greenSPEED at the Oxford Battery Modeling Symposium 2025

24. & 25. July 2025 | Oxford, England

At the Oxford Battery Modeling Symposium 2025, our colleague Florian Feyersinger from Virtual Vehicle presented a poster introducing a hybrid-model digital twin for silicon battery production. This approach enables fast and realistic battery simulations.

Silicon anodes are widely regarded as a promising step forward in energy storage thanks to their higher capacity. However, they also present significant challenges, including voltage hysteresis and accelerated degradation. To address these issues, our approach combines machine learning with physics-based models. This reduces error margins and improves the accuracy of real-world battery performance predictions.

Early results are promising: a clear reduction in voltage error for non-silicon cells has already been observed within trained data ranges, and simulations for silicon-based cells are now underway.

Why this matters: Better models help us understand battery behaviour more accurately, enabling the design and use of batteries that last longer, perform better, and reduce environmental impact by minimizing waste, extending lifespan, and guiding more sustainable materials and recycling strategies.

This research, carried out together with Frank Sehnke (ZSW), highlights the value of EU-funded projects such as greenSPEED in fostering collaboration across institutions and disciplines to address complex battery challenges.