Authors: Logeshwaran V, Bhuvaneswari B

Abstract: In recent years, the exponential growth of dataintensive applications and cloud-based services has drastically increased energy consumption in data centers worldwide. This surge poses significant challenges, including high operational costs, carbon emissions, and sustainability concerns. To address these issues, this paper proposes a comprehensive AI-based framework for energyefficient data center optimization. The approach integrates intelligent control algorithms, predictive analytics, and adaptive resource allocation mechanisms that dynamically manage workloads, cooling systems, and server utilization. Using techniques such as machine learning (ML), reinforcement learning (RL), and neural network-based control, the proposed model achieves significant energy savings while maintaining service-level agreements (SLAs). The research demonstrates how AI- driven systems can autonomously predict server loads, optimize cooling parameters, and minimize idle energy consumption. The findings contribute to the design of sustainable, selfoptimizing data centers for the next generation of green computing infrastructure.

Published by: vikaspatanker

Authors: G Vishal, G K Sharath Kumar, Karthik N Ganiga,, Sanjay Kumar, Prof. Merlin B

Abstract: An important factor in a nation's economic development is the agricultural sector. In particular, it provides millions of people in rural India with a vital source of income. One of the major issues influencing the agriculture industry is plant disease. Diseases can infect plants for a variety of reasons, such as synthetic fertilizers, outdated methods, environmental factors, etc., which affect farm productivity and thereby hurt the economy. Researchers have looked into a variety of AI and machine learning-based applications to identify plant diseases in order to address this problem.

 

Published by: vikaspatanker

Authors: Akshat Parekh, Jemit Patel, Deep Upadhyay, Jay Shah

Abstract: The increasing number of abandoned animals underlines the necessity of reliable and transparent adoption systems. Existing adoption practices often rely on scattered sources such as social media, which lack consistency, verification, and accountability. This paper presents Adoptly, a mobile-first application designed to streamline pet adoption by ensur- ing secure authentication, verified listings, and seamless communication between adopters and shelters. Built using React Native (Expo) for cross- platform compatibility, Clerk for user authentication, and Firebase for real-time storage and database services, Adoptly integrates features such as verified pet profiles, favorites, adoption requests, story sharing, and an in-app chat system. Developed under an agile methodology, the system reduces misinformation, enhances trust, and improves adoption efficiency compared to unstructured methods. Early testing demonstrates reliabil- ity in authentication, smooth navigation, and real-time synchronization. The findings highlight the potential of mobile-first systems to revolu- tionize adoption processes by bridging gaps of communication, trust, and scalability.

Published by: vikaspatanker

Authors: Aaditya Deo, Shah Saurabh, Anurag Tiwari, Ishordev Chaudhary, Dr. Sneha Soni

Abstract: Traditional digital marketing frequently encounter high costs, authenticity gaps, and declining consumer trust. Simultaneously regular customers and small creators who organically promote products seldom receive benefits. This paper present a concise research version of the project “Social Media Monetization: A Modern Advertising System” which enable customers to earn cashback by promoting purchased product from partnered brands on Instagram. The system leverage user generated con- tent to provide brands with authentic, cost effective promotion while offering customers a transparent reward mechanism. We summarize the problem context, objectives, system methodology, modules, expected outcomes, and future directions strictly based on the original project report, and reserve space for illustrative figures aligning with the submitted work. Customer Based Advertising, Cashback Reward, Digital Marketing, Social Media Promotion, Instagram Marketing, Authenticity, Secure Payment Integration, Marketing Ana- lytics, User Engagement Tracking, AI Based Advertisement Optimization, Scalable Marketing Model.

Published by: vikaspatanker

Authors: Akshat Parekh, Jemit Patel, Deep Upadhyay, Jay Shah

Abstract: The increasing number of abandoned animals underlines the necessity of reliable and transparent adoption systems. Existing adoption practices often rely on scattered sources such as social media, which lack consistency, verification, and accountability. This paper presents Adoptly, a mobile-first application designed to streamline pet adoption by ensur- ing secure authentication, verified listings, and seamless communication between adopters and shelters. Built using React Native (Expo) for cross- platform compatibility, Clerk for user authentication, and Firebase for real-time storage and database services, Adoptly integrates features such as verified pet profiles, favorites, adoption requests, story sharing, and an in-app chat system. Developed under an agile methodology, the system reduces misinformation, enhances trust, and improves adoption efficiency compared to unstructured methods. Early testing demonstrates reliabil- ity in authentication, smooth navigation, and real-time synchronization. The findings highlight the potential of mobile-first systems to revolu- tionize adoption processes by bridging gaps of communication, trust, and scalability.

Published by: vikaspatanker

Authors: D. Damodara Reddy, Soma Vivekananda, Vennam Gopala Krishna

Abstract: Primary energy consumption doubled since the 2000s, between 440 and 880 million TOE. It is projected to double within the next 20 years, reaching approximately 1,900 million TOE, and to reach 1,500 million TOE by 2030. It is anticipated to double over the next 20 years to around 1,900 million TOE, and by 2030, it will reach 1,500 million TOE. The manufacturing sector in India uses the most energy. Energy use in the global industrial sector accounts for one-third of total consumption, according to a review of energy analysis. So that effort has been made to improve the energy efficiency (EE) of the industry to enhance performance. Energy efficiency means using less energy to do an identical task while lowering energy costs and emissions. A key component of the all-encompassing plan to decarbonize industrial operations is energy efficiency. This research intends to investigate the most current systematic literature evaluations on energy efficiency in the industrial sector that were published between 2017 and 2023, taking into account this vast amount of information. The current study creates and establishes six distinct groups that reflect the state of the field's research after conducting qualitative and topical evaluation: Energy Conservation and Innovation, Energy Diagnostics, Energy Monitoring, and Energy Optimization. It consists of the automated and comprehensive formulation of measures for energy efficiency utilizing energy efficiency analysis, broad and flexible modeling of consumption of energy at various production stages to determine technological efficiency possibilities, and the comprehensive evaluation and sorting method taking into account the relationships among methods.

DOI: http://doi.org/10.5281/zenodo.17618019

Published by: vikaspatanker

Authors: Hridayjyoti Deka, Dikshita Medhi

Abstract: Women's healthcare necessitates comprehensive approaches to address unique and ubiquitous health related issues of women, which include nutritional, reproductive, mental, and chronic diseases. On the other hand, rural healthcare must overcome tangible and intangible barriers, including geographic isolation, inadequate physical and digital infrastructure, sociocultural resistance, a shortage of healthcare professionals, and policy paralysis. Artificial Intelligence and Machine Learning have revolutionized the healthcare paradigm through developments such as deep learning enabled medical imaging and diagnostics, predictive analytics, drug discovery, real time monitoring of disease surveillance, precision and personalised medicines, robotic surgery, robotic neurorehabilitation, etc. However, the benefits of these breakthroughs are mostly being received by the advanced societies; the greater rural masses still expect miracles of the trickle down effect. The healthcare issues and challenges pertaining to women are peculiar and need special focus from researchers, particularly in low resource settings like rural India. Nevertheless, the researchers have long started exploring Artificial Intelligence and Machine Learning based solutions for problems related to women’s healthcare and rural healthcare. In this article, focusing on the challenges and approaches, we review the state of the art of Artificial Intelligence and Machine Learning in women’s healthcare that carries significant potential for implementation in the rural healthcare system in India

DOI: http://doi.org/10.5281/zenodo.17617845

Published by: vikaspatanker

Authors: Siddalingaswamy R, Yogeesh N, Rajathagiri D T, M. S. Sunitha, Jagadeesha K C

Abstract: This study develops a practical modeling pipeline to treat epistemic uncertainty in sustainability-focused partial differential equations governing environmental and urban systems. We represent imprecise forcings and parameters with fuzzy numbers (triangular/trapezoidal membership functions) and propagate uncertainty via α-level analysis: for each α, parameters are mapped to compact intervals and a deterministic diffusion–reaction problem is solved to yield envelopes of feasible states. The workflow integrates (i) fuzzy parameterization and α-cut computation, (ii) numerically stable parabolic solvers (implicit/Crank–Nicolson discretizations with Dirichlet boundaries), and (iii) a stylized robust multi-objective design that visualizes trade-offs between expected performance and sustainability risk. Two representative applications illustrate relevance: groundwater-style storage under uncertain recharge–demand balance and urban heat mitigation with uncertain material/forcing properties. Results include interpretable membership curves and α-cut bounds, α-dependent terminal profiles, time-evolution bands that communicate worst-plausible excursions, and Pareto fronts clarifying yield–risk compromise under policy intensity. A grid-refinement study indicates indicative second-order spatial convergence in the smooth-solution regime, supporting numerical consistency. Beyond these cases, the framework is modular and extensible to nonlinear physics, higher dimensions, and hybrid fuzzy–stochastic formulations, while remaining transparent for expert elicitation and decision support. Overall, the approach preserves uncertainty structure without imposing unwarranted probability models, providing decision-makers with conservative, policy-ready indicators for risk-aware planning in data-sparse contexts

DOI: http://doi.org/10.5281/zenodo.17617766

Published by: vikaspatanker

Authors: L.Subhalakshmi, Dr. N.R.Santhi Maheswari

Abstract: This paper deals with pseudo irregular fuzzy soft graphs. The definition of pseudo irregular graphs is introduced with some properties. The pseudo edge irregular fuzzy soft graphs are illustrated with examples. The properties of the defined graphs are studied. Highly, neighbourly, strongly pseudo irregular graphs are explained with examples. Also some pseudo edge irregular fuzzy soft graphs are illustrated. The relation between strongly pseudo irregular fuzzy soft graphs with highly and neighbourly pseudo irregular FSG is given. Results on total pseudo irregular FSG and total pseudo edge irregular FSG is examined.

Published by: vikaspatanker

Authors: Atharva Rajesh Gosavi

Abstract: raffic sign recognition (TSR) is a critical component of autonomous driving systems, enabling vehicles to understand and respond to road regulations in real time. Traditional TSR approaches typically separate classification and localization tasks, resulting in increased computational cost and reduced robustness in complex driving environments. This paper proposes a multi-task deep learning framework that performs simultaneous traffic sign detection, classification, and attribute prediction using a shared feature-extraction backbone. The model leverages multi-task learning to exploit interrelated features across tasks, improving overall accuracy while reducing inference time—an essential requirement for self-driving applications. Extensive experiments conducted on benchmark datasets such as GTSRB and GTSDB demonstrate that the proposed approach outperforms single-task baselines, achieving higher precision in both recognition and localization. The results show that multi-task learning enhances generalization under challenging conditions, including occlusion, varying illumination, and high-speed motion. This work highlights the potential of unified deep learning architectures to deliver efficient and reliable traffic sign recognition for next-generation autonomous vehicles.

Published by: vikaspatanker