BMW and FSB students from Zagreb are developing an AI model for batteries

The artificial intelligence network uses data from testing and production, in real time, to “accurately predict process parameters in battery cells and their performance data.”

In this way, the testing phase of battery cells could be significantly shortened, with a smaller number of tests, but also the preservation of quality, according to BMW.

New AI-based systems reduce the amount of material and the time required for individual steps in the process by more than 50 percent, they say.

According to them, predictive models could eliminate the need for “quarantine”.

Newly produced cells are usually stored, after initial charging, in a special storage area under controlled conditions.

INSIGHT’s AI-based systems “can fully analyze battery cells in advance,” which could potentially end the “quarantine” phase in the future, the release noted.

We are currently working on the expanded use of new models, and outside the “prototype environment”, said in a press release the head of the development of lithium-ion battery cell technology of the German company Christian Siedelhofer, noting that, in addition to the production of battery cells, it could also be used in the “broader production system” of the BMW Group.

It is conducted at the CRTA in Zagreb and at the BMW Competence Center for Battery Cells (BCCC) in Munich. The project team includes professors, doctoral students and students of the University of Zagreb and researchers from the BMW Group.