US teen scientist develops new way to predict health of aging EV batteries

An 18-year-old student from Palo Alto has developed a model that could help tackle one of the biggest long-term challenges facing electric vehicles: figuring out how healthy an aging battery really is. Colin Jie Chu, a student at The Nueva School in California, developed a framework for estimating the state of health of lithium-ion batteries with a reported prediction error of just 2.36 percent. The project earned Chu recognition as a finalist in the 2026 Regeneron Science Talent Search, one of the United States’ most prominent science and mathematics competitions. A difficult problem hiding inside every EV EV batteries gradually degrade as they are charged, discharged, and exposed to changing temperatures and operating conditions. But determining exactly how much useful life remains inside a battery is not straightforward. A battery management system cannot simply look inside a cell and measure its degradation directly. Instead, engineers must infer its condition from electrical behavior. According to the Society for Science, Chu analyzed data from 22 batteries deliberately aged using electrical signals designed to mimic changing vehicle driving behavior. He then developed a model combining equivalent circuit modeling with machine learning to estimate each battery’s state of health. The resulting framework achieved a reported error margin of 2.36%. Accurately estimating battery health could improve EV safety, maintenance planning, and battery lifespan. Combining physics with machine learning Chu’s approach did not rely entirely on artificial intelligence. Instead, he combined machine learning with an equivalent circuit model. A method that represents the complex electrical behavior of a battery using mathematical components that can be analyzed more easily. According to reports, Chu conducted the research through Stanford University’s Young Investigators Program, working at Professor Simona Onori’s Stanford Energy Control Lab alongside researchers and industry partners. By combining a physics-based battery model with machine-learning regression, the framework was designed to estimate battery health even as operating conditions changed. The research was later presented at the Modeling, Estimation, and Control Conference in Chicago and published in the Journal of The Electrochemical Society. Why knowing battery health matters A lithium-ion battery does not suddenly fail when it begins to age. Instead, its usable capacity and performance gradually decline, affecting an EV’s range and potentially influencing how the battery should be maintained, reused or eventually recycled. Better state-of-health estimation could allow battery management systems to make more informed decisions about charging, maintenance, and a battery’s remaining useful life. However, the 2.36 percent figure should be viewed in context. The model was tested using research data, and further validation across different battery chemistries, ages, vehicle platforms and real-world driving conditions would be needed before it could become a universal battery-life prediction system. Still, the project demonstrates how combining physical models with machine learning could provide a more reliable alternative to approaches that depend entirely on either traditional modeling or data-driven AI. From high school research to advanced battery engineering Chu began the project in 2024 through Stanford’s Young Investigators Program, giving him access to a university research environment while he was still in high school. His work has since expanded into research involving battery “kneepoints,” the stage at which battery degradation can begin accelerating, and lifetime estimation. For EV owners, the question of battery health may eventually become as important as mileage. And if researchers can predict degradation more accurately, they can extend battery life, improve maintenance decisions, and reduce uncertainty around one of the most expensive components in an electric vehicle. Chu’s research is still part of an ongoing scientific effort, rather than a finished technology ready to be installed in millions of cars. But at just 18, he has already contributed to one of the most technically important problems facing the rapidly growing EV industry. Get the latest in engineering, tech, space & science - delivered daily to your inbox.Kaif Shaikh is a journalist and writer passionate about turning complex information into clear, impactful stories. His writing covers technology, sustainability, geopolitics, and occasionally fiction. A graduate in Journalism and Mass Communication, his work has appeared in the Times of India and beyond. After a near-fatal experience, Kaif began seeing both stories and silences differently. Outside work, he juggles far too many projects and passions, but always makes time to read, reflect, and hold onto the thread of wonder.