Authors: Manogna.Velamakanni, DurgaBhavani.Makkena, Sindhu.Merugu, Harshika.Lekkala, Indira Lakshmi Borigorla

Abstract: Agriculture is very important for food security, but over 40% of farmers use too little or too much fertilizer, which causes low crop yield, soil damage, and financial loss. Smart recommendation systems can help farmers by giving accurate advice on the right type and amount of fertilizer. Many machine learning methods like Decision Trees, Random Forest, Gradient Boosting, and Neural Networks have been used for crop and fertilizer recommendations. However, these methods often need a lot of computing power, do not work well with small or noisy data, and can be hard for farmers to understand. To solve these problems, we propose a Support Vector Machine (SVM)-based fertilizer recommendation model. SVM works well with small and unbalanced datasets, reduces overfitting, and handles complex patterns while needing fewer resources, making it suitable for real farming. Using soil nutrient values and crop needs, the model gives reliable predictions. Tests show that the SVM model achieves 96.77% accuracy, making it effective for smart agriculture and proper fertilizer use.

DOI: https://doi.org/10.5281/zenodo.20066724

Published by: vikaspatanker

Authors: Sk.Firdaus Fathima, M.Anusha, S.Vennela, Sk.Sadiya Kousar, B.Gayathri

Abstract: Driver drowsiness is one of the major causes of road accidents globally, leading to serious injuries, deaths, and economic loss. To combat this, a real-time Driver Drowsiness Detection System has been implemented using machine learning algorithms combined with IoT hardware. The system tracks the driver's eyes continuously through a real-time video feed obtained via a webcam. With the OpenCV and dlib libraries, the Eye Aspect Ratio (EAR) is computed to obtain a measurement of the degree of eye closure, which is a good predictor of drowsiness. Upon detection of prolonged eye closure, the system sends a serial communication command to an Arduino Uno microcontroller to activate a buzzer alarm and commence a progressive motor deceleration, mimicking a safe vehicle stop. This two-stage mechanism reduces the risk of accidents by giving both an initial warning and an automatic safety measure. Experimental data show that the designed system has a detection accuracy of 96.8% for different illumination conditions, with a response time of less than one second. The approach is cost-effective, non-invasive, and easily implementable on contemporary vehicles, ensuring it to be a promising solution for improving road safety.

DOI: https://doi.org/10.5281/zenodo.20066634

</p

Published by: vikaspatanker

Authors: V Manogna, B Durga, P Sravani, N Yamuna, B Reshma

Abstract: Alzheimer’s disease (AD) is a progressive brain disorder that leads to memory loss and a gradual decline in thinking and reasoning abilities. One of the major challenges in dealing with Alzheimer’s is detecting it early and accurately using MRI brain scans. Traditional manual analysis of these scans can be slow, complex, and prone to human mistakes over the years, different machine learning (ML) models like Decision Tree, Random Forest, Logistic Regression, and K-Nearest Neighbors have been used to identify Alzheimer’s. However, these models often face issues such as overfitting, lower accuracy, and weak performance when dealing with complex and high-dimensional MRI data.to overcome these limitations, the proposed approach uses SVM model for detecting and classifying Alzheimer’s disease. The SVM model is well-suited for handling non-linear and complex data. It can effectively separate different disease categories by using advanced kernel functions and optimal hyperplane techniques. This leads to more precise and stable classification results, even with smaller datasets compared to existing ML models, the proposed SVM model achieves higher accuracy, sensitivity, and specificity, making it more dependable for automatic Alzheimer’s detection. It not only reduces errors but also helps in identifying the disease at an early stage, which is crucial for better treatment and patient care. With 98.16% classification accuracy, it outperforms current architectures significantly in the Alzheimer Detection and Classification Using SVM.

DOI: https://doi.org/10.5281/zenodo.20065684

Published by: vikaspatanker

Authors: D Anveshini, K Chinna Reddemma, M Venkata Akshara, Ch Sai Samanvitha, Ch Sai sirisha

Abstract: Sudden cardiac arrest is a critical medical emergency that demands immediate recognition and timely intervention to improve a patient’s chances of survival. Although hospital-based cardiac monitoring systems are dependable, their high expense and lack of portability make them impractical for everyday personal monitoring. This work presents an Internet of Things (IoT)-based wearable framework that enables continuous and real-time cardiac health observation in non-clinical environments. The device integrates multiple biosensors including electrocardiogram (ECG), pulse oximeter, body temperature, and galvanic skin response (GSR) to acquire physiological data. The acquired data are uploaded to a cloud environment, where algorithms evaluate and categorize the user’s cardiac condition as normal, borderline, or severe. The system is linked to a companion mobile application that visualizes real-time readings and automatically issues alerts to caregivers and medical professionals when abnormalities are detected. Through the integration of wearable sensors, edge–cloud data analysis, and IoT communication, the proposed system delivers an economical approach for early cardiac distress prediction and prompt emergency support.

DOI: https://doi.org/10.5281/zenodo.20065529

Published by: vikaspatanker

Authors: P.Silpa Chaitanya, Ch. Lakshmi Mounika Priyadarsini, N. Sowmya, B. Pujitha, M. Lakshmi Triveni

Abstract: The rapid growth of Artificial Intelligence (AI) and machine learning has transformed automation in various fields, including education and corporate sectors. Traditional attendance systems that depend on manual entry or RFID cards are often time-consuming, inaccurate, and susceptible to proxy attendance. To address these issues, this paper proposes an AI-based smart attendance system that utilizes face recognition technology for real-time, contactless attendance marking. The system employs computer vision and deep learning models, particularly convolutional neural networks (CNN), to detect and recognize faces with high precision. Live camera input captures facial features, which are processed through image enhancement and feature extraction algorithms before being matched with a pre-trained dataset for identity verification. The system is scalable and capable of handling multiple users simultaneously. Experimental analysis shows that the model achieves over 95% accuracy under different lighting, facial poses, and occlusion conditions. This automated and secure approach reduces human intervention and offers a reliable, efficient, and intelligent alternative to traditional attendance methods.

DOI: https://doi.org/10.5281/zenodo.20065110

Published by: vikaspatanker

Authors: Mahantesh M. Ganganallimath, Dr. K. Vizayakumar, Dr. Umesh M. Bhushi

Abstract: By providing cast components to the automobile, aerospace, railroad, construction, defence, and heavy engineering industries, the Indian foundry sector is essential to the manufacturing sector. Casting flaws, process unpredictability, material waste, high rejection rates, energy inefficiency, and growing international competitiveness are some of the industry's major obstacles. In this regard, sustainable industrial growth now depends on quality assurance and quality-centric methods. The necessity of methodical quality assurance procedures, process control systems, and continuous improvement techniques in Indian foundries is examined in this study. The study highlights that quality-driven systems enhance customer satisfaction and product dependability while simultaneously lowering costs and promoting long-term competitiveness and environmental sustainability. The combination of Industry 4.0, automation, and statistical quality tools for stable growth is further supported by recent research on KPI-driven foundry quality systems and sustainable control models. An important part of the manufacturing sector, the Indian foundry industry greatly boosts employment and economic growth. This study looks into how top management influences quality management procedures in Indian foundries of different sizes. The study examines implementation difficulties, strategic quality efforts, and leadership commitment at various operational scales. The results show that whereas major foundries use organized quality management systems, small foundries encounter obstacles because of limited resources, ignorance, and opposition to change. The report suggests a framework to improve quality performance in the Indian foundry industry and emphasizes the necessity of a leadership-driven quality culture.

DOI: https://doi.org/10.5281/zenodo.20060718

Published by: vikaspatanker

Authors: Piyush Dewangan, Shivam Vishwakarma, Nikhil Yadav, Prahlad Yadav, Himanshu Mokashe, Deepak Sahu

Abstract: Global climate change poses severe threats to agricultural and forested ecosystems that underpin terrestrial carbon balance, biodiversity, and food security. This paper presents a comprehensive investigation into how Artificial Intelligence (AI)—encompassing machine learning, convolutional neural networks (CNNs), long short-term memory (LSTM) networks, transformers, and generative adversarial networks (GANs)—is transforming climate change responses across agriculture and forestry. Drawing on peer-reviewed literature and documented case studies, we examine AI applications including precision irrigation, crop disease detection, yield forecasting, satellite-based deforestation monitoring, wildfire risk prediction, acoustic biodiversity surveillance, and hydrological flood modeling. A three-tiered analytical framework maps causal pathways from technological deployment to environmental, economic, and social outcomes, while critically addressing structural barriers including data scarcity, algorithmic bias, computational inequity, and governance deficits. Principal findings confirm that AI delivers measurable gains in climate mitigation and adaptation efficiency; however, transformative societal potential remains contingent on equitable data access, open-source computational infrastructure, and coherent multilateral policy frameworks.

DOI: https://doi.org/10.5281/zenodo.20060622

Published by: vikaspatanker

Authors: Sujash Kumar Mandal

Abstract: Our present world is the world of AI, Machine learning, quantum physics where study has been shifted from bookish knowledge to online learning. Every sector has been growing and growing up rapidly; especially in higher education it shifted to online learning or virtual mood. Many courses and exams are conducted through online and also certificate is provided digitally. The present study focuses on about the attitude of digital literacy among post graduate students. Hundred samples has taken for the study; a self-made Likert scale used for the data collection by the researcher and t-test, SD, Mean deviation also used for statistical treatment. The researcher finds that there is no significant difference among post graduate students towards digital literacy according to their race, gender, locality basis.

DOI: https://doi.org/10.5281/zenodo.20056572

Published by: vikaspatanker

Authors: Dr. Harikesh Singh, Sunny Kumar, Suryanshu Singh, Shivam Singh, Smit Verma

Abstract: It is a time marked by increasing environmental degradation and disconnect from traditional knowledge systems, and the preservation and dissemination of information of medicinal plants has become critically important. The Vrikshveda project represents an attempt in creating such a comprehensive digital library documenting India’s rich botanical heritage, emphasising on medicinal plants and their therapeutic applications. This research paper presents the development and implementation of Vrikshveda, a modern web and app platform designed to distill knowledge to society about diverse plant reserves, their medicinal properties, traditional uses, and build a community around it. The system combines botanical information, traditional knowledge, scientific research, 3D AI detection and social features to create an accessible repository that serves researchers, healthcare practitioners, students, and the general public. We use web and app technologies including React, Node.js, Kotlin and MongoDB, to make Vrikshveda provide an interactive platform where users can explore detailed information about medicinal plants, including their taxonomic classification, morphological characteristics, therapeutic applications, and other content around it. The project is an attempt to answer the urgent need for documentation of indigenous botanical knowledge before we may lose it to modernization, and simultaneously promoting awareness about using them for medical benefits. Through comprehensive data collection from genuine sources, field research, and collaboration with traditional healers and botanists, Vrikshveda aims to fill the gap between centuries of wisdom and contemporary scientific understanding.

Published by: vikaspatanker

Authors: Dr. Jonathan Reed, Dr. Emily Carter, Michael Thompson, Dr. Sophia Reynolds, Andrew Richard

Abstract: The increasing deployment of machine learning (ML) systems in high-stakes domains such as healthcare, finance, criminal justice, and autonomous systems has significantly intensified concerns about transparency, accountability, reliability, and societal trust. While modern ML models particularly deep neural networks have demonstrated superior predictive performance, their complex, non-linear architectures often render them opaque, leading to criticism that they function as “black boxes” whose internal reasoning is difficult for humans to interpret, audit, or validate. This lack of interpretability poses risks in safety-critical and regulated environments, where stakeholders require clear, understandable justifications for automated decisions. In response to these challenges, Explainable Artificial Intelligence (XAI) has emerged as a crucial and rapidly evolving research area aimed at designing methods that make AI systems more interpretable, transparent, and aligned with human values, ethical principles, and legal requirements. This article presents a systematic review of Explainable AI techniques developed between 2000 and 2021, focusing on their role in enabling trustworthy machine learning systems by structuring the landscape of XAI into intrinsic (interpretable-by-design) and post-hoc (after-the-fact explanation) approaches, examining representative and widely adopted techniques such as LIME, SHAP, and Integrated Gradients, and critically discussing the methodological and practical challenges in evaluating explanation quality. Furthermore, the review analyzes how XAI intersects with broader principles of trustworthy AI including fairness, accountability, transparency, robustness, and human oversight while identifying key research gaps and outlining future directions for developing more reliable, human-centered, and socially responsible AI systems.

DOI: https://doi.org/10.5281/zenodo.20065857

Published by: vikaspatanker