Authors: Ishaan Sreejith, Vaishnav P, Aby Jacob Varghese

Abstract: This report describes the design and construction of a small wheeled robot that senses obstacles with an HC-SR04 ultrasonic sensor and stops before hitting them. An Arduino Uno R4 Minima measures distance from the echo time of each ultrasonic pulse and drives two DC motors through an L298N H-bridge driver using PWM. The robot moves forward at a fixed duty cycle and halts when an object lies closer than 20 cm or when no echo returns. The hardware, wiring, working principle, software, build procedure and troubleshooting notes are documented so the project can be reproduced.

DOI:

Published by: vikaspatanker

Authors: Jai Akkayarajkumar V J

Abstract: ADCs play an important role in converting information from the analog world to the digital one; with modern applications requiring higher resolutions, speed, and power efficiency, it has become very difficult to design them. Traditionally, the design of an ADC follows an iterative design process that includes stages like specification, behavioral modeling, transistor sizing, schematic simulation, layout, and post-layout verification, which is laborious and requires much designer expertise due to the consideration of process-voltage-temperature variations and yield. In this work, we perform a review of the latest advancements in the use of machine learning (ML) to tackle the challenges of ADC design in recent years, with systematic search from 2015 to 2026 in IEEE Xplore, ACM Digital Library, and arXiv. Rather than organizing the literature on different ML algorithms, this paper categorizes ML approaches such as regression, tree-based models, SVM, ANN, CNN, GNN, reinforcement learning, evolutionary algorithms, and Bayesian optimization in ADC design stages such as behavioral modeling, transistor sizing, calibration of comparators and capacitor arrays, calibration of ADC, layout optimization, prediction of PVT/yield variations, and design space exploration. Comparative analysis includes the benefits provided by each technique such as speed and accuracy improvements, as well as their main drawbacks. Open challenges are then presented, including lack of data, poor generalization performance, low explainability, and the largely unexplored topic of co-optimization across multiple stages. Finally, future trends in the field are presented in terms of new directions such as generative AI, GNN-based layout optimization, physics-informed learning, digital twins, and human-AI collaborative design, as well as risks and potential problems such as manufacturability and IP in generative approaches.

DOI: https://doi.org/10.5281/zenodo.23721242

Published by: vikaspatanker

Authors: Dr. Nagendra Pal

Abstract: – Artificial intelligence (AI) and wearable fitness technologies are increasingly discussed as tools for monitoring workload, interpreting physiological signals and personalising sports training. Yet quantitative indicators alone cannot explain how collegiate athletes understand algorithmic feedback, negotiate its use with coaches, or experience the practical constraints of technology-assisted training. This paper presents a qualitative research protocol to explore athletes’ perceptions and experiences of AI-enabled applications and wearable fitness technologies in relation to sports performance and training effectiveness in Western Uttar Pradesh, India. Guided by an interpretivist orientation, the proposed study will use purposive, maximum-variation sampling and semi-structured interviews with approximately 20–25 collegiate athletes, with the final sample determined by information power. Interviews will explore perceived usefulness, training decisions, coach–athlete communication, trust in data, accessibility, privacy and unintended consequences. Reflexive thematic analysis will be undertaken through iterative familiarisation, coding, theme development and interpretive writing, supported by reflexive memoing and transparent documentation. Ethical procedures will include institutional approval, informed consent, secure data management and safeguards against coercion. As interviews have not yet been conducted, this manuscript does not report participant quotations, emergent themes or empirical findings. It offers a theoretically informed and operationally specified protocol capable of supporting a rigorous future qualitative investigation.

DOI: https://doi.org/10.5281/zenodo.23687691

Published by: vikaspatanker

Authors: Dr. Nagendra Pal

Abstract: Background: Sports participation may contribute to psychological well-being through physical, motivational and social pathways, yet its effects depend on participation contexts. Objective: This critical narrative review examines psychological benefits, potential harms, institutional barriers and opportunities relevant to college students in Western Uttar Pradesh, India. Approach: Published international and Indian research on sports, physical activity and student mental health was critically interpreted, with attention to differences in populations, study designs and geographical relevance. Findings: The literature generally associates physical activity and sport with favourable mental-health and social outcomes, but does not establish uniform causality or directly characterise organised sports participation across Western Uttar Pradesh. Recreational and competitive participation may involve different psychological demands. Contribution: An Institutional–Participation–Psychological Well-being (IPPW) framework is proposed to connect access, participation quality, psychological mechanisms and outcomes. Implications: Institutions should evaluate inclusive access and participant experiences alongside sporting achievement, and regional longitudinal research is needed. This paper is a critical narrative synthesis, not a systematic review or a report of primary fieldwork.

DOI: https://doi.org/10.5281/zenodo.23687228

Published by: vikaspatanker

Authors: Ahamed Asif Nagoor Meerasha

Abstract: Time overrun remains one of the most persistent problems in major construction projects, and more than half of UK projects still exceed their agreed programmes. Four-dimensional building information modelling (4D BIM), which links the construction schedule to a 3D model, has been promoted as a way to make planning more reliable, yet its use in practice remains uneven. This paper examines the principles of implementing 4D BIM for time management in major project delivery. It follows a qualitative approach that combines a structured review of literature, industry guidance and policy with a comparative analysis of published surveys and interview studies of 4D BIM use, mainly from the UK and Sweden. The factors that shape adoption are grouped into technical, procedural and people-related categories, and a practical implementation sequence is set out, from reliable data capture and collaborative planning to the integration of model and programme. Applications across the construction phase are then mapped, covering tender planning, construction method planning, programme communication, design interrogation, resource and workspace planning, safety and logistics. The comparative analysis shows a consistent pattern: awareness of 4D BIM is high, but routine use lags behind, and the technology is valued more for communicating and checking programmes than for creating them. The main barriers are a shortage of 4D skills, limited client demand, poor interoperability between authoring and scheduling tools, mismatched levels of development and traditional procurement. On this basis the paper proposes three practical approaches, clarity, collaboration and adaptation, supported by SMART 4D objectives, early involvement of the 4D team, staged training and a ring-fenced implementation budget. The main limitation is that the time savings are not quantified on live projects, which remains a subject for further research

DOI: https://doi.org/10.5281/zenodo.23685128

Published by: vikaspatanker

Authors: Peace Onya Ali, Ikechukwu Sampson Chikwe, Olubunmi Abiola Adewusi, Tasie Chimezie Michael, Boniface Tersoo Iorhuna, Edjere Akpoobaro Obadejire, Erienu Obruche Kennedy

Abstract: Environmental pollution is a major global challenge that threatens environmental quality, human health, biodiversity, and ecological sustainability. This review examines the sources of environmental pollutants, pollutant removal techniques, their effectiveness, and their impacts on environmental quality and ecological sustainability. Major sources of pollution include industrial activities, agricultural practices, domestic waste, petroleum operations, transportation, and improper waste disposal. Common pollutants include heavy metals, petroleum hydrocarbons, pesticides, plastics, particulate matter, and emerging contaminants. Conventional removal techniques, such as filtration, adsorption, chemical precipitation, and biological treatment, are discussed alongside emerging approaches, including phytoremediation, nanotechnology, membrane filtration, electrochemical treatment, and advanced oxidation processes. The review highlights that the effectiveness of these techniques depends on pollutant characteristics, treatment conditions, operational costs, and environmental requirements. Although pollutant removal technologies can improve air, water, and soil quality, some generate secondary waste, consume substantial energy, or produce potentially harmful transformation products. These limitations emphasize the need for integrated treatment strategies, effective waste management, pollution prevention, and continuous environmental monitoring. Sustainable pollution control should prioritize affordable, energy-efficient, and environmentally safe technologies that support ecosystem restoration, biodiversity conservation, and resource recovery. The review concludes that combining conventional and emerging remediation techniques with stronger environmental regulations and sustainable production practices is essential for reducing pollution and promoting long-term ecological sustainability.

DOI: https://doi.org/10.5281/zenodo.23659503

Published by: vikaspatanker

Authors: Bhuvaneswaran R, Dharshan M, Nandhakumar S, Selvaragavan J, Indhuja N

Abstract: Email remains a common channel for personal, educational, and business communication, but unsolicited messages and deceptive email campaigns create security and productivity concerns. This paper presents an AI- and natural-language-processing-based workflow for classifying email text as spam or legitimate (ham). The proposed pipeline validates labelled email data, normalizes text, converts messages into numerical features using Term Frequency–Inverse Document Frequency (TF-IDF), and applies a supervised classifier. Multinomial Naive Bayes is used as the baseline model, while Logistic Regression and a linear Support Vector Machine can be considered for comparison. The workflow separates dataset preparation, feature extraction, training, prediction, and evaluation so that the same fitted pipeline is used for new messages. Model quality is assessed using accuracy, precision, recall, F1-score, and a confusion matrix. The project report does not provide verified numerical evaluation results; therefore, performance values should be obtained by running the implementation on the selected labelled dataset and should not be assumed. The design provides a reproducible educational prototype that can be extended with URL analysis, multilingual processing, and email-client integration.

DOI: https://doi.org/10.5281/zenodo.23657931

Published by: vikaspatanker

Authors: Ajay Saagar P, Kavihai Arasu S, Ramachandran K, Samuel S

Abstract: Artificial intelligence tools for conversation, programming and speech are usually built and used in isolation, which forces users to carry context manually from one tool to another. This paper presents the COLLABRO AI Ecosystem, a multi-agent platform that unifies conversational assistance, autonomous software engineering and audio intelligence within a single coordinated architecture. Natural Language Processing (NLP), Machine Learning (ML) and Deep Learning (DL) techniques are used to interpret user input, infer intent and decompose goals into tasks. An orchestrator agent routes these tasks to specialised agents for dialogue, code understanding, generation and debugging, and speech and audio analysis. A shared context store lets the agents maintain a unified understanding of the user’s goal, exchange intermediate artefacts and improve through feedback. A governance layer embeds data privacy, transparency and responsible automation into every agent interaction. We describe the architecture, the coordination and routing workflow, candidate implementation techniques and the ethical safeguards, and we propose an evaluation protocol with per-agent and cross-agent metrics. The modular design supports software development, enterprise automation, content creation and accessibility applications. This paper reports design and methodology; no quantitative performance results are claimed, and measured evaluation is identified as future work.

DOI: https://doi.org/10.5281/zenodo.23657542

Published by: vikaspatanker

Authors: Harshini R, Poojabharani S, Sanjana S, Saranya R

Abstract: Lip reading is the task of recognising spoken words from the visible movements of the lips without using any audio signal. This paper presents a real-time visual speech recognition system that captures live video from an ordinary webcam, locates the mouth region, and predicts the spoken word or phrase as text. The system combines 3D convolutional layers, which learn spatial and temporal lip-movement features, with bidirectional LSTM layers that model the whole sequence in both directions, and it is trained with Connectionist Temporal Classification (CTC) loss so that no frame-level alignment is needed. A lightweight pipeline of MediaPipe landmark detection, mouth-region preprocessing and a sliding frame buffer allows the model to run on a standard computer. On a small custom dataset of four phrases, the system achieved a word error rate (WER) of 16.67% and a character error rate (CER) of about 4–5%. The results indicate that a compact, audio-free lip-reading system is feasible for assistive communication and for noisy environments.

DOI: https://doi.org/10.5281/zenodo.23615156

Published by: vikaspatanker

Authors: Tejas Nanasaheb Rahane, Sakshi Vikram Shelke, Shravani Sambhaji Hinge

Abstract: Edge detection is an important operation in digital image processing because it identifies boundaries where image intensity changes sharply. These boundaries provide useful information for object detection, shape analysis, machine vision and robotics. This paper presents a proposed FPGA-based image edge detection system using the Sobel operator. The input image is acquired from a camera or stored image source and is converted to a suitable grayscale pixel stream. A 3×3 sliding window is formed with line buffers and registers. The Sobel horizontal and vertical kernels calculate gradient components, which are combined and compared with a threshold to generate an edge image. FPGA parallelism and reconfigurable logic can support high-throughput, low-latency processing for suitable designs. The design can be described using Verilog HDL or VHDL and developed with tools such as AMD Xilinx Vivado or Intel Quartus Prime. Performance must be verified on the selected board using measured timing, resource utilization and image-quality results.

DOI: https://doi.org/10.5281/zenodo.23595201

Published by: vikaspatanker