today face several challenges including structural deterioration, water losses, changing climatic conditions, and inadequate maintenance. The present study focuses on the Palampur region of Himachal Pradesh to examine the problems associated with traditional Kuhl irrigation and explore modern solutions for improving irrigation efficiency and agricultural sustainability. The study highlights the need for technological improvements, conservation measures, and integrated water management practices to preserve this traditional irrigation heritage while meeting present-day agricultural demands.

DOI: https://doi.org/10.5281/zenodo.20390582

Augmented Reality-Based Interactive Solar System Visualization

Authors: Aryan Baban Repe, Atharv Narayan Rane, Harshvardhan Vijay Desai, Keshiraj Mahesh Lad, Sumit Vasant Bhatane, Mrs. Anuradha S. Solanki

Abstract: Augmented Reality (AR) has gained considerable traction in educational settings, offering interactive three-dimensional experiences that go beyond what conventional two-dimensional instructional materials can provide. This paper describes the design, development, and evaluation of an AR-Based Interactive Solar System Visualization system built using Unity 2022 and the Vuforia Engine. The system employs markerless ground-plane detection to overlay a fully interactive three-dimensional Solar System model onto the user’s physical surroundings, supporting planetary orbital revolution, axial rotation, touch-based planet selection, and dynamic educational information panels. The primary contribution of the proposed system is the integration of stable ground-plane AR tracking, structured educational interfaces, and a modular software architecture within a lightweight mobile deployment requiring only a standard Android smartphone. Prototype evaluation on Android devices yielded an average frame rate of 45–60 FPS, AR tracking accuracy of approximately 92%, an interaction response time below 100 ms, and a user satisfaction score of 88%, indicating measurable gains in learner engagement and conceptual retention relative to conventional instructional methods.

Design And Implementation Of An IoT-Based Smart Blind Assistive Stick For Visually Impaired Individuals

Authors: Prof. P. Prasanna, Arjun C, Chinmay J, Deepak B B, Dhanush S Yadav

Abstract: Visually impaired individuals face severe challenges in independent navigation. Traditional white canes provide only contact-based obstacle detection and cannot warn users of overhead hazards, wet surfaces, or distant obstacles—significantly limiting their mobility, safety, and self-reliance. Existing com-mercial smart canes, while technologically superior, remain prohibitively expensive and require specialised training. This paper presents the design and implementation of a Smart Blind Assistive Stick: a low-cost, IoT-enabled, real-time navigation aid built around an ATmega328 microcontroller. The system integrates an HC-SR04 ultrasonic sensor for non-contact obstacle detection up to 2 m, a moisture sensor for wet-surface identification, and multi-modal feedback through a vibration motor and buzzer. An optional GSM/GPS module enables real-time location sharing with caregivers via SMS. The firmware was developed in Embedded C on the Arduino IDE, simulated in Proteus, and physically prototyped. Testing confirmed obstacle detection accuracy within a 2 m range, sub-100 ms system response time, reliable wet-surface detection, and successful emergency SMS transmission with live GPS coordinates. The device operates for 6–8 hours on a rechargeable lithium-ion battery pack and maintains a compact, lightweight form factor suitable for daily indoor and outdoor use. Results demonstrate that a well-integrated multi-sensor embedded system can effectively bridge the technological gap in assistive mobility devices.

DOI: http://doi.org/10.5281/zenodo.20391169

RAG & LLM Based TNEA Student Assistant For Academic Guidance

Authors: Ms. K. Sabitha, B. Monish, M. Nithishkumar, S. Mohammed Al Ameen, S. Samvarthini

Abstract: The process of selecting an appropriate engineering course and college has become increasingly challenging due to the large volume of information and the complexity of admission procedures such as TNEA. Students often face difficulty in understanding cutoff trends, identifying suitable colleges, and making informed decisions because the available information is scattered and sometimes unreliable. To address this issue, this project proposes an intelligent academic assistance system that combines Retrieval-Augmented Generation (RAG) with Large Language Models (LLMs). The system is designed to provide accurate and user-friendly guidance by retrieving verified academic data and presenting it through an interactive conversational interface. The retrieval component ensures that information such as cutoff marks and college details is obtained from structured datasets, while the language model supports explanation-based queries related to courses and career paths. The system is implemented as a web-based application using modern technologies, enabling real-time interaction between the user and the system. By combining data retrieval techniques with intelligent response generation, the proposed solution improves accuracy, reduces misinformation, and enhances user experience. This approach simplifies the decision-making process and helps students choose suitable academic paths with confidence.

“High Risk And Low Risk Patients’ Prediction In Icu Using Ml Algorithms”

Authors: B. M. Promod Kumar, Namith Kumar Y, Pruthvi M C, Poorvik K V, Jagadeesh M

Abstract: This concept is based on patient’s classification in an Emergency Department in a hospital according to their critical conditions. Machine learning can be applied based on the patient’s condition to quickly determine if the patient requires urgent medical intervention from the clinicians or not [1]. Basic vital signs like Systolic Blood Pressure (SBP), Diastolic Blood Pressure (DBP), and Respiratory Rate (RR), Oxygen saturation (SPO2), Random Blood Sugar (RBS), Temperature, and Pulse Rate (PR) are used as the input for the patients’ risk level identification [2]. High-risk or non-risk categories are the outcome for patient classification. ML algorithms such as Gaussian NB, KNN or DT are applied for the data analysis and for the classification. We'll use a many of supervised learning methods before deciding which one is best for the model. Existing systems rely on classical learning models, which are inefficient and imprecise. They aren't as accurate as the proposed model and take a little longer to process. Many existing topics on patient’s classification where they have built models and shown results generated using R language, Python language and data science tools. All existing works are just models, cannot be applied as application useful in real time. In our project work we build an application with ML models that can classify high risk patients and non-risks patients in an emergency department and provides doctors with the information of how to handle patients and treat better [5]. Our proposed work is a real-world medical system useful for hospitals and doctors and built using trending tools such as Visual Studio code, PYTHON and MYSQL Server.

Data Driven Housing Intelligence: A Comprehensive Analysis Of Infrastructure,Contractor And Community

Authors: Burhan Sheikh, Vighnesh Muppawar, Jiya Khan Pathan, Suyog Madavi, Prof. Sachin Dhawas

Abstract: CompThe Pradhan Mantri Awas Yojana (PMAY) is a flagship initiative by the Government of India aimed at providing affordable housing to the rural and urban poor. However, at the local administrative level (Zila Parishad), the management of beneficiary data involves processing massive, decentralized Excel datasets. This manual approach leads to data redundancy, lack of real-time monitoring, and significant difficulty in identifying stalled projects. This paper proposes "Data Driven Housing Intelligence," a web-based analytical application designed to automate and centralize PMAY data management. Developed using the Python Flask framework, the system integrates bulk data ingestion, automated data cleaning using the Pandas library, and persistent storage in a structured SQLite database. It features a multi-level drill-down dashboard with interactive visualizations (Bar and Pie charts) for block-wise analysis. Furthermore, it introduces a "Visual Audit" mechanism allowing stage-wise photo uploads for transparency and an automated algorithm to flag delayed projects. The implementation results demonstrate a significant reduction in administrative overhead and improved data accessibility for government officials.

Smart Door: Motion,face-recognition And Voice Recognition IOT Security

Authors: Mr.B.Ajantha Reddy, Mr.K.Ch.Malla Reddy, Immadi Venkata Naga Sai Pujitha, Byragani Manasa, Bathula Varshini, Shaik Fathima

Abstract: The purpose of this project is to make home or office or any area secure. When someone presses the doorbell, then the doorbell makes a video call to the registered number. If someone roams in front of the door it notifies you by sending message. Then he can see the person who is roaming in front of our door. So, if the person is known we can open the door otherwise we can be alert. And also, we can talk to the person through mobile only and the person can reply there itself, because it contains the audio speaker so that we can hear the outside people talks trough the mobile once we pick up the video call. If someone tries to steal it then the steal alarm will be activated.

DOI: http://doi.org/10.5281/zenodo.20394540

Automated Patient Health Tracking System

Authors: Dr.A.Ranganayakulu, Dr.D.Satyanarayana, Allam Lakshman, Tadikamalla Madhav Kumar, Pasala Harikrishna, Sandala Dinesh Kumar

Abstract: Immediate diagnosis and management are crucial in preventing serious consequences for older and mobility-impaired patients, since falls are a leading cause of injury in these populations. Using Arduino and Bluetooth, this project aims to create a patient fall monitoring system that can detect falls in real-time and notify caregivers immediately. A microelectromechanical system (MEMS) sensor detects abrupt and uncontrollable falls by continuously monitoring the patient's movement and assessing changes in posture, speed, and acceleration. By analyzing sensor data, a fall detection system that is based on thresholds may distinguish between normal activities and actual falls. The technology immediately notifies registered mobile devices over Bluetooth and activates a buzzer alarm for local alerting the moment a fall is detected. A wide variety of fall occurrences, such as sliding, tripping, and abrupt loss of balance, were detected with excellent sensitivity in the experiments. It was determined that the system reliably and without noticeable delay sent emergency alerts regardless of ambient conditions like vibrations and the range restrictions of Bluetooth. The open-source, ever-improving architecture of Arduino makes it possible to include Wi-Fi or cellular connection for expanded coverage in future expansions. The system's ability to monitor, notify, and intervene in real-time enhances patient safety.

DOI: http://doi.org/10.5281/zenodo.20394577

Future Ready Low-power 4-bit Multiplier For Portable VLSI Systems

Authors: Mr.K.Ch.Malla Reddy, Mr.M.Ramana Reddy, Muthakapalli Sai Sravani, Tadi Kranmai, Chinamanagonda Ranga Gayathri, Kothakota Vara Lakshmi

Abstract: A multiplier is an essential part of many extremely large scale integration systems and finds extensive application in digital circuits for innumerable arithmetic calculations. One of the most fundamental operations in digital technology is multiplication, and hardware multipliers are essential for quick computing and efficient data processing. There have been several design approaches to multipliers that focus on surface area and energy efficiency in response to the desire for high-performance, low-power multipliers. Our study presents a 4-bit field multiplier with improved energy efficiency compared to conventional designs, using a modified gate-diffusion input (MGDI) cell architecture. The energy consumption of the MGDI-based multiplier is significantly reduced to 1.109861 mW without sacrificing any of the essential operating efficiencies. Thorough simulations have been conducted to showcase the performance of the 4-bit MGDI multiplier, which was meticulously constructed utilizing Tanner EDA tools. Low power, 4-bit multiplier, time delay, transistor count, modified gate diffusion input (MGDI) are some of the keywords.

DOI: http://doi.org/10.5281/zenodo.20394671

Efficient MAC Architecture Using Different Parallel Adders

Authors: Mr.A.Prasad, Mrs.N.Swarupa Rani, Batchu Bala Bhargavi, Chinni Yamini, Chityala Lakshmi Devi, Chilakala Anjali

Abstract: Because of its capability to do arithmetic operations at fast speeds, the Multiply-Accumulate (MAC) Unit is an essential part of all digital signal processor applications. An 8-bit MAC Unit that can do addition and multiplication is the target of this study. While the MAC Unit uses the same multiplier, it incorporates other adders, including the Kogge-Stone, Ladner-Fischer, Carry Look-Ahead, and Ripple Carry adders. Xilinx ISE was used to implement the structures that were created in Verilog Hardware Description Language (HDL), and ModelSim was used for simulation.

DOI: http://doi.org/10.5281/zenodo.20394713

Food Delivery Website Using Tsp

Authors: Prof. Mahesh Dumbere, Vivek Thakre, Pratik Patil, Ritesh Magare, Mahesh Gajbhiye

Abstract: The demand for effective delivery route optimization has grown due to the quick expansion of online meal delivery services. In order to reduce delivery time and expense, this study suggests a web-based meal delivery system that makes use of the Travelling Salesman Problem (TSP). For a delivery agent to visit several clients and return to the starting point, the system calculates the quickest path. To solve TSP, a variety of algorithms are examined, including Nearest Neighbor, Genetic Algorithm, and Dynamic Programming. The suggested solution increases customer satisfaction, lowers fuel consumption, and improves delivery efficiency. According to experimental results, optimized routing performs noticeably better than conventional delivery techniques.

Intellihome: FPGA-Driven Smart Automation System

Authors: Dr.P.Prasanna Murali Krishna, Mrs.N.Swarupa Rani, Mudamala Manoj, Dudekula Mastan Vali, Patibandla Jayanth, Baddigam Koti Reddy

Abstract: The Xilinx Zynq-7000 system on a chip (SoC) is the basis for the home automation system described in this article. The FSM logic processes the signals from the sensors (the fire sensor requires 5V digital and the buzzer 5-12V) using Verilog HDL. The RTL schematics and waveforms were validated using Cadence tools. The three-second detection threshold for fire and intruder alarms is one of the most important features, along with automatic lighting and temperature adjustment. The scalability of the modular system allows for the easy integration of more devices, which in turn increases functionality and security.

DOI: http://doi.org/10.5281/zenodo.20394814

Enhanced Power Distribution Through IOT Based Under Ground Cabel Fault Detection

Authors: Mr.N.B.Jilani, Mr.M.Ramana Reddy, Dudekula Chinna Khasim Vali, Nukala Sai Kumar, Shaik Mahaboob Subhani, Pengani Ganesh, Uppaladinne Ram Charan

Abstract: Accurately locating faults in subterranean cable connections is a prevalent challenge, particularly in cities, and the suggested approach aims to tackle this problem. A power supply and an Arduino microcontroller kit form the basis of the system, which measures the wire length in kilometers using current measurement circuits linked to the microcontroller's Analog to Digital Convertors (ADC) device. A relay circuit is used to regulate the relays, and switches are employed to simulate faults. You can see the specifics of the issue as they happen on a 16×2 LCD screen. Short circuit problems in the underground cable may be caused by environmental factors such as rain, subterranean pollution, drain leakage, and so on. Among the many possible problems with subterranean cable, the identification of short circuit faults is the primary emphasis of this work. By monitoring voltage variations across the resistor and calculating the distance from the source feed point, the ADC digitizes the data and displays it on the LCD, pinpointing the exact position of the short circuit defect. The Wi-Fi module 8266 helps to save the specifics of the defect location in the cloud, allowing for subsequent investigation. Because of this, the suggested design for the Arduino microcontroller can pinpoint the precise location of the problem in terms of kilometers from the base station. Additionally, in the event of a malfunction, a buzzer alerts field personnel to the urgency of the situation. This innovative approach provides a dependable and efficient means of locating and identifying issues with underground cable cables. This reduces downtime and facilitates service.

DOI: http://doi.org/10.5281/zenodo.20394874

Maximizing Area And Power Efficiency With A Modified Karatsuba Multiplier For Cryptography Algorithms That Avoid Errors

Authors: Dr. A. Ranganayakulu, Mr. A. Prasad, M. Ramana Reddy, B. Ajanta Reddy, Dr. D. Satya Narayana

Abstract: Using efficient finite field multipliers becomes vital in elliptic curve cryptography (ECC), where data security and authentication are critical. These multipliers do affect performance, however, because they use quite a lot of hardware resources. The Karatsuba algorithm and its variations are explored in this study as a means to enhance hardware efficiency on FPGA devices. Although performance is improved with the overlap-free Karatsuba algorithm. Problems with recombining intermediate findings cause them to add 20% mistakes. We present a modified Karatsuba method that can compute four key outputs for 2-bit inputs error-free to solve this problem. The revised design tested on Artix-7 FPGA and implemented in Verilog HDL, cuts power consumption by 73.95% and area utilization by 95% when compared to the original Karatsuba algorithm. Its overall efficiency is much improved while accuracy is guaranteed, despite a 3.22% increase in area and an 11.12% increase in power compared to the overlap-free version.

DOI: http://doi.org/10.5281/zenodo.20394919

High-performance FPGA- Based ALU Using Reversible And Quantum-inspired Logic

Authors: Mr. B. Ajantha Reddy, Mr. A. Prasad, Dr. A. Ranganayakulu, Pagadala Ananthalakshmi, Jestadi Divya Jyothika, Vyja Swetha, Bathula Persis

Abstract: Computing devices, mobile phones, and computers all rely on the Arithmetic Logic Unit (ALU), a critical subsystem inside processors, to carry out the arithmetic and logical operations necessary for digital system operations. There is an urgent need for more energy-efficient alternatives in digital system design to standard ALUs designed using non-reversible logic gates, which are known for their considerable power consumption. Our proposed solution to this problem is a 32-bit ALU that makes use of reversible logic gates; this will allow us to cut down on power consumption while simultaneously increasing computing performance. Our suggested 32-bit ALU uses reversible logic gates to provide a solution that reduces power consumption and increases computational efficiency; this should lead to a revolution in digital system design. Not only do we want to reduce power consumption, but we also want to increase the ALU's usefulness and adaptability in other computing contexts by adding a full set of sixteen separate operations. Our goal is to set a new standard for ALU design with this novel method, one that puts computing power and power efficiency first. Our 32-bit ALU's use of reversible logic gates is a giant leap forward in the quest for power-efficient digital computers that don't sacrifice processing capability. We seek to solve the essential demand for energy-efficient solutions in today's technology-driven world by contributing to the progress of digital system design with painstaking attention to detail and a focus on innovation. Metaphors: Arithmetic Logic Unit (ALU), Reversible Logic Gates, Digital System Design, Energy Efficiency, Versatility, Innovative Architecture, Low Power Consumption, Computational Performance.

DOI: http://doi.org/10.5281/zenodo.20394930

Comparative Analysis Of Pixel-Based Segmentation Model For Accurate Detection Of Impacted Teeth

Authors: Dr. Deepika, Sneha K M, Sudhanshu Sharma, Sujendra T R, Varshini J

Abstract: Impacted teeth, particularly third molars that fail to erupt properly due to insufficient space or improper angulation, represent a common dental condition that can lead to severe complications including infection, cyst formation, and damage to adjacent structures. Traditional diagnosis relies heavily on manual interpretation of panoramic dental X-ray images by clinicians, a process that is time-consuming, subject to human variability, and lacks pixel-level precision. This paper presents an AI-based impacted tooth detection system using the U-Net deep learning architecture, a convolutional neural network specifically designed for biomedical image segmentation. The proposed system performs pixel-level segmentation of impacted tooth regions from panoramic dental X-ray images, providing precise boundary delineation that conventional object detection methods cannot achieve. The system integrates data annotation, model training using PyTorch, and deployment via a Flask-based web application into a unified end-to-end pipeline. Preprocessing steps including grayscale conversion, resizing to 256 × 256 pixels, and pixel normalization ensure consistent input quality. The trained model achieved an overall segmentation accuracy of approximately 87%, with precision of 85%, recall of 89%, and an F1-score of 87%. Experimental results and confusion matrix analysis confirm that the proposed system reliably detects impacted tooth regions while maintaining a low rate of false predictions. The system demonstrates strong real-time performance through a user-friendly web interface, making it a practical diagnostic support tool for dental professionals.

DOI: http://doi.org/10.5281/zenodo.20395549

Real-Time Collaborative Code Editor Using WebSockets

Authors: Pruthviraj Pawar, Niranjan Rasal, Pruthviraj Deshmukh, S. B. Dighe

Abstract: Collaborative programming platforms are becoming increasingly important in modern software development, online education, and distributed teamwork environments. Traditional methods such as manual file sharing, screen sharing, or repeated version control synchronization are often inefficient during live coding sessions. This paper presents the design and implementa- tion of a Real-Time Collaborative Code Editor developed using WebSockets and Socket.IO. The proposed system allows multiple users to edit source code simultaneously through a browser-based interface with minimal synchronization delay. The frontend is developed using HTML, CSS, JavaScript, and Monaco Editor, while the back- end is implemented using Node.js and Express.js. Socket.IO is used to establish persistent bidirectional communication between connected clients and the server. Experimental observations demonstrate synchronization la- tency below 100 milliseconds under normal conditions. Perfor- mance analysis confirms that WebSocket-based communication provides significantly lower delay and better bandwidth efficiency than traditional HTTP polling techniques.

Towards Sustainable Cloud Computing: Limitations Of Revenue-Optimized Resource Scheduling Policies

Authors: Dr. Megala.R, Dr.D. Balasubramanian

Abstract: Cloud computing has become an essential platform for delivering scalable and on-demand computational services. However, most existing cloud resource scheduling policies are designed primarily to maximize provider revenue and infrastructure utilization, often neglecting sustainability concerns such as energy efficiency, carbon reduction, thermal management, and fair resource allocation. This paper critically examines the limitations of revenue-optimized resource scheduling policies in cloud computing environments. The study evaluates traditional scheduling techniques including First-Come-First-Serve (FCFS), Round Robin, Priority Scheduling, and profit-aware heuristic scheduling approaches. Major issues identified include excessive power consumption, increased carbon emissions, resource starvation, thermal imbalance, and reduced long-term infrastructure sustainability. To address these limitations, a Sustainable Multi-Objective Scheduling Framework (SMOSF) is proposed that integrates energy efficiency, QoS maintenance, fairness, and profitability objectives. Comparative analysis demonstrates that sustainability-aware scheduling policies can significantly reduce energy consumption and environmental impact while maintaining acceptable service quality and operational profitability. The proposed framework contributes to the advancement of green cloud computing and sustainable data center management.

DOI: http://doi.org/10.5281/zenodo.20404448

Smartphone Remote Detection

Authors: Arya Parashram Kamble, Akshata Hemant Bansode, Vaishnavi Vitthal Shinde, Mrs D.N Ghatage

Abstract: Smartphones have become an essential part of daily life, storing sensitive personal information and enabling communication, navigation, and financial transactions. However, increasing smartphone usage has also led to serious concerns such as device theft, loss, and personal safety risks. Existing applications provide limited functionality and fail to offer a complete solution for real-time monitoring and emergency response. This paper presents a Smartphone Remote Detection System (Safety Guard) that provides an integrated platform for device tracking, remote control, and emergency assistance. The system is developed using Flutter and Firebase technologies, enabling real-time data synchronization, push notifications, and background processing. The application allows users to track device location, send remote commands such as ring, vibrate, and lock, and trigger SOS alerts to notify supporters and authorities.The system ensures continuous operation even when running in the background, making it reliable in critical situations. Additionally, it includes a complaint management module to report incidents effectively. The proposed solution enhances both personal safety and device security by providing a fast, efficient, and user-friendly system for real-time monitoring and emergency response.

Iot Based Coal Mine Safety Monitoring And Alerting System

Authors: Hari Priya S, Mrudhul NR, Saran H, Subarssini V, Vishal K

Abstract: Coal mining is one of the most hazardous industries, with risks such as gas leaks, rising underground temperatures, flooding, poor air quality, and structural hazards. This paper presents an IoT-based Coal Mine Safety Monitoring and Alerting System that continuously monitors environmental conditions and provides real-time alerts to miners and supervisors. The system employs sensors to detect gases (MQ-4, MQ-7, MQ-135), temperature and humidity (DHT11, DHT22), water levels (float switch, conductivity sensor) all integrated with an ESP32 microcontroller. Sensor data is transmitted via LoRa communication modules to a central monitoring unit, where it is analysed for any signs of danger. When hazardous conditions are detected, alerts are sent instantly through visual, auditory, or wireless notifications. Compared to conventional wired systems, this IoT-based approach offers greater reliability, faster response times, and reduced costs. By enabling proactive safety management, the system helps prevent accidents and injuries while improving operational oversight, ultimately supporting safer and more sustainable mining practices.

Automated Parallel Hybrid Data Extraction and Entity Resolution for Sports Data Aggregation: Architecture, Challenges, and Trade-offs

Authors: Olalekan Oluyinka

Abstract: This paper presents the design and implementation of an architecture for automated hybrid data extraction, integration, and entity resolution for sports aggregation. The system consolidates inconsistent records from multiple heterogeneous sources into a centralized, deduplicated interface for sports event discovery and streaming access. Data is extracted in real time across seven heterogeneous sources and directly ingested in the automated pipeline. A multi-step entity resolution algorithm, combined with data pre-processing within a Single Source of Truth (SSOT) framework transforms heterogeneous data into a unified, deduplicated index. The architecture employs edge caching and batching to reduce latency and improve operational performance in constrained environments. A prototype further demonstrates the practicality of automated multi-source sports event aggregation through entity resolution.

DOI: https://doi.org/10.5281/zenodo.20406780

Compensation and Reward Management Practices at Wipro: A Secondary Research Analysis

Authors: Prachi Saini, Dr. Pooja Kohli

Abstract: This research paper analyzes the compensation and reward management practices of Wipro Limited, a leading multinational information technology company in India. The study is based on secondary data collected from academic journals, industry reports, company publications, and previously published research. It focuses on understanding how Wipro structures its compensation system, including salary components, performance-based incentives, employee benefits, and recognition programs. The paper is supported by key motivational theories such as Maslow’s Hierarchy of Needs, Herzberg’s Two-Factor Theory, and Adams’ Equity Theory. These theories help explain how different types of rewards influence employee motivation, satisfaction, and performance. The findings suggest that Wipro adopts a comprehensive total rewards approach that combines fixed pay, variable pay linked to performance, and a wide range of financial and non-financial benefits. In addition, the company uses formal recognition programs to appreciate employee contributions and encourage high performance. However, the study also identifies certain challenges in Wipro’s compensation system. Issues such as lack of transparency in pay communication, limited personalization of non-monetary rewards, and gaps in career growth opportunities may affect employee satisfaction. The paper suggests that improving communication, offering more customized rewards, and strengthening career development frameworks can help Wipro enhance employee engagement and reduce attrition. Overall, the research highlights the importance of a balanced and transparent reward system in achieving organizational success.

DOI: https://doi.org/10.5281/zenodo.20407064

Design And Implementation Of A Wearable Multimodal Hand Gesture Vocalizer For Assistive Communication

Authors: Aryan Patel, Era Mane, Tanvi Sonawane, Dr. Vineeta Philip

Abstract: The persons who have difficulties communicating with society because of hearing and speech problems face challenges that restrict their communication ability with society. While sign language is one possible solution, everyone may not comprehend it. The design and application of a low-cost wearable hand gesture vocalizer that produces both visual and aural outputs from predefined hand gestures are presented in this paper. The suggested system uses flex sensors built into a glove to record finger movements, which are then operated by an Arduino Nano microcontroller. An external audio replay module is used to translate recognized gestures into corresponding audio output and display them on a 16×2 I2C-based LCD at the same time. To improve approachability for users with hearing or vision impairments, the system places a strong emphasis on multimodal feedback, appropriate wearable design and un- complicated hardware. The prototype developed shows reliable gesture identification with minimal latency and offers a extensible platform for future improvements.

DOI: http://doi.org/10.5281/zenodo.20407122

Smart Navigation Stick For Blinds

Authors: Dr.Vineeta Philip, Manasi Owhal, Piyush Khonde, Soham Khulage

Abstract: Smart Navigation Stick for the Blind is an aid that helps blind or visually impaired people move independently and safely. The system consists of several sensors and communication modules interfaced to an Arduino Uno (SMD) controller for real- time obstacle detection, hazard alerting and emergency mobil- isation purposes. It uses ultrasonic sensors to detect potential obstructions at various distances (i.e., pointing in all directions) and offers timely feedback, aided by the vibration motor and buzzer. Adding a flame sensor is responsible for identifying fire or high-temperature threat, and the inclusion of a level sensor help with water-filled or unlevel surfaces. The user can activate emergency functions when needed using a manual switch. The built-in GPS and GSM modules allow tracking of individual locations in emergency scenarios, as well as auto-issue alert texts to pre-selected people. This is a compact, low-cost energy-efficient and easy-to-use system that can be used every day outdoors or indoors. All in all, the smart navigation stick seeks to elevate situational awareness, safety, and confidence of visually impaired users with intelligent sensing as well as feedback initialized in real time.

DOI: http://doi.org/10.5281/zenodo.20407703

Online Coffee Shop Management System

Authors: Vaibhav Mali, Vaibhav Mane, Prajwal Zanje, Prof. N.B.Khade

Abstract: Digital image classification plays a significant role in the early detection and analysis of medical conditions. Traditionally, diagnosis is performed manually by ophthalmologists through examination of retinal fundus images. However, this process is time-consuming, requires expert knowledge, and may sometimes lead to errors due to human limitations. In contrast, automated digital image classification systems provide a faster, more consistent, and cost-effective solution by analyzing large volumes of medical images efficiently. This work focuses on the application of digital image classification techniques for identifying different stages of diabetic retinopathy. Additionally, different image preprocessing, feature extraction, and classification methods are discussed. The study also summarizes the key image features commonly used in previous research for accurate classification of retinal images into different disease categories.

NexusOps: A Secure Agentless Framework For Real-Time Telemetry And Automated Self-Healing In Multi-Server Infrastructure

Authors: Pranit Dattatraya Patil, Avdhoot Arunkumar Sakate, Ajinkya Anil Dhane, Prof. Uchale B. S

Abstract: As cloud-native environments scale, maintaining the high availability of virtual private servers (VPS) has become paramount. Traditional server monitoring tools rely heavily on daemon agents installed on host machines, exposing host environments to resource taxations and security vulnerabilities. This paper presents NexusOps, a premium agentless server management and self-healing platform. By utilizing Java Secure Channel (JSch) tunnels directly to host Operating Systems, NexusOps extracts real-time telemetry metrics (CPU, RAM, Disk, active processes) without target-side exporters. Telemetry metrics are streamed through a centralized Spring Boot REST and WebSocket engine into a responsive React frontend interface, facilitating live command execution, remote service control, and visual analytics. Furthermore, NexusOps introduces a mathemat- ical health-score heuristic model and a multithreaded automated self-healing controller to autonomously resolve critical errors (e.g., service failures, storage spikes) and dispatch alerts via external push notification channels. Our empirical evaluation demonstrates that NexusOps achieves equivalent telemetry ac- curacy and latency (sub-100ms response times) as traditional systems while eliminating persistent CPU and memory footprints on target nodes.

IoT-Driven Demand Forecasting Integrated With Blockchain-Enabled Resilient Supply Chain Model And Disruption Mitigation

Authors: Dr. Srimathi Kannan

Abstract: The global supply chain network is currently more vulnerable to disruptions that can be caused by pandemics, geopolitics, and other reasons. In such cases, centralized and opaque logistics infrastructures are exposed to risks. This study recommends a reliable supply chain management framework that utilizes blockchain technology, IoT sensors, a hybrid deep learning algorithm to forecast consumer demand, and disruption management through smart contracts. The proposed architecture relies on the Hyperledger Fabric, a permissioned blockchain network, and guarantees data immutability and transparency. A temporal convolutional network with an attention mechanism enables forecasting demand at 95.2% accuracy over a 12-week time frame. After detecting a disruption, the automated smart contract system will engage in dynamic routing, inventory redistribution, and supplier substitution. Simulating the proposed solution on a multi-tier supply chain network with over 100 nodes resulted in 67% faster disruption resolution compared to conventional models and 94% customer satisfaction during disruption events, while conventional models were able to serve just 62% of consumers.

DOI: https://doi.org/10.5281/zenodo.20411670

Neuromarketing Signals And Consumer Purchase Intent Prediction Using EEG And Computer Vision

Authors: Dr. A. Sathiya, Dr. P. Jeyanthi

Abstract: Purchase prediction and understanding its nuances are essential aspects of marketing, but standard approaches do not provide any information about subliminal neural processes which lead to actual purchases. In this paper, a novel multimodal system is proposed which combines EEG neuromarketing data and computer vision features related to visual attention to achieve accurate prediction of consumer purchase intent. A dataset comprising 120 participants who viewed 500 e-commerce images is used for extraction of both EEG-based features (frontal asymmetry of alpha activity, theta/beta ratio, and late positive potential) and visual attention features based on computer vision approach (fixations density, saccades dynamics, and pupils size). Hybrid model consisting of two branches – Temporal Convolutional Network for processing EEG signals and Graph Attention Network for mapping visual attention – reaches 88.3% accuracy and an area under curve equal to 0.94 in predicting consumer purchase intent, while unimodal EEG and visual models reach 74.2% and 72.8% respectively.

DOI: https://doi.org/10.5281/zenodo.20411755

AI-Driven Competency Mapping Framework For Future-Ready Talent Development

Authors: Mr. Shrikant Karampuri, Dr. P. Jeyanthi

Abstract: The fast development of technologies, automation, and digitalization processes have led to a considerable disruption of how workforce planning is traditionally performed and have created a significant gap between the skills of the existing workforce and the skills that will be necessary in the future for the success of organizations. The process of competency mapping which includes the identification, evaluation, and alignment of skills with strategic goals is extremely important to ensure future readiness via skill development. This paper introduces a new approach to the competency mapping using artificial intelligence, which utilizes NLP algorithms for skill extraction from unstructured sources (resumes, job descriptions, and performance evaluations), GNNs for skill adjacency and competency modeling, and BKT for prediction of the evolution of individual skills. When applied to a database of 50,000 employees in a multinational technology company, our approach yields an accuracy of 89.7% for skill extraction, 82% for skill adjacencies, and 76% for the prediction of future skill gaps. The proposed approach allows us to create personalized learning paths and reduce time-to-competency by 34% in six months.

DOI: https://doi.org/10.5281/zenodo.20411976

A Study on the Impact of Upi Usage on Digital Payment Preferences in India

Authors: Dr. Ashish Saxena, Priya Kumari

Abstract: Indias financial world has changed a lot because of technology and digitalization. Now digital payments are a part of our daily transactions. They are fast, easy, convenient and safe. The Unified Payments Interface or UPI is a leading platform for payments. It was developed by the National Payments Corporation of India. UPI helps people transfer money instantly using their phones. They do not need to share their bank details for each transaction. This is possible because many people have smartphones, internet access and use banking. This study looks at how UPI affects users, businesses and Indias digital payment system. It explores how people are changing their behavior to use transactions more. UPI plays a role in helping India become a cashless economy. We look at how UPI simplifies money transfers, bill payments, online shopping and transactions with merchants. We also check how satisfied customers are with UPI, how easy it's to use how fast it is, how secure it is and how reliable it is. The research uses data from questionnaires and surveys well as information from other sources like journals, government reports and websites. We use methods to understand this data and see how UPI affects financial transactions. We found that UPI makes payments more efficient and convenient. It helps reduce the use of cash and increases the use of services. The reasons for this are that UPI is accessible, cost, fast and has many apps. However, there are still some challenges. These include cyber threats, connectivity problems, technical issues and a lack of knowledge. In conclusion UPI has changed Indias payment systems for the better. It has helped include people in the financial system and created opportunities, for businesses, consumers and institutions. As UPI continues to grow it promises to create a secure and cashless financial system that supports a digitally empowered economy.

DOI: http://doi.org/10.5281/zenodo.20412253

Federated Learning With Privacy Preservation For Healthcare Analytics

Authors: S Jayashree Ananth, Naveen V S

Abstract: The digitization of the healthcare industry has resulted in massive collection of personal health information among hospitals, clinics, and research institutions. But strict privacy laws (HIPAA, GDPR), along with other institutional obstacles, hinder data collection in a centralized manner, resulting in data silos that prevent the construction of efficient machine learning models for predicting diseases, estimating treatment, and managing public health issues. In this paper, we introduce a framework for privacy-preserving federated learning (PPFL) in healthcare. Our proposed framework includes three techniques: (1) Federated Averaging with differential privacy (DP-FedAvg) for model privacy, (2) Secure Multi-Party Computation (SMPC) for private aggregation of gradients, and (3) Homomorphic Encryption (HE) for performing computations on encrypted data. Our PPFL framework is evaluated on three real-life datasets of healthcare applications (mortality prediction from ICU records, diabetic retinopathy classification, and diagnosing COVID-19 patients) and outperforms federated learning with centralization in terms of model accuracy (within 3.2%) and provides differential privacy guarantees with ε=1.0 and δ=10⁻⁵.

DOI: https://doi.org/10.5281/zenodo.20415139

Sentiment Classification of Imdb Movie Reviews Using Naturl Language Processing Techniques

Authors: P. Anusha, E. Naveen Kumar, G. Sravanthi, E. Rohitha

Abstract: Sentiment analysis is a crucial task in natural language processing (NLP) that aims to determine the overall sentiment or opinion expressed by a reviewer towards a movie. This study focuses on the sentiment analysis of IMDB movie reviews using various machine learning and NLP techniques. The findings indicate that feature selection can enhance the accuracy of sentiment-based classification, but the effectiveness depends on the specific method and number of features selected. The paper also presents a comprehensive comparison of traditional machine learning techniques and advanced transformer-based models for sentiment analysis of IMDB movie reviews. The results provide insights into choosing appropriate methods for accurate and timely sentiment analysis on IMDB data. The study employs feature extraction techniques such as bag-of-words (BoW), term frequency-inverse document frequency (TF-IDF), and word2vec. Feature selection using methods like chi-square is shown to improve classification performance.

DOI: http://doi.org/10.5281/zenodo.20423234

Intelligent Flight Delay Prediction Using Machine Learning

Authors: P. Anusha, Syed Mannan Uddin, T.Sree Chandana, V.Vaishnavi

Abstract: Accurate flight delay prediction is fundamental to establish the more efficient airline business. Recent studies have been focused on applying machine learning methods to predict the flight delay. Most of the previous prediction methods are conducted in a single route or airport. This paper explores a broader scope of factors which may potentially influence the flight delay, and compares several machine learning-based models in designed generalized flight delay prediction tasks. To build a dataset for the proposed scheme, automatic dependent surveillance broadcast (ADS-B) messages are received, pre-processed, and integrated with other information such as weather condition, flight schedule, and airport information. The designed prediction tasks contain different classification tasks and a regression task. Experimental results show that long short-term memory (LSTM) is capable of handling the obtained aviation sequence data, but overfitting problem occurs in our limited dataset. Compared with the previous schemes, the proposed random forest-based model can obtain higher prediction accuracy (90.2% for the binary classification) and can overcome the overfitting problem.

DOI: http://doi.org/10.5281/zenodo.20423275

Dna Sequence Predictions Using Nlp And Ml

Authors: K. Vigneshwar, P. Shruthi, J. Rahul Naik, P. Khaleel Basha

Abstract: Deoxyribonucleic acid (DNA) is a biological macromolecule. Its main function is information storage. At present, the advancement of sequencing technology had caused DNA sequence data to grow at an explosive rate, which has also pushed the study of DNA sequences in the wave of big data. Moreover, machine learning is a powerful technique for analyzing largescale data and learns spontaneously to gain knowledge. It has been widely used in DNA sequence data analysis and obtained a lot of research achievements. Firstly, the review introduces the development process of sequencing technology, expounds on the concept of DNA sequence data structure and sequence similarity. Then we analyze the basic process of data mining, summary several major machine learning algorithms like Multinomial NB Classifier & Random Forest, and put forward the challenges faced by machine learning algorithms in the mining of biological sequence data and possible solutions in the future. Then we review four typical applications of machine learning in DNA sequence data: DNA sequence alignment, DNA sequence classification, DNA sequence clustering, and DNA pattern mining. We analyze their corresponding biological application background and significance, and systematically summarized the development and potential problems in the field of DNA sequence data mining using Multinomial NB Classifier & Random Forest. Finally, we summarize the content of the review and look into the future of some research directions for the next step.

DOI: https://doi.org/10.5281/zenodo.20425270