Overview
- Researchers at Chalmers University announced on May 20 that their AI charging strategy lengthened simulated lithium‑ion battery life by about 23 percent while keeping fast-charge times nearly the same.
- The system uses reinforcement learning trained in simulation to adjust charging current to each battery’s state of health so it reduces lithium‑plating, the chemical process that speeds capacity loss and can cause short circuits.
- The team says the software can be delivered as a battery management system update without new hardware but requires calibration for different cell chemistries and pack designs.
- All results so far come from simulation models, so the next steps are direct tests on real cells, safety and regulatory checks, and integration trials with vehicle manufacturers before any consumer rollout.
- If validated, the method could cut warranty costs, raise resale values and help high‑use vehicles such as taxis and delivery fleets get more life from batteries while easing demand for critical raw materials.