 Abstract photovoltaics and Li-ion batteries are becoming increasingly popular sources of renewable energy. However, due to their intermittent nature, it is important to ensure the safety and longevity of the batteries used in conjunction with them. This paper proposes a novel diagnostic methodology based on machine learning algorithms that can be used to detect potential problems with the batteries before they become serious. The algorithm was tested on synthetic data generated by a solar panel in Hawaii and found to have an accuracy of 98%. This article was authored by Matthew Jubry, Nawel Costa, and Dax Matthews.