As part of the “AI-for-Fusion-Engineering” program, the Federal Ministry of Education and Research (BMBF) is funding Proxima Fusion with more than €6.5 million. Project partners include the University of Bonn, the Jülich Research Centre, and the Technische Universität München (TUM). All project partners bring together interdisciplinary expertise in the fields of plasma physics, machine learning, optimization, and computer science.
Together, they aim to develop AI-powered simulation tools that can integrate physical and technical simulations. These tools will optimize critical components of stellarators, such as high-temperature superconducting magnets, materials in contact with fusion plasma, and cooling systems. This technology is intended to provide clean energy from fusion power plants in the future.
Francesco Sciortino, co-founder and CEO of Proxima Fusion, explains:
“Stellarators represent the clearest and most robust path toward commercially viable fusion energy. They can enable stable and continuous energy generation, but their development is complicated due to their 3D geometries. This project will accelerate stellarator development, reduce costs, and improve the reliability and performance of these devices.”
The interdisciplinary collaboration is intended to enable a technological breakthrough in nuclear fusion. On this, Daniel Cremers, holder of the chair for computer vision and artificial intelligence at the TUM School of Computation, Information and Technology, says:
“By applying data-driven optimization techniques, geometric machine learning, and uncertainty-aware surrogate modeling, we can leverage the latest technological advances in stellarator development in scalable computational solutions.”
Zorah Lähner, leading researcher in geometric deep learning at the University of Bonn and the Lamar Institute, sees further potential:
“The tools and technologies developed as part of this project have broader applications beyond fusion energy. The innovative and practical solutions will also be incredibly interesting for complex engineering challenges in aerospace and the automotive industry.”
As the lead partner, Proxima Fusion is responsible for overall project coordination. The startup focuses primarily on the development and design of economically viable stellarator fusion power plants. By the mid-2030s, fusion energy is to be market-ready and integrated into the grid.
