Proceeding ICSEMT Feb 2026

Design of 5g Based Smart City Communication Prototype

Authors: Bommisetty Srihari, K Balasubrahmanyam, Mareddy Sai Kotireddy, Dr. U. Saravanakumar, Mr. E. Vinoth Kumar

Abstract: The rapid growth of urban populations has increased the demand for intelligent and highly connected city infrastructures. Traditional communication technologies like 4G, Wi-Fi, and ZigBee face limitations in bandwidth, latency, reliability, and scalability for handling numerous IoT devices. This project presents a 5G-based Smart City Communication Prototype, demonstrating how next generation networks enable real-time monitoring, data processing, and automated control of essential city services. The prototype integrates IoT sensors, a microcontroller (ESP32/Raspberry Pi), a 5G module, and cloud/edge computing to create an intelligent communication framework. The prototype highlights 5G’s advantages for smart city applications: higher data rates, massive device connectivity, improved reliability, and seamless integration of multiple services on a single network. Applications include smart traffic control, environmental monitoring, public safety, energy management, and emergency response. The results confirm that 5G enhances communication speed, responsiveness, and scalability compared to existing technologies. Overall, the project demonstrates that 5G-enabled communication is a robust solution for building sustainable, automated, and intelligent urban ecosystems.

DOI: https://doi.org/10.5281/zenodo.18194645

Design And Implementation Of Planar Phase Shifter

Authors: G.Jithendra, Dr.R.Praveena, N.Prabhakar, T.Sivashankar

Abstract: The implementation of a planar phase shifter using Substrate Integrated Waveguide (SIW) technology is presented for 5G high-frequency applications. SIW enables waveguide-like performance within a compact planar implementation, offering low loss, strong field confinement, and stable phase behavior. The implemented design employs a slot-loaded SIW section to produce controlled variations in the propagation constant, thereby achieving the required phase shift without significantly increasing insertion loss. The structure is realized on a Rogers RT5880 substrate and validated through full-wave simulation in CST Microwave Studio. Key performance metrics – return loss, VSWR, transmission coefficient, and phase response are measured and analyzed to verify the implementation. The results demonstrate that the implemented SIW section yields accurate phase delay, low loss, and reliable guided-wave operation, making the implementation suitable for 5G and other millimeter-wave RF systems.

DOI: http://doi.org/10.5281/zenodo.18219384

The Role Of 5G In Advancing The Internet Of Things

Authors: Mr.S. Gowtham, K. Priyadharshini, G. Pugalnethi

Abstract: The Internet of Things (IoT) has emerged as a transformative paradigm that connects physical devices to the digital world. However, the rapid increase in connected devices and data-intensive applications demands a communication in- frastructure capable of delivering high speed, ultra-low latency, massive connectivity, and reliable performance. Fifth-generation (5G) wireless technology addresses these requirements and acts as a key enabler for next-generation IoT systems. This paper presents a comprehensive study on the role of 5G in advancing IoT, focusing on network architecture, enabling technologies, applications, quality of service, security challenges, deployment issues, and future research directions.

DOI: http://doi.org/10.5281/zenodo.18299998

The Intensified Academy: A Pre-Post Pandemic Analysis Of Workload Redistribution And Well-being Among Higher Secondary Teaching Faculty In Namakkal District, Tamil Nadu

Authors: S. Shyamsundar, Dr. M. Uma Raman

Abstract: The global COVID-19 pandemic catalysed an unprecedented and involuntary transition of education from traditional classrooms to digital platforms, profoundly transforming the landscape of academic labour. This dissertation examines the lasting ramifications of this transformation on both the workload and subjective well-being of higher secondary school faculty in Namakkal District, Tamil Nadu a region representative of India’s semi-urban educational milieu. Utilizing a sequential explanatory mixed-methods approach, the research commenced with the collection of quantitative data from 327 teachers through a stratified random sampling survey, which assessed multiple facets of workload (instructional, administrative, digital, emotional) alongside well-being indicators. Subsequently, qualitative insights were garnered via phenomenological interviews with 24 purposefully selected participants to elucidate their lived experiences in depth. The findings reveal a pronounced post-pandemic intensification and reconfiguration of academic labour, not only in terms of increased working hours but also through a qualitative broadening of responsibilities, encompassing digital content curation, continuous online communication heightened emotional labour. Importantly, this restructuring has significantly blurred conventional work-life boundaries, precipitating notable declines in faculty well-being and elevating the risk of burnout. Although institutions demonstrated resilience, this was often achieved at considerable personal cost to teachers, with the burden unevenly distributed along lines of gender and school type. The study ultimately contends that the post-pandemic educational milieu constitutes a regime of intensified academic workload. It advocates for urgent policy measures that extend beyond mere technological integration to holistically address the work ecology of teachers, positioning faculty well-being as a core determinant of both educational quality and sustainable institutional resilience.

DOI: http://doi.org/10.5281/zenodo.18464791

One-pot Synthesis Of Graphene—γ-Fe2O3 Composite Through Hydrolysis In A Glycol Medium

Authors: Jyothi Sequeira

Abstract: A composite comprising graphene and γ-Fe2O3 nanoparticles can be prepared in a glycol medium by a one-pot synthesis starting from graphite oxide, GO, and FeCl3.6H2O. The precursors were dispersed in 1,2-propanediol along with urea as the hydrolysing agent and n-octylamine as the capping agent and the mixture was refluxed. Fe3+ ions undergo hydrolysis to give γ-Fe2O3 nanoparticles and GO gets reduced by glycol simultaneously leading to the formation of graphene—γ-Fe2O3 composite.

DOI: http://doi.org/10.5281/zenodo.18468090

Smart Detection of Medicinal Plants and Skin Care Applications

Authors: Ms. Divya P, Saran A, Shayam Sharan C, Sidharth M

Abstract: This project presents a Smart Detection of Skin Diseases using Traditional diagnostic systems mostly use simple image Convolutional Neural Networks (CNN). It aims to identify various comparison methods or rule-based strategies. Although these skin conditions and suggest suitable herbal remedies. The system techniques are capable of identifying outwardly apparent uses deep learning and image processing techniques to analyze anomalies, they frequently fall short in recognizing intricate user-uploaded images and detect possible skin diseases with high accuracy. By using a CNN-based classification model trained on labeled dermatological datasets, the system can automatically don't offer recommendations for appropriate treatment that are recognize disease patterns such as acne, eczema, psoriasis, and specific to the diagnosed condition. A deep learning-based fungal infections. Once detected, it recommends appropriate system that can learn from various datasets and correctly herbal or natural treatments, offering an alternative and eco- classify diseases is needed to get around these restrictions. To preprocessing to remove noise, segmentation to extract affected regions, and feature extraction for efficient classification. A web- system should also encourage natural and affordable based interface allows users to upload images and instantly receive remedies. diagnostic results along with herbal suggestions. This solution is modular and scalable, making it easy to integrate with healthcare platforms, mobile applications, or wellness systems. By combining AI-driven diagnosis with natural remedy recommendations, the system aims to improve early detection, promote awareness of traditional medicine, and provide an affordable, accessible tool for personalized skin health management.

DOI: https://doi.org/10.5281/zenodo.18492639

Personal AI Research Agent Using RAG And MCP

Authors: Mital Kadu, Abhilasha Bhagat, Vinay Alapure, Pushkar Borse, Shruti Deshmukh, Snehalata Gujar

Abstract: The rapid expansion of academic literature and digital research resources has significantly increased the complexity of modern research activities. Researchers are required to navigate large volumes of documents across multiple platforms, often leading to fragmented workflows and difficulties in maintaining contextual continuity over extended research periods. While Large Language Models (LLMs) have improved natural language interaction with research content, standalone LLM-based approaches often suffer from limitations such as hallucinations, weak factual grounding, and lack of persistent memory, reducing their effectiveness in document-centric research tasks. Recent advancements have introduced hybrid frameworks that augment LLMs with external retrieval and orchestration mechanisms to address these challenges. This review paper examines key developments in Retrieval-Augmented Generation (RAG) and orchestration frameworks such as the Model Context Protocol (MCP), focusing on their application in AI-assisted research systems. RAG enhances response reliability by grounding language model outputs in relevant source documents, while MCP enables structured coordination between LLMs and external tools, including web search and summarization services. The review also highlights the growing importance of persistent memory mechanisms for supporting long-term research continuity and cumulative knowledge building. By synthesizing findings from recent studies, this paper identifies common architectural patterns and design principles adopted in modern AI-driven research assistants. The analysis discusses the strengths and limitations of existing approaches with respect to retrieval quality, context management, scalability, and security. Additionally, open challenges related to standardized evaluation, long-term memory management, and secure tool orchestration are highlighted. Overall, this review provides a consolidated perspective on current practices and emerging trends in AI- assisted research systems, offering insights that can guide future research and development in this rapidly evolving field.

DOI: https://doi.org/10.5281/zenodo.18585351

Autonomous Border Surveillance Framework on Edge-Level Object Recognition

Authors: Tummalapalli Jaswanth Nagasairam, Dasa Divya Santhoshi, S.Deepajothi

Abstract: Growing international security threats also mean that autonomous surveillance research systems are required to undertake real-time perception, adaptative assessments of threats and decentralized functioning within limited communication conditions. The traditional monitoring systems are dependent on the centralized processing and human monitoring that lead to slow detection of threats, high bandwidth usage, and inability to be responsive to complex terrain situations. To surmount such constraints, an autonomous defence surveillance research framework of border threat intelligence is provided. The architecture involves the integration of high resolution optical imaging, thermal sensing module and unmanned monitoring platform to make persistent observation of the perimeter. Streams of sensor data are processed at the local edge, and thus can be analyzed in a low-latency manner and still operated continuously in distant areas. An object recognition and multi-object tracking mechanism using deep learning allows identifying intruders, vehicles, wildlife, aerial objects and evaluating temporal trajectory allows behavior interpretation and analysis of intrusion patterns. A rule-based threat intelligence model combines classification result, movement patterns, and contextual limitations to determine the degree of threat severity and create takeable alerts. Edge-centric inference minimizes network dependency, maximizes the level of operational confidentiality, and speed up the execution of responses. An average detection accuracy of 96.4 was experimentally assessed and the proposed output implementation model was capable of constant real-time performance under a variety of conditions with illumination levels, weather and terrain changes. Effective situational awareness, credible threat classification and efficient decision making are confirmed by observed results to improve the effectiveness of autonomous border surveillance operations.

DOI: https://doi.org/10.5281/zenodo.18628269

Health AI: Personal Health Advisor Mobile Application for People

Authors: Walid Bebal, Joydeb Nandi, Taha Ghole, Prof.S.E.Gawali

Abstract: The rapid evolution of mobile technology has cre- ated opportunities for new types of mobile applications, including applications for providing healthcare information. This paper discusses the design and implementation of a mobile health information software application with a Node.js backend. The primary function of the application is to distribute health-related information via a mobile device. The proposed system adopts a less complex architecture due to the absence of complex backend systems and databases. The system is designed such that the mobile application and backend server communicate via standard HTTP protocols. The proposed system focuses on minimizing the time for implementation and enhancing system responsiveness and simplicity. The results of the evaluation of the system’s performance under typical use conditions indicate that the application is responsive and reliable. The developed system demonstrates an efficient use of mobile and server-side technologies for educational and informational healthcare.

DOI: https://doi.org/10.5281/zenodo.18630389

DirectSmart: Resource Management Using Artificial Intelligence For Circular Economy — Bio-Bitumen Production From Agricultural Waste In India

Authors: Taranjeet Singh, Manu Raj Moudgil

Abstract: India faces interconnected challenges of agricultural residue mismanagement, environmental pollution, and increasing demand for sustainable infrastructure materials. Large quantities of rice straw are generated annually after paddy harvesting, particularly in northern India. Due to limited management alternatives, farmers frequently resort to open-field burning, resulting in severe air pollution, soil degradation, and public health impacts. Simultaneously, the road construction sector depends heavily on petroleum-based bitumen, contributing to greenhouse gas emissions and economic vulnerability due to crude oil imports. This paper proposes DirectSmart, a comprehensive Artificial Intelligence (AI)-driven resource management framework aligned with circular economy principles and the Swachh Bharat Mission. The framework integrates satellite-based identification of rice-growing regions, machine learning-based biomass forecasting, mathematical optimization of collection and logistics, AI-assisted bio-bitumen quality prediction, and governance-level environmental impact assessment. Unlike conventional approaches where AI is loosely associated with sustainability, this study demonstrates how AI functions as the central enabling mechanism that ensures technical feasibility, economic viability, environmental sustainability, and social acceptance of rice straw-based bio-bitumen production in India. The proposed framework is suitable for scalable implementation and policy-level adoption in developing economies.

DOI: http://doi.org/10.5281/zenodo.18631488

CFD-based ANN Prediction Of Fin Surface Nusselt Number For A Specific Laptop Heat Sink Design

Authors: Yogesh Chouksey, Nitin Shrivastava, Sunil Kumar

Abstract: Thermal performance prediction of compact heat sinks is important during early design stages of laptop cooling systems. While computational fluid dynamics (CFD) provides accurate evaluation of heat transfer behavior, repeated simulations are computationally intensive. In this study, a simple artificial neural network (ANN) model is developed as a surrogate tool to predict the surface-averaged Nusselt number of fins in a specific laptop heat sink design using CFD-generated data. A total of 30 data samples corresponding to six Reynolds numbers and five fin configurations were used for training and testing. Reynolds number and fin type were employed as input parameters, while the surface Nusselt number served as the output. The ANN was implemented using a single hidden-layer feedforward architecture. Model performance evaluated on the test dataset yielded a root mean square error (RMSE) of approximately 4.33 and a mean absolute percentage error (MAPE) of about 6.6%, indicating satisfactory predictive capability. The results demonstrate that the ANN successfully captures the relationship between Reynolds number, fin geometry, and convective heat transfer performance, providing a computationally efficient approach for preliminary thermal assessment of the specified heat sink configuration.

DOI: http://doi.org/10.5281/zenodo.18639815

AD-GCRS: A Generalized Clinical Reliability System For Multistage Alzheimer’s Classification Leveraging Transfer Learning Across Heterogeneous MRI Datasets

Authors: Shaik Shameer Basha, Prof. B. Sathyanarayana

Abstract: Objective: Standard Deep Learning(DL) evaluations for Alzheimer’s Disease (AD) often prioritize raw accuracy over clinical safety and cross-institutional generalization. This research proposes the Alzheimer’s Disease Generalized Clinical Reliability System (AD-GCRS), a novel framework designed to evaluate model stability and clinical risk across heterogeneous environments. Methods: The AD-GCRS framework leverages four transfer learning architectures VGG16, Xception, ResNet50, and EfficientNetB0 to classify MRI image scans into four progressive stages of impairment. We introduce a specialized evaluation suite, Clinical Deviation Error (CDE) to penalize stage-skipping misclassifications, the Index of Model Stability (IMS), and the Correct Class Index (CCI) for cross-dataset validation between ADNI and OASIS repositories. Results: Internal validation achieved accuracies exceeding 97%. However, cross-dataset testing revealed a significant "generalization gap," where the Xception model emerged as the most robust architecture with a CCI of 0.9708 and a restricted Major Error Rate (MCR) of 23.96%. Furthermore, we successfully resolved the "Scaling Paradox" in EfficientNetB0. It can be possible through a custom Lambda layer, restoring its diagnostic capability. Conclusion: The AD-GCRS provides a transparent pathway. For deploying AI in clinical settings by quantifying not just if a model is wrong, but how safely it fails

DOI: https://doi.org/10.5281/zenodo.18721750

Proximate Analysis Of The Leaves Of Landolphia Oweriences (White Vine Rubber) In Southern Nigeria Ecosystems For Nutritional Evaluation

Authors: Odimgbe Ezekiel Izudike

Abstract: The proximate composition of the leaves of Landolphia oweriences, a plant belonging to the Apocynaceae family and commonly known as the white rubber vine, was thoroughly investigated to determine its nutritional profile. Proximate analysis was conducted using the AOAC Method (2005), revealing several key components that highlight its potential as a food source. The results indicated that the leaves contained a moisture content of 8.09 ± 0.21%, which is typical for plant leaves and indicates moderate hydration. In terms of protein content, the leaves were found to contain 15.86 ± 1.46%, suggesting a good source of plant-based protein, which is essential for various physiological functions. The ash content was 4.74 ± 0.14%, which reflects the mineral content of the leaves, important for micronutrient intake.The fat content was measured at 1.07 ± 0.42%, indicating a relatively low fat level, while the crude fiber content was significantly higher at 32.51 ± 1.15%. This high fiber content contributes to digestive health and can aid in managing cholesterol levels. Additionally, the total carbohydrate content (or nitrogen-free extract) was 37.74 ± 2.01%, with soluble carbohydrates accounting for 11.59 ± 0.69%. These figures suggest that the leaves of Landolphia oweriences offer a good balance of macronutrients, making them suitable for human consumption, particularly in regions where other food sources may be limited.

DOI: https://doi.org/10.5281/zenodo.18757716

Ai Based Skin Disease Diagnostic System

Authors: Adhithya Lp, Harish A.T, Madhan.A, Lakshmi Roopa

Abstract: Skin diseases represent a significant global health concern due to their increasing prevalence and the limited availability of dermatology specialists, particularly in resource- constrained regions. Early and accurate diagnosis plays a crucial role in improving treatment outcomes and reducing disease progression. This paper presents an AI-Based Skin Disease Diagnostic System that leverages deep learning and transfer learning techniques for automated skin disease classification. The proposed system employs the MobileNetV2 architecture, chosen for its computational efficiency and strong performance on image classification tasks, making it suitable for real-world deployment. The model is trained and evaluated using the HAM10000 dataset, consisting of dermatoscopic images spanning seven different skin disease classes. Image preprocessing and data augmentation techniques are applied to improve generalization and robustness. The system achieves an overall classification accuracy of 78%, demonstrating its potential as a clinical decision-support tool. The proposed solution is implemented using TensorFlow and deployed via a Streamlit-based web interface for interactive usage. Ethical considerations are emphasized, and the system is explicitly designed as a supportive diagnostic aid, not a replacement for professional dermatologists.

DOI: https://doi.org/10.5281/zenodo.18796978

Smart Energy Meter Using Arduino Board

Authors: Challa Sreenu, Kalluri Jaya Krishna, Shaik Sajid, Dr. U. Saravanakumar, Mrs. N. Lalithakumari

Abstract: The rapid growth in electricity demand has increased the need for efficient energy monitoring and management systems. A Smart Energy Meter is an advanced digital device designed to measure electrical energy consumption accurately and transmit the data automatically to utility providers and consumers. The proposed system utilizes a microcontroller-based architecture integrated with voltage and current sensors to monitor real-time power usage. Communication technologies such as Wi-Fi or GSM enable remote data transmission, eliminating the need for manual meter reading. This system enhances billing accuracy, supports real- time monitoring, and provides consumers with detailed insights into their energy consumption patterns. Additionally, smart energy meters help in detecting abnormal usage and power theft, thereby improving grid reliability and operational efficiency. The implementation of smart energy metering plays a crucial role in the development of smart grids, energy conservation, and sustainable power management systems.

DOI: https://doi.org/10.5281/zenodo.18797471

Determination And Comprehensive Evaluation Of Microbial Load On Smoked Fish Sold In Owerri Metropolis Markets

Authors: Etus Patrick Chimuanya

Abstract: Microbial assessment of smoked Scomber scombrus spp. of fish procured from two major markets in Owerri metropolis was carried out to evaluate their microbiological quality and safety for human consumption. Freshly smoked fish samples were randomly purchased from selected vendors in each market and transported under aseptic conditions to the laboratory for analysis. The samples were analyzed using standard bacteriological and fungal culture media to determine the total viable counts and to identify the microorganisms present. The results revealed that the average bacterial counts ranged from 3.1 × 10⁶ to 6.8 × 10⁶ colony-forming units per gram (cfu/g), indicating a relatively high microbial load. In contrast, the average fungal counts ranged from 0.0 to 0.3 × 10⁶ cfu/g, which, although lower than the bacterial counts, still suggests fungal contamination in some samples. A total of four bacterial species were isolated and identified: Staphylococcus aureus, Escherichia coli, Bacillus spp., and Salmonella spp. Additionally, three fungal species were isolated, namely Mucor spp., Yeast spp., and Aspergillus spp. The presence of these microorganisms, particularly pathogenic species such as Staphylococcus aureus, Escherichia coli, and Salmonella spp., raises significant public health concerns. The higher levels of microorganisms identified from smoked fish purchased in the markets can be attributed to poor handling practices, inadequate hygiene during processing, exposure to contaminated environments, and improper smoking and storage methods adopted by fish mongers. Improved sanitary measures and proper smoking techniques are therefore recommended to ensure product safety.

DOI: https://doi.org/10.5281/zenodo.18875111

The Impact Of Digital Marketing Strategies On Sales Performance In Modern Businesses

Authors: Sibiraj.k, M. Kripalakshmi

Abstract: Ask any business owner whether they do digital marketing and the answer is almost always yes. Ask whether it is actually driving sales and things get complicated quickly. There is a real gap between spending on digital channels and understanding what those channels are genuinely doing for revenue. That gap costs money — because budgets end up allocated to what looks active rather than what demonstrably converts. This study surveyed 130 respondents — business owners and marketing professionals drawn from small, medium, and large businesses across retail, services, technology, and manufacturing — to examine how five core digital marketing strategies affect sales performance. Simple percentage analysis and cross-tabulation were used throughout to keep the findings readable and actionable. What the data showed is more specific than a general verdict on digital marketing. Email marketing leads on direct sales impact, named as the primary revenue channel by 42% of respondents. SEO follows at 31%. Paid search contributes meaningfully at the conversion stage for businesses with adequate budget. Social media, despite near-universal adoption, is named as a direct sales driver by very few respondents — its real value is earlier in the funnel, at awareness and audience-building. Content marketing shows modest short-term attribution but meaningfully improves repeat purchase behaviour and customer trust over longer timeframes. The businesses performing best digitally are not those spending the most — they are those who understand what each channel actually does and use it accordingly.