Smart Predictive Tool for Detection of Lithium Ion Defects

Physical Sciences Inc. (PSI) will develop an algorithm that uses electro¬chemical data to identify defects in lithium-ion cells throughout their entire lifetime. The algorithm will be capable of detecting defect traces in electrochemical data results that are normally imperceptible. The algorithm will rely on a suite of tests that PSI will develop to characterize defects in electrochemical data. PSI, in partnership with Dr. Zenyuk at the University of California Irvine (UCI), will perform 3D-tomogrophy imaging on Li-ion cells with micron level resolution and use corresponding electrochemical data to confirm the algorithms effectiveness. Various cells with and without defects will undergo electrochemical testing to train the algorithm to identify defect fingerprints in data results curves. During the Phase I effort, PSI will focus on delivering an algorithm and test protocol than can accurately predict and identify the defects present within a cell using electrochemical test data collected during cell formation. During the Phase II effort, PSI will improve the functionality of the algorithm while developing a front-end UI for use in a manufacturing setting.