Technology Outline (Process Description)

The battery health monitoring system consists of an image capturing unit, algorithms for system regulation, and a data-driven model for fast predicting the State of Health (SoH) of batteries, fault and uneven aging in the cells. The image capturing unit captures color and thermal images of the batteries. Color images are processed to detect battery locations, while continuous thermal images are analysed to extract thermal features such as mean, standard deviation, skew, entropy, energy, and kurtosis. Algorithms have been developed to process color and thermal images, including thresholding, contour detection, and image segmentation. The data bank generation for the predictive model is involves charging batteries with known SoH using a predefined current for a specific time and storing thermal features at fixed intervals, and these are used to train the predictive model for determining battery health status.