Predicting the Current and Future State of Batteries using Data-Driven Machine Learning

Machine learning is a specific application of artificial intelligence that allows computers to learn and improve from data and experience via sets of algorithms, without the need for reprogramming. In the field of energy storage, machine learning has recently emerged as a novel approach for battery modelling, not only to determine the current state-of-charge of batteries, but also predict their future state-of-health and remaining useful life. In this review, we first discuss the two most studied types of battery models in the literature for battery state prediction: the equivalent circuit and physics-based models. Based on the current limitations of these models, we showcase the promise of machine learning techniques for fast and accurate battery state prediction, as well as the major challenges involved, especially in highthroughput data generation. In addition, we propose the incorporation of physics and domain knowledge to develop machine learning models that are more explainable and interpretable. Overall, we see data-driven machine learning as a promising modelling technique that can open up new, exciting opportunities in battery manufacturing, usage, and optimization in the future.