John Florence Gnana Poovathy

In a world actively moving towards sustainable growth, the efficient management of Battery Management Systems (BMS) in Electric Vehicles is critical. The precise estimation of essential parameters such as State of Charge (SoC), State of Health (SoH) and State of Power (SoP) play a pivotal role in the effective management of batteries through timely decision making by battery management systems (BMS). There are many machine learning (ML) base models to predict health, energy, power and other battery parameters. The proposed work combines predictive modeling and time series analysis with an extended Kalman filter to develop a framework for SoH, SoC and SoP prediction. This includes long short term memory (LSTM) based machine learning (ML) algorithms. In this article it is demonstrated that the suggested approach has better accuracy and resilience against modeling errors. The main finding of this study is that combining predictive modeling and time series analysis with extended Kalman filter can increase the accuracy of estimation of state variables and other essential parameters to more than 88% for various temperatures such as 0°C, 25°C, and 45°C that would suit most of the tropical countries.

Conventional door locks are easily bypassed, posing security threats. Recent IoT advancements have led to smart door locks, but many rely on single authentication methods. This work presents an IoT-enabled multi-factor authentication door lock system, enhancing security with RFID, numeric keypad, and access control via an ESP32 microcontroller and Arduino Uno board. It requires two out of three methods: PIN, facial recognition, and RFID tag verification, unlocking within 10 seconds and re-locking in 5-6 seconds. The average success rate in face recognition with and without disturbances such as spectacles, facial hair etc., are 97% and 92% respectively. Features include robust encryption, secure communication, and fail-safe mechanisms, with a modular design for smart home integration. Thoroughly tested, the system outperforms existing solutions in accuracy and flexibility, providing enhanced security through multi-factor authentication

Typically, drones are utilized in locations where people are unable to carry out essential activities. This article discusses the feasibility study of a drone that can move in both air and water. There are several drones that can fly, robots that can swim, and robots that can move on the ground, among other things. Such multimodal drones may find use in military and agricultural settings as a result of this concept. The multimodal drone's operation in both air and water is examined in this paper through thorough explanations and simulation. The theoretical viability of a drone has been shown by the simulation findings. The drone used a calculated thrust of 51.03N to achieve steady flying in the air and 673.39N to sustain regulated movement underwater. A calculated drag force of 0.765N in air and 623.125N in water, which highlights the difference in operating conditions, is one of the intriguing insights into the drone's performance that the study also uncovers. Furthermore, the analysis shows that when completely submerged in water, the buoyancy force is 45.07N.

Robots have become a priceless remedy that can be deployed in places and applicative areas where humans have restrictions and risks. For example, painting of hi

Image fusion can be described as the process of combining multiple input images for obtaining a single enhanced image. It is not only constructs the output imag

This research article is aimed at improving the efficiency of a computer vision system that uses image processing for detecting cracks. Images are prone to noise when captured using drones or under various lighting conditions. To analyze this, the images were gathered under various conditions. To address the noise issue and to classify the cracks based on the severity level, a novel technique is proposed using a pixel-intensity resemblance measurement (PIRM) rule. Using PIRM, the noisy images and noiseless images were classified. Then, the noise was filtered using a median filter. The cracks were detected using VGG-16, ResNet-50 and InceptionResNet-V2 models. Once the crack was detected, the images were then segregated using a crack risk-analysis algorithm. Based on the severity level of the crack, an alert can be given to the authorized person to take the necessary action to avoid major accidents. The proposed technique achieved a 6% improvement without PIRM and a 10% improvement with the PIRM rule for the VGG-16 model. Similarly, it showed 3 and 10% for ResNet-50, 2 and 3% for Inception ResNet and a 9 and 10% increment for the Xception model. When the images were corrupted from a single noise alone, 95.6% accuracy was achieved using the ResNet-50 model for Gaussian noise, 99.65% accuracy was achieved through Inception ResNet-v2 for Poisson noise, and 99.95% accuracy was achieved by the Xception model for speckle noise.