Biometric-based access control systems are gaining popularity over traditional security methods such as keys orpersonal identification numbers (PINs), offering stronger protection for homes and organizations as a safer and smarter alternative. This research presents a sequential fingerprint authentication framework in which multiple biometric inputs must be verified in a predefined order to grant access. Unlike conventional single-fingerprint authentication systems, the proposed design transforms biometric authentication into a biometric passcode mechanism, requiring both correct identity and correct order of fingerprints. This additional layer significantly increases resistance to unauthorized access attempts while maintaining a low-cost embedded implementation suitable for practical access control applications. This solution is flexible and affordable, and it can be used in residential, commercial and industrial environments. By combining fingerprint authentication with sequential verification, the proposed system provides a robust, scalable and user-friendly security solution. The integration with Arduino and relay control further enhances its flexibility, making it a practical choice for modern smart security systems.

The surge in digital currencies as alternative investment assets has been accelerated by real-time financial data and the inherent volatility of digital currencies. This study presents Coinmeter, an innovative tool that leverages dynamic visualization and real-time data to enhance digital asset analysis. Coinmeter is built on a robust technology stack, featuring Chart.js for interactive data visualization, Material User Interface (UI) for modern interface design, and React.js for front-end development, creating a user-friendly platform that delivers comprehensive market capitalization and exchange rate data with seamless integration of the CoinGecko Application Programming Interface (API). Coinmeter efficiently retrieves real-time data using effective caching techniques that optimize speed and performance. The modular architecture of Coinmeter’s design ensures scalability and ease of maintenance, making it adaptable to the evolving demands of the virtual currency market. Coinmeter also aligns with the United Nations Sustainable Development Goals (SDGs), especially those focused on financial literacy and innovation. This study demonstrates how Coinmeter functions as both a powerful analytical tool and an educational resource, empowering users to make informed investment decisions in a rapidly changing financial landscape. The results underscore how cutting-edge web technologies can revolutionize financial data analysis and visualization, fostering a more informed and engaged user base in the future.

Accurate detection of tomato ripening stages is crucial for optimizing harvesting, minimizing post-harvest losses, and ensuring consistent quality control throughout the agricultural supply chain. This study presents a real-time method for detecting tomato ripening stages, utilising deep learning through YOLOv8 and the portability of Raspberry Pi. The YOLOv8 model was trained on the Laboro dataset, comprising varied photos of tomatoes at various stages of development. The algorithm accurately categorises tomatoes into three ripening stages: green, half-ripened, and completely ripened, attaining a classification accuracy of 93%. The trained model was implemented on a Raspberry Pi, connected with a Pi Camera to enable live video streaming for real-time detection and classification. A Python-based script was implemented to capture frames and run inference on the Pi, enabling instant categorization of tomatoes based on ripening stages. This low-cost, portable, and energy-efficient solution is designed to support farmers and agribusinesses, especially in rural or resource-limited settings, by automating the process of fruit quality assessment. The system seeks to optimise crop monitoring efficiency, diminish reliance on labour, and refine post-harvest decision making. The proposed approach demonstrates a scalable method for smart agriculture practices using edge AI technologies.

The rapid expansion of 5G networks offers ultra-reliable low-latency communication (URLLC), enhanced mobile broadband (eMBB), and massive machine type communication (mMTC) demands for intelligent and adaptive optimization strategies. This is particularly critical in ultra-dense urban environments, where interference, traffic variability, and energy inefficiencies pose major challenges. This study offers a next-generation optimization framework driven by machine learning. It combines random forest classifiers, long short-term memory (LSTM) networks, and gradient boosting regressors. Together, these models improve key performance indicators, such as the signal-to-interference-plus-noise ratio (SINR), reference signal received power (RSRP), latency, throughput, and energy efficiency. The proposed system demonstrated a strong predictive performance. It achieved an R 2 score of 0.89, mean squared error (MSE) of 0.25, and mean absolute percentage error (MAPE) of 5.20%. Under high load and dense interference conditions, the framework delivered a 22.4% increase in average throughput, a 17.8% reduction in packet loss, and a 19.5% drop in end-toto-end latency. The LSTM model effectively responded to changes in signal strength caused by user mobility. The random forest classifier detects interference with 93.4% accuracy, enabling quick mitigating action. Moreover, the energy efficiency of the proposed system is improved by 18.7%, making it both optimized and cost-effective for better performance evaluation. These outcomes show that future 5G infrastructure depends on artificial intelligence-driven self-organizing, self-optimizing networks. This framework provides a reliable path toward scalable, resilient, and adaptive service delivery in smart cities.

The objective of this work is to propose a low-cost vehicle driver control system that aims to prevent road accidents caused by various factors, such as intoxication, negligence in seatbelt usage, reckless driving, and drowsiness. This system integrates preventive and deductive measures to ensure driver and passenger safety, while providing immediate access to emergency services. The preventive aspect ensures the driver's sobriety using a gas sensor, verifies seatbelt usage with an optical IR encoder, and detects rash while driving through an accelerometer/gyroscope. Additionally, a camera module, as part of the integrated system, identifies drowsy driving and initiates appropriate measures. The deductive segment utilizes a load cell to detect accidents caused by sudden changes in velocity, whereas the Global Navigation Satellite System (GNSS) determines the location of the accident for notification transmission via the GSM module to emergency services. The smooth integration of individual units is facilitated by Arduino microcontrollers. This development endeavours to enhance road safety by implementing proactive accident prevention measures and facilitating swift access to medical assistance during an accident.

The objective of this paper is to develop an AIpowered system for translating natural language into SQL queries, intended to streamline database interactions for nontechnical users. The proposed system generates executable SQL statements from user inputs by leveraging the Gemini API, which naturally facilitates retrieval augmented generation. A Streamlit-based interface enables users to upload structured Excel datasets and execute Create, Read, Update and Delete (CRUD) operations using conversational commands. The process encompasses schema parsing, sample data extraction, and prompt engineering to ease query generation. When analyzed using a student information dataset, the system attained a query generation accuracy of 92% and an execution success rate of 90 %. Use case scenarios such as data insertion, filtering, updating, and deletion illustrate the effectiveness of the model in understanding the user intent. Despite its dependable functioning, issues such as response latency and sporadic misreading of natural language inquiries persist. This study illustrates the practical capabilities of large language models, like Gemini, in connecting natural language with structured data access, thus improving database interactions.

Simultaneous Localization and Mapping (SLAM) is a key technology for autonomous robot navigation in new settings. This project focuses on the development of a SLAM based mapping and navigation system in MATLAB. The system uses MATLAB's Robotics System Toolbox and Computer Vision Toolbox to simulate a mobile robot traversing an environment and creating a map. The SLAM method, which is based on the Extended Kalman Filter (EKF-SLAM) and the Particle Filter, allows for real-time localization and mapping. Autonomous navigation uses path planning techniques such as A* and Dijkstra's algorithm. The simulation results show precise mapping, efficient obstacle avoidance, and optimal path planning. This project offers a low-cost approach to explore SLAM techniques in MATLAB, making it useful for research and instruction in autonomous robots.

The purpose of this work is to design of an artificial pancreas for precise blood glucose insulin regularization for a diabetes patient. The author proposed a continuous Glucose monitoring system. Where Insulin will be injected externally into the patient using an insulin pump known as the artificial pancreas. Bergman's simple model is used in a closed-loop system that employs a PID controller and MATLAB/SIMULINK. This regularization system determines the required amount of external Insulin infusion through the control actions involved according to the meal disorders and other factors involved. The artificial pancreas simulation model includes a continuous glucose monitor, an insulin injection pump, and a potential PID controller. This study uses a combinational mathematical model to provide a simple glucose-insulin monitoring and regularisation method for type 1 diabetes patients. The findings of the SIMULINK simulation have demonstrated that when the intake of glucose during meals is balanced, the body's insulin sensitivity can be stabilized and the hypoglycemia condition effectively regulated. The results show that this model is reliable and useful, and it may be utilized to lay out information for research on blood glucose regulation.

Uninhabited Aerial System (UAS), or drones, are aircraft that could be remotely controlled and managed by a person or have varying techniques, like autopilot support, and even fully autonomous modes that do not require human intervention. In this paper, the drone can be remotely controlled and used for spying, among other things. The Pixhawk flight controller, combined with a transmitter and a receiver to transmit and receive radio signals for the drone’s remote control, makes up the brains of the drone. The main components of this system are accompanied by four propellers for flight. The use of an electronic speed controller (ESC) has been implemented to control and regulate the drone’s speed. Moreover, a lithium polymer battery has been used to power up the drone. As previously indicated, the installation of the ESP Camera module to this drone has been implemented, which will be utilized for live footage taken throughout its flight and to be able to relay that footage to the user. The footage is analyzed for image processing and identifying objects in the video using the latest You Only Look Once (YOLO) algorithm as surveillance is the primary function of this system.

To ensure data security and safeguard sensitive information in society, image encryption and decryption as well as pixel data modifications, are essential. To avoid misuse and preserve trust in our digital environment, it is crucial to use these technologies responsibly and ethically. So, to overcome some of the issues, the authors designed a way to modify pixel data that would hold the hidden information. The objective of this work is to change the pixel values in a way that can be used to store information about black and white image pixel data. Prior to encryption and decryption, by using Python we were able to construct a passcode with hand gestures in the air, then encrypt it without any data loss. It concentrates on keeping track of simply two pixel values. Thus, pixel values are slightly changed to ensure the masked image is not misleading. Considering that the RGB values are at their border values of 254, 255 the test cases of masking overcome issues with the corner values susceptibility.

Rainfall prediction is vital for various applications such as agriculture, flood control, and water resource management. Accurate rainfall prediction can help farmers to make decisions about irrigation and crop selection. It can also help authorities to take necessary measures to prevent or mitigate the impact of floods. In the proposed study, multiple algorithms are examined and compared in terms of how accurately they predict rainfall. The accuracy of rainfall forecasts may be increased by using machine learning algorithms, which have shown considerable promise in predicting rainfall patterns. The authors solely used the AUC score and categorial standards to analyze six ML algorithms. Based on their accuracy score, AUC, precision, recall, and f1 score, authors compared algorithms including Random Forest, K-Nearest Neighbours (KNN), Logistic Regression, Cat Boost, Naive Bayes, and Gradient Boost. Since they provide a useful solution for categorization and prediction of the amount of rainfall as well as precise rate prediction, the applied algorithms are supervised learning algorithms.

This paper deals with the development of a Simulink model for the control, protection and monitoring system of a diesel engine. Various sensors and switches involved in the engine protection and monitoring systems such as throttle position sensor (TPS), lube oil pressure sensor (OPS), coolant temperature sensor (CTS) and idle validation switch (IVS) are discussed in detail. Then, a comparison study of the speed control of the diesel engine was done using P, PI and PID controllers. Finally, a mathematical model of the diesel engine is derived using the System Identification Technique. The results obtained by formulating an accurate mathematical model and a precise controller allow the operator to establish the required engine protection parameters such as the coolant temperature, oil pressure and the speed of the engine.

Twenty one genotypes were grouped into eight clusters based on D2 values for 11 characters. From the inter cluster D2 values of the eight clusters, it was found that the highest divergence occurred between cluster II and VI followed by cluster I and V, cluster IV and VIII, cluster V and VI, cluster V and VIII and cluster IV and V suggesting that the crosses involving varieties from these clusters would give wider and desirable recombination. The variety PTB-9 from cluster VIII, Swarnaprabha from cluster VI and Uma from cluster I having high mean values for grain yield per plant may be directly used for adaptation or may be used as parents in future hybridization programme. Among the 11 quantitative characters studied the most important character contributing to the divergence was filled grains/panicle followed by grain yield/plant, number of grain /panicle, number of spikelets/panicle, number of grains/panicle, plant height and pollen fertility.

In control of mobile robots, precision plays a key role in path tracking. In this paper we have intended to use hybrid stepper motors for precise control of the two wheeled robot. A control algorithm was developed to control the robot along different trajectories. We have found that stepper motors are more accurate for path tracking than normal DC motors with wheel encoders and one can obtain the implicit coordinates of the robot in runtime more precisely. Getting the precise coordinates of the robot at runtime can be used in various SLAM and VSLAM techniques for more accurate 3D mapping of the environment.

This paper examines the performance of Model Predictive Control (MPC) scheme for an Active suspension. A vehicle suspension is designed to provide superior ride comfort and road handling characteristics. Unlike passive suspensions, the Active suspension can change the dynamic of suspension in real-time by injecting force into the system. MPC allows the active suspension to provide better and consistent passenger comfort and road handling capabilities for different road profile. Even though long back, the idea of active suspension conceived, the prohibitive cost and complexity restricted its usage. In recent years active suspension is receiving more and more attention with users preferring a high-end car. In an active suspension for the real-time adjustment of the control force, need a design of a controller. In literature, many controllers used such as Proportional Integral Derivative (PID), Linear Quadratic regulators (LQR), Fuzzy logic controller, Artificial Neural Networks (ANN). In this paper, revealed a model predictive control arrangement for Active suspension model. MPC is an optimal control scheme which uses a model of plant for predicting the future output. The control inputs are optimized such that these predicted outputs meet the desired level of performance. Tested the MPC control scheme is using a bench-scale replica of Quarter active suspension model from QUANSER. To better appreciate the capabilities of the MPC Control Scheme, compared the performance of the active suspension with that of an LQR control scheme, and passive suspension.

Our project involves the generation of a path for modern driver assistant systems. It provides a cognizance of the objects ahead of the driver, which can play a major role in preventing accidents. Our algorithm is inspired from the data sets involving displacement and time. This provides the accurate position and velocity of the vehicle. To define the trajectory, polynomial equations can be used to explain this. The velocity and acceleration can be calculated according to coefficients of the polynomial equation. The number of coefficients determines the degree of the polynomial. By making use of a simulator, the trajectory generated can be studied. The objective of detection and trajectory generation is to provide a system that alerts the driver to the hurdles ahead so he/she is better placed to avoid a collision while the vehicle is moving.

This paper puts forward an adaptive predator-prey optimization algorithm to solve the weight selection problem of linear quadratic control applied for vibration control of vehicle suspension system. The proposed technique addresses the two key issues of PSO, namely (a) the premature convergence of the particles, and (b) the imbalance between exploration and exploitation of the particles in finding the global optimum. The main principle behind this optimization algorithm is that the inertia weight is adaptively updated based on the success rate of the particles to increase the convergence, and the predator-prey strategy is reinforced to avoid the particles getting trapped in a local minimum thereby, guaranteeing convergence of the particles towards the global optimal solution. The convergence of the particles towards the global minimum is guaranteed on the basis of a passivity argument. Moreover, the strength of this new adaptive optimization technique to tune the gains of linear quadratic regulator is validated experimentally on a laboratory scale active vehicle suspension system for improved ride comfort and passenger safety. (C) 2018 Elsevier B.V. All rights reserved.

Twenty one Kerala rice varieties were crossed with four CMS lines in LxT fashion.The hybrid developed from twenty two crosses between identified restorers and 4 CMS lines were evaluated for heterosis in yield and yield contributing traits.Identified promising hybrids were UPRI95-17A x Aiswarya, UPRI95-17A x Neeraja, UPRI95-17A x Remya and CRMS31A x Kanakom based on high mean grain yield per plant and high standard heterosis over standard check Uma with respect to grain yield and yield contributing traits.Identified superior hybrids can be released for commercial purpose after trail.