Application of deep learning in materials microscopy to evaluate Lithium-Ion batteries

Major advances have been made recently in the field of machine learning (ML) and specifically deep learning following decades of intensive research. The consequence of this sudden progress extends to almost every industry and the field of material science is no exception. Complex relationships can be better understood and interpreted, and structural characteristics tracked over a longer period of time e.g. production variants correlated with microstructural parameters or microstructure deviations determined for the quality evaluation In this presentation we use ML and here specifically algorithms of deep learning for the quality assessment of large prismatic Li-ion batteries. The performance of Li-ion battery is intrinsically linked to the electrode microstructure. Quantitative measurements of key structural parameters will enable optimization as well as motivate systematic numerical studies for the improvement of the battery performance. The aim is to automatically evaluate various cell components (cathode, anode, and separator) geometrically in high-resolution and also to automatically identify various types of defects such as metal particle contamination etc.