I am delighted to have been part of a team writing a forward-looking Perspective on the future of data-driven Li-ion battery research with focus on improved safety. The Perspective was a collaboration with co-authors from NREL, MIT, Tsinghua, and Imperial College and is now available in the academic journal Joule. The full text can be accessed at The Application of Data-Driven Methods and Physics-Based Learning for Improving Battery Safety.

One of the most intriguing sections (Practical Implementation for Commercial Applications) takes inspiration from the field of robotics and automation research where a neural network is used to combine data from simulation and real-world experiments to achieve high accuracy with comparatively small datasets. Further resources for a deeper dive into this approach1 is available through the KTH twitter account.


  1. Martin Hwasser, Danica Kragic, Rika Antonova “Variational Auto-regularized Alignment for Sim-to-Real Control” ICRA2020