An algorithm which I proposed in Physics-based vs. equivalent circuit cell models:

1. Estimate most cell parameters at once.
2. Train a model which estimates cell failure risk using cell parameters and the examples of cell failures induced during cell testing at the end of the manufacturing line.
3. Use the model in the BMS.

I think a model trained on the examples of artificially induced cell failures will work well enough in the field. Li-ion cells are predictable enough for interpolative algorithms to cover most of the ways in which cells can fail. It's highly unpredictable whether any given cell will fail and when, but overall, failures should fall into a limited number of recognisable scenarios.

Estimating the risk of cell failure is an algorithm of Automatic battery problem discovery and Battery safety.

More generally, ‣.

Related:

• Onboard BMS should detect anomalies and report failures itself