This year’s KOMSO Academy was a great success — fully booked, highly interactive, and driven by the exceptional engagement of all participants.
Hosted at the Robert Bosch Research Campus in Renningen, the event brought together experts from mathematics, engineering, computer science, and industry to explore current advances in Deep Learning for partial differential equations (PDEs).
A central highlight was the strong industrial presence:
35 participants from industry joined the Academy, underscoring both the current relevance of the topic and the growing integration of mathematical AI methods into real-world industrial applications.
Hands-on sessions with TorchPhysics allowed participants to experiment with Physics-Informed Neural Networks (PINNs) and advanced operator-learning approaches such as DeepONet, FNO, and PCA-Net. The interactive format fostered lively discussions, critical reflection, and valuable knowledge exchange.
A heartfelt thank you goes to:
the organizing team at the University of Bremen (ZeTeM),
our colleagues at Robert Bosch GmbH,
and all speakers and participants whose contributions shaped the depth and collaborative spirit of the event.
Together, we demonstrated how powerful the connection between mathematics, AI, and industrial innovation has become — and how much potential lies ahead.
We are already looking forward to the next KOMSO Academy and to continuing this shared journey toward a data- and model-driven future.
