A long life for batteries

Machine learning makes predictions about the endurance of energy storage devices more reliable

It’s hard to predict how long a lithium-ion battery in an electric car or a cell phone will last. The forecasting methods used today are based on measuring current and voltage during the charging and discharging of the battery, but not all the relevant parameters can be recorded. Now researchers at Argonne National Laboratory at the US Department of Energy have succeeded in predicting the capacities of a wide range of different chemical battery types. They did this by means of a machine learning algorithm. To train the computer program, the scientists collected data from experiments with hundreds of batteries. In a very short time, the simulation provides insights that would take several years with real cycles. Based on the research results, the scientists plan to develop higher-performance battery materials in the future.