NEWS AND VIEWS
04 February 2026 Machine learning slashes the testing needed to work out battery lifetimes A machine-learning method that reasons and adapts has been developed to solve one of the most time-consuming bottlenecks in battery development. By Chao Hu 0 Chao Hu Chao Hu is at the School of Mechanical, Aerospace, and Manufacturing Engineering, University of Connecticut, Storrs, Connecticut 06269, USA. View author publications PubMed Google Scholar
Lithium-ion batteries are increasingly used as power sources for diverse applications ranging from electric vehicles to medical devices. However, for any new battery design, it still takes a frustratingly long time to answer one fundamental question: how long will the batteries last? Conventional approaches to answering this question require batteries to be repeatedly cycled (charged and discharged) until they reach an end-of-life threshold, a process that can take months or even years. Writing in Nature, Zhang et al.1 present a machine-learning method that short-circuits this waiting game, predicting battery lifetime with an accuracy that is sufficient for real-world use in one week or less — even for designs that the model has never encountered.
Nature 650, 41-42 (2026)
doi: https://doi.org/10.1038/d41586-026-00168-w
References Zhang, J. et al. Nature 650, 110–115 (2026). Ramadass, P., Haran, B., Gomadam, P. M., White, R. & Popov, B. N. J. Electrochem. Soc. 151, A196 (2004). Ramadesigan, V. et al. J. Electrochem. Soc. 158, A1048 (2011). Birkl, C. R., Roberts, M. R., McTurk, E., Bruce, P. G. & Howey, D. A. J. Power Sources 341, 373–386 (2017). Shen, S., Sadoughi, M., Li, M., Wang, Z. & Hu, C. Appl. Energy 260, 114296 (2020). Figgener, J. et al. Nature Energy 9, 1438–1447 (2024). Severson, K. A. et al. Nature Energy 4, 383–391 (2019). Download references
Competing Interests The author declares no competing interests.
Related Articles
Subjects