Project description
Current methods to evaluate Li-ion batteries safety, performance, reliability and lifetime represent a remarkable resource consumption for the overall battery R&D process. The time or number of tests required, the expensive equipment and a generalised trial-error approach are determining factors, together with a lack of understanding of the complex multi-scale and multi-physics phenomena in the battery system. Besides, testing facilities are operated locally, meaning that data management is handled directly in the facility, and that experimentation is done on one test bench.
The FASTEST project aims to address these challenges by developing and validating a fast-track testing platform that integrates a strategy based on Design of Experiments (DoE) and robust testing results, combining multi-scale and multi-physics virtual and physical testing.
Among other contributions, Fraunhofer IISB leads the work package “Design of Experiments, boundary conditions and methodologies” which aims to provide innovative model-based DoE to accelerate and improve the battery testing process and to smartly combine physical and virtual testing for optimal time and cost reduction. To this end, the AI-augmented simulation group at Fraunhofer IISB leverages Physics-Informed Neural Networks (PINN) to accelerate and improve DoE methodologies which offer a pathway to the reduction of time and costs in battery testing.
Publications
Parametrized Physics-informed Deep Operator Networks for Design of Experiments Applied to Lithium-Ion-Battery cells
by P. Brendel et al., Journal of Energy Storage 128 (2025)
Fraunhofer Institute for Integrated Systems and Device Technology IISB
