Autonomous Screening and Optimization of Battery Formation and Cycling Procedures

Optimizing battery performance currently relies on empirical testing using arbitrary parameters, under-validated physiochemical models, and limited data analysis of summary trends. The Chueh lab at Stanford utilized closed loop optimization to develop an adaptive optimal experimental design algorithm capable of reducing battery development time by up to two orders of magnitude. For each cell in a 48-cell array, the OED algorithm analyses data from the first 100 cycles to predict the cell lifetime. Once there is enough data to predict with sufficient confidence, the OED determines the next cycling parameters for an individual cell and sends them to the potentiostat before repeating the process. As the 48 cells are tests in parallel, what is typically a 560-day process to evaluate 224 charging protocols only takes 16 days. This technology applies to fast charging, formation protocols, and state-of-health monitoring for overall lower development cost and time.

Stage of Research