Volume 14 | Issue 3 |

 Abstract

Demand forecasting is a critical component in modern supply chain management, enabling organizations to optimize inventory levels, reduce costs, and enhance customer satisfaction. Traditional forecasting techniques often fail to capture complex patterns and dynamic market conditions. This paper proposes a machine learning-based demand forecasting system using time series analysis and the SARIMAX model. The system incorporates data pre processing, feature engineering, and model evaluation using performance metrics such as MAE, RMSE, and MAPE. Additionally, a Flask based web application is developed for real-time forecasting and visualization.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Machine Learning, Demand Forecasting, Supply Chain Management, SARIMAX, Time Series Analysis

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The quality assessment of fruits and vegetables is essential for ensuring food safety and reducing post-harvest losses. Traditional methods such as visual inspection and laboratory testing have several limitations. Visual inspection can only evaluate the external appearance and cannot detect internal spoilage, hidden damage, or contamination, while laboratory testing is costly, destructive, and time-consuming. This paper presents Nutri Fusion X, a smart non-destructive fruit and vegetable quality inspection system based on multi-sensor technology and IoT. The system integrates gas, optical, UV, NIR sensors, ultrasonic sensing, and image analysis to evaluate fruit freshness, ripeness, spoilage, pesticide contamination, and internal damage (e.g., bruises or soft spots) without cutting the fruit. An ESP32 microcontroller processes sensor data and transmits results in real time to a mobile application via Wi-Fi. The application displays quality indicators, including freshness percentage, ripeness level, spoilage risk, and safety alerts. The proposed system is portable, low-cost, and user-friendly, making it suitable for farmers, vendors, and the food industry while helping to reduce food waste and improve food quality monitoring.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Fruit Quality Detection, IoT, ESP32, Multi-Sensor System, Non-Destructive Testing, Smart Agriculture, Food Safety, Image Processing

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

India's Electric Vehicle (EV) sector is experiencing rapid growth, fuelled by government incentives, rising environmental concerns and technological advancements. With initiatives like the Faster Adoption and Manufacturing of Hybrid and Electric Vehicles (FAME) scheme, India aims to significantly increase EV adoption, revolutionizing its transportation landscape towards sustainability and innovation. The country has set an ambitious target to elevate EV sales to 30% of private cars, 70% of commercial vehicles, 40% of buses and 80% of two- and three-wheelers by 2030, translating to approximately 80 million EVs on Indian roads, alongside a focus on complete domestic EV production through the 'Make in India' initiative. According to ICRA, India's auto component industry will get investments of Rs. 25,000-30,000 crore (US$ 2.89-3.46 billion) in FY26 for capacity expansion and EV parts, up from Rs. 15,000-20,000 crore (US$ 1.73-2.31 billion) in FY25.Due to the nearly low to zero carbon emissions, low noise, great efficiency, and flexibility in grid operation and integration, electric vehicles (EVs) are a viable technology for creating a sustainable transportation industry in the future.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Electric Vehicle, FAME Scheme, Government Incentives, Environmental Concerns, Technological Advancements, Financial Year

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Financial inclusion has emerged as a cornerstone of sustainable development, particularly in rural economies where access to formal financial services remains limited. This paper explores the intricate relationship between financial inclusion and rural development, emphasizing how inclusive financial systems can empower marginalized communities, enhance entrepreneurship, and foster socio-economic growth. Despite significant policy interventions, rural populations in many developing countries continue to face barriers such as inadequate infrastructure, low financial literacy, and socio-cultural constraints. Through a systematic review of existing literature, this study highlights the evolution of financial inclusion research, identifies gaps in rural outreach, and underscores the transformative potential of digital finance. The need and scope of the study are rooted in the recognition that rural development cannot be achieved without equitable access to credit, insurance, savings, and payment systems. The objectives include analysing the role of financial inclusion in rural empowerment, evaluating policy frameworks, and proposing recommendations for strengthening rural financial ecosystems. Methodologically, the paper adopts a qualitative approach, synthesizing secondary data from scholarly articles, government reports, and international development agencies. Findings suggest that financial inclusion enhances rural resilience by promoting entrepreneurship, reducing poverty, and enabling social mobility. However, challenges such as digital divides, gender disparities, and institutional inefficiencies persist. Recommendations include strengthening financial literacy programs, leveraging fintech innovations, and fostering public-private partnerships. The conclusion reiterates that financial inclusion is not merely a policy goal but a developmental imperative for achieving inclusive growth in rural economies.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Financial Inclusion, Rural Development, Digital Finance, Financial Literacy, Poverty Reduction, Entrepreneurship, Inclusive Growth

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Abstract-- The efficiency of photovoltaic (PV) systems is significantly influenced by operating temperature. Elevated temperatures reduce power output and overall system performance. This paper presents a comprehensive study on modern PV cooling technologies with a focus on improving energy efficiency and sustainability. Various cooling techniques, including passive, active, and hybrid methods, are reviewed. Additionally, a conceptual experimental setup for a hybrid PV cooling system is proposed. The findings highlight the importance of thermal management in solar energy systems and demonstrate the potential of emerging cooling technologies in enhancing PV efficiency.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Photovoltaic systems, cooling techniques, renewable energy, thermal management, solar efficiency, hybrid cooling.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Maintaining physical fitness is essential for a healthy lifestyle, yet many individuals lack access to personalized fitness guidance. Traditional fitness applications often provide generic workout plans that do not consider individual characteristics such as age, body parameters, and fitness goals. With the advancement of Artificial Intelligence (AI) and Machine Learning (ML), intelligent systems can now analyze user data and generate customized health recommendations.This paper proposes an AI-based Personalized Fitness and Wellness Coach that provides tailored workout recommendations using machine learning techniques. The system collects user inputs including age, height, weight, gender, and fitness goals through a web interface. These inputs are processed by a machine learning model trained on a fitness dataset to generate suitable workout and wellness recommendations. The proposed system is implemented as a web application using Python Flask for backend processing and HTML, CSS, and Bootstrap for the frontend interface.The developed system demonstrates how AI can be applied to create intelligent and accessible fitness guidance platforms. The results show that the proposed system effectively generates personalized fitness plans and improves user engagement with healthy lifestyle practices. The system can be further enhanced by integrating wearable fitness devices and advanced AI models for real-time health monitoring.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Artificial Intelligence, Machine Learning, Fitness Recommendation System, Personalized Healthcare, Flask Web Application

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Artificial Intelligence (AI) agents increasingly require context awareness to interact with dynamic environments and make intelligent decisions. The Model Context Protocol (MCP) is a standardized mechanism that enables AI agents to access contextual information, tools, and data sources seamlessly. This paper proposes the design and implementation of an MCP-enabled context-aware architecture for AIML-based agents to effectively manage context and improve decision-making. We integrate MCP into traditional AIML conversational agents--enhancing their ability to retrieve external context, maintain semantic coherence, and perform real-time reasoning. Experimental evaluations on response accuracy, context consistency, and scalability demonstrate significant improvements over baseline AIML agents. The contributions include a detailed system architecture, protocol integration methodology, and analysis of context-aware performance in interactive sessions.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

AIML, Context-Aware AI Agents, JSON-RPC, Model Context Protocol (MCP), Agent Communication Protocols

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

To control wildlife and protect its resources, specific identification and behavioural patterns of animal species are required to better comprehend processes in the ecology and to guarantee environmental safety. The traditional methods of monitoring rely wholly on manual observation, which is time-consuming and requires a lot of labour and is prone to human error. In order to remove such constraints, the current study suggests a self-driven framework of identifying the animal species and behaviour through a combination of spatial and temporal features analysis with the assistance of deep learning. A set of 40,000 images of 25 animal species is obtained by assembling three reference repositories and is pre-processed in ways like resizing, normalization, and data augmentation to improve the quality, consistency, and ratio of data. The suggested framework uses the VGG19 convolutional neural network to obtain highlevel spatial features, such as textures and edges, as well as the shape representations. The features extracted are then fed into a Bidirectional Long Short-Term Memory (BiLSTM) network in order to take advantage of the relationships of time by carrying out forward and backward analysis of sequential patterns of the features. The hybrid VGG-19 and Bi-LSTM architecture can be useful in the classification of animal species as well as their respective behavioural activities. Also, an automatic warning system that is active in real-time is provided to impound possible risky behaviour, and active surveillance and intervention in the environment are easy in what concerns wildlife. By experimental analysis, it is proved that the suggested system attains high classification rate and high stability through effective integration of spatial and temporal features learning and can be used in the real-life wild life monitoring and conservation.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Animal detection, VGG-19, Bidirectional Long Short-Term Memory, Deep learning.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The era of the internet has seen social media emerged a powerful platform of political communication. The fast development of the internet connection and smartphone devices has largely increased the scope of social media marketing, such as Facebook, WhatsApp, Instagram, and X are some of the most commonly used social media platform tools in India. Political parties, leaders, and citizens have become a common user of these platforms to disseminate political information, discuss issues, as well as shape popular opinion. This paper will look at how social media can impact on electoral behaviour in India. It discusses the role of social media platforms in enhancing awareness about politics, voter turnout, and campaigning. Another aspect brought out in the study is the influence of social media in forming the opinions of politics and the dissemination of information as well as misinformation via the election times. The study is founded on secondary data that was gathered by use of books, academic journals, articles and reputable sources on the internet. The results show that social media has a considerable effect on the political involvement and the voting patterns to a large extent among young voters. Nevertheless, such problems as the misinformation, political polarization, and digital manipulation are critical issues. The research concludes that, although social media has enhanced political communications, there is need to ensure it is used responsibly and one must be digitally literate to ensure healthy democratic process.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Social Media, Electoral Behaviour, Political Communication, Political Awareness, Digital Platforms and Elections.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Customer reviews play a major role in e-commerce platforms' ability to assist customers in making selections. However, one of the biggest problems facinginternet markets is the increasing number of fraudulent or deceptive evaluations. These fraudulent reviews can mislead consumers anderodetheirfaithintheplatformbecause they are frequently created to falsely promote or harm the reputation of goods and merchants. The enormous amount of onlinefeedbackcreateddailymakesit challenging to manually identify such bogus reviews. The capacity of e-commerce platforms to help customers make decisions is greatly influenced by user feedback. However, the rise in fake or misleading ratings is one of the major issues facing online markets. Because these fraudulent evaluations are often made to falsely promote or damage the reputation of products and merchants, they have the potential to deceive customers and undermine their trust in the platform.Itisdifficulttomanuallyidentify suchfraudulentevaluationsduetothe massive volume of online comments generated every day.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Fake Review Detection, Deep Learning, NaturalLanguageProcessing,E-commerce Platforms, Sentiment Analysis, Machine Learning

  License

Creative Commons Attribution 4.0 and The Open Definition