Journal of Science Engineering Technology and Management Science

Environmental research focuses on renewable energy systems because there is increasing power consumption and limited availability of traditional energy resources. The suggested solar power producing circuit consists of three critical components: solar array and boost converter and boost inverter. A dc-dc boost converter enables both battery recharge and the conversion of photovoltaic array low voltage into high-quality sinusoidal ac voltage. Unfiltered solar energy from the boost inverter directs its output to independent loads. Due to rising power requirements and conventional energy source depletion researchers now pursue renewable energy systems. The solar power generating circuit presents three major components which include the solar array together with boost converter along with boost inverter. The dc-dc boost converter powers up the battery supply by transforming solar array voltage from low to high-quality sinusoidal form. The boost inverter can transmit solar energy directly to its independent load since it does not require intermediate filters or conversion stages.

Keywords : Fuzzy logic controller, solar photovoltaic, boost converter, boost inverter, and total harmonic distortion

Author : Sasidhar Reddy and Sirisha

Title : Boost Converter and Boost Inverter Implementation of a Fuzzy Logic Controller for Photovoltaic Power Generation

Volume/Issue : 2025;02(04)

Page No : 01-07

The main purpose of studying variable drive systems with Switching Reluctance Motors (SRM) for torque density enhancement and efficiency improvement focuses on minimizing torque fluctuations. This ripple occurs due to the change in excitation current between adjacent phases. A smooth torque requires exact control of how switches turn the motor on and off. Simulation of all possible angle selection consequences appears in this research. The excitation current shape in the stator coil and its overlapping point with its adjacent coil depends on how precisely the turn-on and turn-off angles are selected as illustrated by the case of extended turn-on and turn-off excitation timing. Correct selection of turn-on and turn-off angles enables the implementation of transition phase smoothness. Researchers have performed a comprehensive analysis to determine how the modified excitation current interacts with torque ripple based on different dwell angle conditions. Proper selection of these parameters leads to the generation of sinusoidal current forms.

Keywords : Torque ripple, electric motor, switched reluctance motor, FE analysis, and dwell angle

Author : Madhu

Title : Effects of Dwell Angles of Current in Switched Reluctance Motors

Volume/Issue : 2025;02(04)

Page No : 08-14

The proposed study introduces an indirect field-oriented controller to handle five-phase induction motor drives. The controller has its foundation on the fuzzy logic control technique. This paper uses the MATLAB/Simulink package to conduct its simulations. The experimental control system was created through a digital signal processing (DSP) circuit. The performance testing of the proposed system occurs in different operating settings. The proposed controller demonstrates reliability for operation in five-phase induction motor drives with high performance levels. The proposed methods receive verification from experimental trial results and simulation testing.

Keywords : Five-phase induction motor, five-phase inverter, fuzzy logic, and digital signal processor (DSP)

Author : Shalini and Pravallika

Title : Fuzzy Logic Control for DSP-Based Vector Control of a Five-Phase Induction Motor

Volume/Issue : 2025;02(04)

Page No : 15-22

The research concentrated primarily on the meter side of customers because it investigated harmonic effects at K.Wai Perlis' Electrical System Engineering School laboratory. Magnetic fields from harmonic voltage frequencies rotate at the same speed as harmonic frequencies but harmonic currents inside equipment result in operational overheat and performance losses. This disruption produces mechanical noises and equipment vibrations which also causes overheating of equipment and operating efficiency reduction as well as voltage stress against insulation of equipment windings. The research employs Computers and Printers as well as Air Conditioners and Compact Fluorescent Lamps as harmonic-significant loads. The measurement process depends on the Fluke 435 Power Quality Analyzer. The harmonic current parameters fluctuate unpredictably due to the installation of different nonlinear loads. Through this research study the utility obtains substantial knowledge about harmonic effects on their equipment. The utility will gain knowledge about the expenses and performance losses which result from using these pieces of equipment. This paper presents information about harmonic investigation expenses through THD measurement as well as energy loss costs.

Keywords : THD, cost, energy losses, nonlinear loads, and harmonics

Author : Rohit, Shanmukh and Srinivasa Rao

Title : Nonlinear Loads Effect on Harmonics in the Laboratory

Volume/Issue : 2025;02(04)

Page No : 23-26

This research analyses how variations in multi-area power system automatic generation control (AGC) work when an HVDC system functions with an HVAC system taking into account system parameter alterations. A fuzzy logic controller serves to connect four areas through parallel HVAC/HVDC transmission link systems also referred to as asynchronous tie lines. A linear model of HVAC/HVDC becomes available for evaluation purposes while analyzing abrupt load variations in the system. The investigation uses a four-area interconnected thermal power system. A system whose dynamic performance has been improved will deliver appropriate solutions to the automatic generation control problem affecting the four-region electrical power system. The controller robustness evaluation happens through varying system parameters. The simulation output demonstrates effectiveness of this approach. Advanced Fuzzy logic controllers in MATLAB-Simulink perform dynamic system analysis both in presence of and absence of an HVDC link. The suggested system achieves better results regarding settling time and overshoot compared to the existing model.

Keywords : Fuzzy logic controller, HVDC link, automatic generation control, area control error, and generation rate limitation

Author : Ramesh and Vijay

Title : HVDC Links for Automatic Generation Control in a Multi-Area Interconnected Power System

Volume/Issue : 2025;02(04)

Page No : 27-33

The Auto wired high rate of digital technologies development along with the integration of the internet to almost every area has considerably changed consumer behavior, especially in the sphere of retailing. This paper discusses feedback of the consumers on online shopping and the most important and influential characteristics involved in it including convenience, prices, availability of goods, safety, confidence and satisfaction after receiving goods. The study method will be qualitative based on a well-designed questionnaire that will be issued to all age groups, income and geographic disciplines of the consumers. It has been seen that convenience and time saving figures are the main factors influencing the positive perception of online shopping with the second-best position held by availability of a wider variety of products and competitive prices. Nevertheless, issues of the quality of products, lagging deliveries, and information privacy still represent barriers to further adoption. The results also impose the differences in terms of the demographics where young consumers express stronger preference and trust to using online platforms than the older segments of the population. It is found, in the study, that the acceptance of online shopping is being increased and yet it requires longterm consolidation through greater consumer confidence, adequate logistics, and added value. Both findings can be used by e-commerce businesses and marketers to ensure that they align their approaches to changing consumer demands.

Keywords : Consumer perception, Online shopping, E-commerce, Trust, Convenience, Consumer behavior, Digital retail

Author : Raghvendra K.Mishra,Anuj Pathak

Title : EVALUATING CONSUMER SENTIMENTS AND TRUST TOWARDS DIGITAL SHOPPING CHANNELS

Volume/Issue : 2025;02(04)

Page No : 34-41

Star type classification plays a crucial role in astrophysical research and space exploration. Identifying different types of stars helps in understanding stellar evolution, examining their physical characteristics, and exploring the properties of celestial bodies throughout the universe. Accurate classification supports cosmological research, improves models of stellar lifecycles, and enhances the accuracy of the Hertzsprung-Russell diagram. It also benefits practical applications such as spacecraft mission planning, telescope-based observations, and large-scale astronomical surveys through automated categorization of stars. Traditional methods for classifying star types, such as statistical techniques and decision trees, often fall short in performance. These approaches typically struggle with capturing the complex, nonlinear relationships present in astronomical data and underutilize available features. Moreover, manual feature engineering becomes inefficient and impractical when applied to extensive datasets, resulting in lower accuracy and reduced generalization to new star types. In this work, we focus on key features such as temperature, luminosity, radius, magnitude, color, spectral class, and star type labels, including Red Dwarf, Brown Dwarf, White Dwarf, Main Sequence, Super Giants, and Hyper Giants. We conduct a detailed evaluation of multiple machine learning (ML) models for star type prediction and propose an enhanced approach aimed at improving both classification accuracy and computational efficiency

Keywords : Star Classification, Automation, Machine Learning, Space Exploration, Machine learning.

Author : Dr. Sundeep Kumar K, V Bharathi, D. Ramesh, U. Satyanarayana

Title : STAR CLASSIFICATION AUTOMATION USING MACHINE LEARNING ON NASA DATA

Volume/Issue : 2025;02(04)

Page No : 42-48

Embryo classification plays a crucial role in assisted reproductive technology (ART), especially in the context of in vitro fertilization (IVF). In India, IVF has witnessed rapid growth, with more than 1,500 clinics performing approximately 2.5 lakh cycles annually as of 2023, according to the Indian Society of Assisted Reproduction. Traditionally, embryo assessment is performed manually by embryologists using the Gardner grading system, which involves evaluating morphological characteristics such as blastocyst expansion, inner cell mass, and trophectoderm quality under a microscope. However, this method is inherently subjective, often influenced by the embryologist’s experience, fatigue, and inconsistent application of grading criteria. This subjectivity leads to significant inter- and intraobserver variability and limits the accuracy and scalability of embryo selection, contributing to stagnant IVF success rates of 30–40%. To address these challenges, a hybrid deep learning model combining Convolutional Neural Networks (CNN) with a Cat Boost Classifier (CBC) is proposed. This AI-driven approach aims to automate the classification of embryos into categories such as "normal" or "viable," thereby reducing human error and enhancing the predictive accuracy of implantation potential. CNNs are used to extract detailed features from microscopic images of embryos by resizing and normalizing pixel values, while the CBC performs efficient classification based on these features. The model not only improves consistency but also significantly boosts performance, achieving an accuracy of up to 99.06%. By enabling faster, data-driven, and objective decision-making, the proposed system overcomes the limitations of manual evaluation. It enhances embryo selection, increases implantation success rates, and offers a scalable solution for modern IVF practices in India

Keywords : Embryo Classification, Deep Learning, CNN, In Vitro Fertilization (IVF), Artificial Intelligence (AI).

Author : Dr. P. Nagendra Kumar, Sk. Asiff, B. Poojitha, D. Ramesh

Title : MICROSCOPIC EMBRYO CLASSIFICATION USING AN INTEGRATED DEEP LEARNING FRAMEWORK

Volume/Issue : 2025;02(04)

Page No : 49-56

Obesity is a growing global health concern, influenced by a complex interplay of genetic, environmental, and lifestyle factors. Among these, dietary habits are a key contributor, making it essential to understand how eating behaviors impact obesity levels. This research investigates the use of machine learning (ML) techniques to predict obesity based on individuals' dietary patterns and demographic characteristics. A comprehensive dataset containing information on age, gender, eating frequency, food type, physical activity, and obesity levels is utilized. Several machine learning algorithms, including Decision Trees, Support Vector Machines (SVM), Random Forests are implemented to develop predictive models. Feature selection techniques are applied to identify the most influential variables affecting obesity. Model performance is evaluated using standard classification metrics such as accuracy, precision, recall, and F1-score. The results demonstrate that ML algorithms can effectively predict obesity levels, with some models outperforming others in terms of accuracy and generalization. Notably, feature importance analysis reveals specific dietary behaviors and food groups—such as high consumption of fast food, sugary beverages, and low physical activity levels—as strong predictors of obesity. This research underscores the potential of machine learning in public health by offering data-driven insights into the dietary factors contributing to obesity. The findings can inform the development of personalized nutrition recommendations, targeted health interventions, and evidence-based public health policies. Ultimately, this work contributes to the advancement of predictive healthcare analytics and supports efforts to combat the global obesity epidemic through informed decision-making and lifestyle modifications.

Keywords : Obesity, Machine Learning, Dietary habits, Extra Trees Classifier, Health

Author : Sk. Asiff, V. Bharathi, D. Ramesh, B. Poojitha

Title : PREDICTING OBESITY TRENDS: THE ROLE OF DIETARY HABITS THROUGH MACHINE LEARNING MODELS

Volume/Issue : 2025;02(04)

Page No : 57-63

Effective waste management is essential in India to ensure environmental sustainability and safeguard public health. With rapid urbanization and a growing population, the volume of waste generated has increased significantly, placing immense pressure on traditional disposal systems. This has led to environmental pollution, ecosystem damage, and various health risks. Conventional waste management methods—such as manual sorting, rule-based categorization, and exporting waste—are increasingly proving to be inadequate. Manual sorting is time-consuming, labor-intensive, and prone to human error, while rule-based systems lack the flexibility to adapt to varied and changing waste types. Exporting waste, on the other hand, raises serious environmental and ethical concerns due to potential mishandling. To address these issues, this research leverages a large-scale image dataset containing millions of waste item images to build and assess robust waste classification models. It utilizes MobileNetV2 and a lightweight convolutional neural network (CNN) to develop an AI-driven system capable of accurately classifying different types of waste. This intelligent classification system can be integrated into smart bins and recycling infrastructures to automate the sorting process, ultimately improving recycling efficiency and minimizing environmental impact.

Keywords : Smart waste management, Trash classification, Artificial Intelligence, MobileNetV2, CNN.

Author : U. Satya Narayana, B. Poojitha, Sk. Asiff, Dr. P. Nagendra Kumar

Title : SMART WASTE MANAGEMENT: AI-BASED TRASH CLASSIFICATION SYSTEM

Volume/Issue : 2025;02(04)

Page No : 64-71

Since its inception in 1969, the Indian Space Research Organisation (ISRO) has been at the forefront of space exploration, contributing significantly through a wide range of satellite missions in areas such as communication, Earth observation, and navigation. Traditionally, the prediction of satellite lifetimes has relied on empirical models, component testing, and expert judgment. While effective to a certain extent, these conventional approaches often struggle with analyzing large volumes of data and are limited in predictive accuracy. To address these challenges, this research proposes the application of machine learning techniques to model and predict the expected lifetime of ISRO satellites. By leveraging historical mission data and advanced analytical methods, the aim is to enhance the precision of lifetime estimations, ultimately supporting better mission planning, resource management, and operational decision-making. The methodology involves collecting comprehensive datasets from past satellite missions, including operational parameters, environmental influences, and component performance records. This data will undergo preprocessing to identify significant patterns and correlations. Using this processed information, a suite of machine learning models will be developed to estimate satellite lifespans based on variables such as launch conditions, operational anomalies, and degradation behaviors. Additionally, visualization tools will be incorporated to provide actionable insights in an intuitive format, enabling more informed planning by mission teams

Keywords : ISRO Space Missions, Satellite Lifetime Prediction, Time Series Forecasting, LSTM Regressor, Artificial Intelligence

Author : D. Ramesh, N.V.S. Praneeth Ramana, MD. Hameeduddin Ahmed, K. Vishnu Vardhan, K. lakshmi Nandan

Title : ENHANCING ISRO’S MISSION PLANNING WITH AI-DRIVEN SATELLITE LIFETIME PREDICTION MODELS

Volume/Issue : 2025;02(04)

Page No : 72-79

Liver diseases such as fatty liver, cirrhosis, and hepatitis are becoming increasingly widespread due to factors like poor diet, excessive alcohol consumption, and sedentary lifestyles. Given the liver’s vital role in human physiology, early detection of such conditions is critical to reducing the high mortality rates associated with liver dysfunction. Traditional diagnostic techniques, including biopsy, X-ray, CT, and MRI scans, aid in the early identification of liver-related abnormalities. This research introduces an advanced framework for liver disease detection utilizing artificial intelligence (AI) and pre-trained deep learning models. The study primarily focuses on CT scan images for liver disease classification. To enhance computational efficiency and improve model accuracy, the Social Spider Optimization (SSO) algorithm is implemented for feature selection and dimensionality reduction from the CT images. Following feature extraction and reduction via SSO, three established Convolutional Neural Networks (CNNs) are employed to classify the liver images. These CNN models are trained and finetuned using the optimized feature sets produced by the SSO algorithm. A publicly accessible liver CT image dataset is used to assess the performance of the proposed models through various evaluation metrics. This structured approach enables an in-depth analysis of the influence of feature reduction on classification outcomes. By integrating SSO-based feature optimization with powerful pre-trained CNN architectures, the proposed method demonstrates the potential of AI-enhanced tools in medical imaging and accurate liver disease diagnosis

Keywords : Liver Disease Detection, CT Scan Analysis, Deep Learning, Social Spider Optimization

Author : Dr. P. Nagendra Kumar, K. Sai Likith, K. Tanmay, K. Avinash, K. Yashwanth Kumar

Title : FEATURE-OPTIMIZED DEEP LEARNING MODEL FOR LIVER CT SCANS VIA SOCIAL SPIDER OPTIMIZATION

Volume/Issue : 2025;02(04)

Page No : 80-86

This research presents an advanced framework for detecting fake tweets using a dedicated Twitter fake tweet dataset. The goal is to improve detection accuracy by combining natural language processing (NLP) and machine learning techniques. Traditional manual detection methods face several limitations, including time-consuming analysis, susceptibility to subjective bias, and poor scalability in the face of the rapid growth of social media data. These methods often fail to identify subtle linguistic and contextual cues associated with misinformation. The proposed approach addresses these challenges through a multi-step pipeline. It begins with NLP preprocessing to clean and standardize the tweets, followed by FastText embeddings to transform textual data into numerical vectors for detailed analysis. The dataset is then split into training and testing sets to ensure robust evaluation. A deep learning convolutional neural network (DLCNN) is used for feature extraction, capturing complex patterns within the data. These features are then classified using a Random Forest classifier to determine whether tweets are real or fake. The model's performance is rigorously evaluated using appropriate metrics, demonstrating its effectiveness and reliability in real-world scenarios

Keywords : Convolutional Neural Network, Random Forest Classifier, Tweet Classification, Fake tweet Detection, Machine Learning, Deep learning

Author : Dr. P. Babu, A. Sruthi, D. Sumanvitha, J. Jasmitha

Title : DETECTING DEEP FAKE TWEETS WITH FAST TEXT EMBEDDINGS AND CNN-DRIVEN RANDOM FORESTS

Volume/Issue : 2025;02(04)

Page No : 87-93

The project introduces an AI-powered automated diagnosis system designed to accurately and efficiently detect eye diseases from retinal images. By combining advanced deep learning methodologies with an intuitive graphical user interface, the system provides clinicians and researchers with a reliable tool for early disease detection and in-depth analysis. The application begins with robust dataset management capabilities, allowing users to upload retinal images and categorize them into specific diagnostic classes, including Diabetic Retinopathy (DR), Macular Hole (MH), Normal, and Other Diseases/Conditions (ODC). Essential preprocessing steps such as image resizing, normalization, and advanced data augmentation are incorporated to increase dataset diversity and enhance model generalization. At the core of the system are two deep learning model architectures. The first is a pre-existing Deep Neural Network (DNN) trained using the Stochastic Gradient Descent (SGD) optimizer. This model is composed of multiple convolutional layers, batch normalization, pooling, and dense layers, enabling it to effectively extract and learn complex features from retinal images. The second model is a newly proposed Convolutional Neural Network (CNN) utilizing the Adam optimizer with a specific configuration of valid padding (AVP). This architecture incorporates additional dropout layers and batch normalization to reduce overfitting and improve generalization capabilities. Both models are rigorously evaluated using key performance metrics such as accuracy, precision, recall, and F1-score. Detailed classification reports and confusion matrices offer comprehensive insights into the models’ diagnostic performance. Experimental results show that while both models perform reliably, the proposed CNN with AVP outperforms the DNN in terms of diagnostic accuracy and the ability to distinguish between subtle disease features. The system’s modular architecture and extensive use of data augmentation make it adaptable to various clinical datasets and scalable for real-world implementation. This research not only demonstrates the effectiveness of deep learning in the medical imaging domain but also delivers a practical, userfocused solution for automated retinal disease diagnosis contributing significantly to the development of next-generation computer-aided diagnostic systems

Keywords : Retinal Disease Diagnosis, Macular Hole, Medical Image Analysis, Deep Learning, Computer-Aided Diagnosis

Author : U. Satya Narayana, B. Navitha, Ch. Lathasri, M.V.P. Praneetha, N. Sukanya Reddy

Title : A DEEP LEARNING MODEL WITH SMART PADDING FOR EARLY RETINAL DISEASE DIAGNOSIS

Volume/Issue : 2025;02(04)

Page No : 94-104

With the rapid rise in electric vehicle (EV) adoption worldwide, the need for a reliable and efficient charging infrastructure has become more critical than ever. One of the major challenges faced by EV users today is the lack of a proper reservation system at charging stations. Most existing stations operate on a first-come, first-served basis, forcing users to manually check availability upon arrival, which often leads to long queues, wasted time, and user frustration. To overcome these issues, this system introduces an automated EV slot reservation solution designed to streamline the charging process. The proposed system enables users to remotely pre-book charging slots through a userfriendly interface, significantly enhancing accessibility and convenience. It integrates various technologies including IoT devices for real-time monitoring of charging slot status, a Real-Time Clock (RTC) module to ensure accurate scheduling and time tracking, GSM modules for seamless remote communication, and an LCD display for clear on-site user interaction. These components work together to provide live updates on slot availability and ensure bookings are managed efficiently. By allowing users to plan their charging sessions in advance, the system not only reduces congestion at charging stations but also improves the overall user experience. Furthermore, the use of remote communication and automated scheduling enhances the security and management of the charging infrastructure. Ultimately, this smart reservation system aims to make EV charging more efficient, reliable, and user-centric, supporting the continued growth of electric mobility

Keywords : EV Charging Infrastructure, EV Slot Reservation, LCD, Keypad, ESP3

Author : B. Suresh Kumar, Kona Sumanth, Bommisetty Hemendra Siva Sai, Kaki Praveen Kumar

Title : REVOLUTIONIZING EV CHARGING WITH REAL-TIME RESERVATIONS

Volume/Issue : 2025;02(04)

Page No : 105-110

Agriculture serves as the backbone of India’s economy, with Andhra Pradesh standing out as a major contributor in rice production and distribution. Maximizing rice yield requires effective management of resources like water and land, alongside efficient disease control strategies. Crop disease management plays a vital role in improving the economic conditions of farmers by reducing infection rates and ensuring healthier, more productive crops. Traditionally, disease identification has relied on manual inspections by farmers, guided by personal experience, or consultations with agricultural experts who conduct laboratorybased, rule-driven diagnostics. While useful, these conventional methods are time-consuming, error-prone, and require substantial human effort especially in the face of rising disease variants. To address these challenges, this study proposes a deep learning-based approach for automatic rice leaf disease detection. By using rice leaf images as input, the model identifies specific diseases and recommends appropriate pesticides accordingly. This method enhances both the accuracy and speed of disease diagnosis, reducing the dependence on manual inspections and enabling timely interventions. Ultimately, the adoption of this AI-driven system aims to improve rice crop yield, support farmer decision-making, and boost overall agricultural productivity

Keywords : Rice Leaf Disease, Deep Learning, Crop Disease Detection, Agricultural AI, Precision Farming, Image Classification, Smart Agriculture, Pest Control Recommendations, CNN, Andhra Pradesh Agriculture

Author : B. Akhila, Karra Parichitha, Mannepalli Thrisha, Balu Deepthi, Gundemeda Manjula

Title : DEEP LEARNING-BASED DETECTION OF RICE LEAF DISEASES FOR SMART AGRICULTURE

Volume/Issue : 2025;02(04)

Page No : 111-122

The rapid increase in mobile users and data traffic has significantly influenced the evolution of wireless communication systems from 1G to 5G. In real-world urban environments, challenges such as "real call jumps" which lead to rapid signal fluctuations due to reflections and diffractions complicate reliable communication. Moreover, issues like channel propagation errors and overlapping communication paths degrade signal quality, resulting in increased path loss and a higher risk of communication failures. Currently, communication engineers at Base Stations (BS) and Mobile Switching Centres (MSC) primarily rely on traditional path loss calculation methods that focus on signal properties. However, these conventional techniques are becoming inadequate, particularly with the complexity and high frequency demands of 5G networks. They struggle with challenges such as synchronization errors, increased system complexity, and higher error rates. To address these limitations, a novel solution is proposed: leveraging machine learning to predict path loss more accurately. The foundation of this approach is the development of a "Path Loss Propagation" (PLP) dataset, which will be used to train machine learning models. The key goal is to build a system that can effectively predict future path loss values based on the input features within the PLP dataset. Python has been chosen for implementation due to its powerful ecosystem and rich set of libraries for machine learning. This solution aims to significantly improve the accuracy and efficiency of path loss prediction, ultimately supporting the development of more reliable and resilient wireless communication networks

Keywords : Path Loss Prediction, 5G Networks, High-Frequency Bands, Mobile Switching Centres, Propagation Loss Estimation

Author : G. Shanmugavel, Nune Pravallika, Kaluvai Yaswini, Naraga Hemapriya, Alahari Venkata Jyothirmayi

Title : PREDICTIVE PATH LOSS MODELING IN 5G NETWORKS USING MACHINE LEARNING TECHNIQUES

Volume/Issue : 2025;02(04)

Page No : 123-131

Traffic rule violations such as overspeeding and signal jumping are major contributors to road accidents and traffic congestion. This project introduces an AI-powered system designed to detect and respond to such violations in real-time. The system combines various hardware components, including an RPS (Regulated Power Supply) for power, an RFID module for vehicle identification, an ESP32 camera for capturing images, and IR sensors (IR1 and IR2) for detecting vehicle movement. The ESP32 microcontroller processes the collected data and communicates with output devices such as an LCD display, LED indicators, a GSM module, a buzzer, and an IoT module. It also integrates with Google Cloud for real-time monitoring and data storage. When a traffic violation is detected, the ESP32 camera captures an image of the offending vehicle, and the RFID module retrieves its identification details. Alerts are then triggered through LEDs, buzzers, and GSM-based notifications sent to traffic authorities or the vehicle owner. Additionally, the collected data is stored in Google Cloud for future analysis, traffic management, and law enforcement purposes. The IoT integration enables remote monitoring, making this AI-driven system an effective tool for enhancing road safety and traffic regulation.

Keywords : Traffic Violation Detection, Speed Monitoring System, Signal Jump Detection, Esp 32, Google Cloud.

Author : M. K. Chakravarthi, Gunji Vijayaaditya, Gandavarapu Adithya, Kethireddy Mohith Reddy

Title : REDLIGHT SHIELD: SMART AI SYSTEM FOR SIGNAL AND SPEED VIOLATION CONTROL

Volume/Issue : 2025;02(04)

Page No : 132-139

Medical imaging tools play a critical role in the early detection and treatment monitoring of lung cancer. Various imaging modalities, including chest X-ray, magnetic resonance imaging (MRI), positron emission tomography (PET), computed tomography (CT), and molecular imaging techniques, have been widely utilized for lung cancer diagnosis. However, these techniques have limitations, such as the inability to automatically classify cancerous images, making them less suitable for patients with co-existing pathologies. Therefore, there is an urgent need for a highly sensitive and accurate method for early lung cancer diagnosis. Deep learning has emerged as a rapidly advancing field in medical imaging, with promising applications across image-based and texture-based data analysis. Utilizing deep learning-based tools enables clinicians to detect and classify lung nodules more efficiently and with greater accuracy. This study proposes an advanced Convolutional Neural Network (CNN) model for detecting lung cancer from chest CT scan images. The proposed model classifies CT images into four categories: adenocarcinoma, large cell carcinoma, squamous cell carcinoma, and normal tissue. It demonstrates superior accuracy compared to a traditional machine learning approach, the Naive Bayes Classifier (NBC). Additionally, the evaluation metrics confirm the effectiveness of the proposed deep CNN model in supporting clinical experts for improved diagnostic outcomes.

Keywords : Early Lung Cancer Diagnosis, Computed Tomography (CT), Computer-Aided Diagnosis, Adenocarcinoma

Author : G. Suresh, Karikati Teja Sai Krishna, Meduru Harish, Potteti Praveen

Title : DEEPLUNGSCAN: EARLY DETECTION OF LUNG CANCER USING ENHANCED CNN-BASED CT IMAGE ANALYSIS

Volume/Issue : 2025;02(04)

Page No : 140-150

Orthogonal Time Frequency Space (OTFS) modulation is a promising technique known for its robustness in time-varying channels, making it well-suited for high-mobility communication environments. This project focuses on the design and hardware implementation of a low-complexity Zero Forcing (ZF) equalizer for a Single-Input Single-Output (SISO) OTFS system. To reduce computational complexity and hardware overhead, we propose a Very Large Scale Integration (VLSI) architecture that incorporates parallel processing and resource optimization techniques. A key feature of the design is the use of a back-to-back Fast Fourier Transform (FFT) and Inverse FFT (IFFT) structure, which simplifies matrix inversion operations and allows for efficient updates without significantly degrading performance. The architecture is carefully crafted to support real-time processing while satisfying area and latency requirements. Through extensive simulations and evaluations, we study the trade-offs between reduced complexity and equalization performance. Performance metrics such as Bit Error Rate (BER), latency, and hardware area are used to evaluate the design. Synthesis results on a Xilinx 7vx485tffg1157-1 FPGA demonstrate the effectiveness of the proposed architecture, achieving a latency of 440 ns at a 100 MHz clock frequency, utilizing 249,843 Look-Up Tables (LUTs) and 74,611 Flip-Flops (FFs).

Keywords : Zero Forcing Equalizer, Orthogonal Time Frequency Space (OTFS), FFT/IFFT-Based Processing, Low-Complexity Equalization

Author : Ch. Suresh, Pesala Venkata Sai Ganesh, Suram Naga Mukesh Reddy, Pambakula Manohar

Title : EFFICIENT VLSI ARCHITECTURE FOR OTFS MODULATION WITH ZERO-FORCING EQUALIZATION

Volume/Issue : 2025;02(04)

Page No : 151-157

The IoT-Enabled Urban Utility Management System represents a transformative step forward in the management of essential city infrastructure. By integrating advanced Internet of Things (IoT) technologies, this system enables efficient monitoring and control of key urban utilities such as street lighting, water pumps, and drainage systems. At the heart of the system lies the ESP32 microcontroller, a powerful and flexible platform perfectly suited for IoT applications. Acting as the central processor, the ESP32 coordinates data collection and executes control commands across the system. To deliver its comprehensive functionality, the system employs a variety of sensors and modules. A Real-Time Clock (RTC) is included to maintain precise timing, essential for scheduled operations like automatically turning street lights on and off. GPS modules are used to provide accurate location data for each utility asset, which supports asset tracking and simplifies maintenance. Light Dependent Resistors (LDRs) are used to monitor ambient light levels, enabling the automatic adjustment of street light brightness to conserve energy. Additionally, ultrasonic sensors are deployed to detect conditions such as water levels in storage tanks or blockages within drainage systems, enhancing the system’s responsiveness and reliability. The primary objective of this project is to create a smart, responsive, and sustainable urban utility management system. By leveraging IoT technology, the system aims to improve infrastructure efficiency, lower energy consumption, and optimize the use of resources. In doing so, it contributes to the sustainability of urban operations and enhances the reliability of essential services, ultimately benefiting city residents through more consistent and intelligent infrastructure performance.

Keywords : Ultrasonic Sensor, Light Dependent Resistor (LDR), GPS Module, Smart Grid Integration, LCD.

Author : G. Tejaswi,L. Naveen, Perisetla Joshika, Pinniboyina Sai Sirisha, Konduru Sruthi, Sirimamilla Yagnadeepika

Title : ECOURBAN: SUSTAINABLE UTILITY MANAGEMENT THROUGH IOT INNOVATION

Volume/Issue : 2025;02(04)

Page No : 158-163

Traditional logistics and container monitoring systems are lacking real-time tracking and facing issues like data breaches and inefficiency. The logistics industry adopts cloud-based solutions for scalability, real-time data access, and efficiency. This project proposes a robust framework for cloud-based logistics, integrating IoT to address modern challenges. This project consists of a fingerprint sensor for securing the container, a GPS for live tracking, a temperature and humidity sensor for monitoring the inside environment of the container, a vibration sensor for Accident detection, and a gas sensor for detecting gas leakage in the container. These sensors are integrated into a microcontroller and connected to IoT. This project introduces a secure, scalable, and interoperable cloud-based logistics framework. It ensures data integrity, enables real-time tracking, and enhances. The system aims to reduce costs, optimize delivery efficiency, and provide a seamless experience for stakeholders.

Keywords : Cloud Computing, Buzzer, DC fan, DC motor, Vibiration senors,

Author : Dr. Ch. Amarnatha Sarma, Pathan Ismail, Valleti Prabhas, Shaik Anas

Title : CLOUD BASED ROBUST SECURITY FRAMEWORK FOR LOGISTIC SURVEILLANCE

Volume/Issue : 2025;02(04)

Page No : 164-170

The ultimate vision of 6G technology is such a network, which is capable of automatically operating and smartly taking over everything from the radio signals to the core cloud. In this paper, we suggest an innovative, human-like method to create that future: a layered AI agents team. Just think of a pyramid-like structure where the lowest agents deal with the very urgent matters such as spectrum allocation, while the higher-level manager agents are in charge of the long-term goals such as quality of service from end to end. These AI colleagues will learn to collaborate with each other by negotiating and exchanging information to make the best possible decisions. We lay out the design for this agentic structure, discussing how it permits the performing of seamless, self-directed orchestration. This method has the potential to deliver not only the 6G networks that are immensely more efficient and prompt but also the ones that are completely self-driving.

Keywords : The ultimate vision of 6G technology is such a network, which is capable of automatically operating and smartly taking over everything from the radio signals to the core cloud. In this paper, we suggest an innovative, human-like method to create that future: a layered AI agents team. Just think of a pyramid-like structure where the lowest agents deal with the very urgent matters such as spectrum allocation, while the higher-level manager agents are in charge of the long-term goals such as quality of service from end to end. These AI colleagues will learn to collaborate with each other by negotiating and exchanging information to make the best possible decisions. We lay out the design for this agentic structure, discussing how it permits the performing of seamless, self-directed orchestration. This method has the potential to deliver not only the 6G networks that are immensely more efficient and prompt but also the ones that are completely self-driving.

Author : Bhaskara Raju Rallabandi

Title : AI-NATIVE 6G: AUTONOMOUS HIERARCHICAL AGENTIC RAN AND CORE ORCHESTRATION

Volume/Issue : 2025;02(04)

Page No : 171-177

This study examines differences in school enrolment and school choice across the 33 districts of Telangana using UDISE+ data for 2024–25. The results show a clear regional imbalance, with 38% of students concentrated in Hyderabad, Rangareddy, and Medchal-Malkajgiri, where private schools account for most enrolment. In rural and tribal districts, government and welfare schools serve the majority of students. The findings suggest that urbanisation, household income, and school availability influence the choice between private and public education. The study also highlights the importance of welfare schools in improving access in less-developed districts and suggests ways to make public education spending more effective.

Keywords :

Author : Dr. Chanda AppaRao

Title : REGIONAL INEQUALITY IN SCHOOL ENROLMENT IN TELANGANA: AN ECONOMIC ANALYSIS OF DISTRICT-WISE EDUCATIONAL ACCESS IN 2024–25

Volume/Issue : 2025;02(04)

Page No : 178-187

To address the challenges of few-shot aerial image semantic segmentation, where unseen-category objects in query aerial images need to be parsed with only a few annotated support images, we propose a novel approach by integrating a U-Net architecture with Efficient Net. Typically, in few-shot segmentation, category prototypes are extracted from support samples to segment query images in a pixel-wise matching process. However, the arbitrary orientations and distribution of aerial objects in such images often result in significant feature variations, making conventional methods, which do not account for orientation changes, ineffective. The rotation sensitivity of aerial images causes substantial feature distortions, leading to low confidence scores and misclassification of same-category objects with different orientations. To overcome these limitations, we propose an enhanced solution by combining U-Net for robust semantic segmentation with Efficient Net for efficient feature extraction and scale adaptability. This architecture, which we refer to as Efficient U-Net, introduces rotation-invariant feature extraction to handle the varying orientations of aerial objects. By leveraging Efficient Net’s scalable Convolutional layers for feature extraction, we ensure that the network can capture orientation-varying yet category-consistent information from support images. This approach enhances the segmentation accuracy by aligning same-category objects, irrespective of their orientation, thereby minimizing the oscillation of confidence scores and improving the detection of rotated semantic objects. This Efficient U-Net model provides a scalable, rotation-invariant solution to the few-shot segmentation of aerial images.

Keywords :

Author : Shaziya Noorien, Syeda Muskaan Begum, Syeda Zehra Fatima Razvi, Ms. Sadaf Jahan

Title : AERIAL IMAGE SEMANTIC SEGMENTATION USING U-NET AND EFFICIENTNET

Volume/Issue : 2025;02(04)

Page No : 188-194