Artificial Intelligence as a Catalyst for Fusion Energy
By Alex F. Savin, Harry Westhead, Philip Horton, and Richard A. Johnson
Published in Nuclear Future 21.6
DOI: https://doi.org/10.63198/nuclearfuture216410
Download PDF
SUMMARY
- AI models, such as digital twins and surrogate models,offer significant potential to speed up and improve the design and optimisation of fusion reactors, but must be applied in targeted areas where sufficient high-quality data exists to ensure trustworthy results.
- The scarcity of experimental data in fusion presents challenges for AI implementation; thus, a problem-specific approach is recommended, focusing on well-benchmarked aspects like transport phenomena and adaptive reactor control.
- A symbiotic relationship between the AI and fusion sectors could accelerate innovation in both fields—fusion provides sustainable energy for AI’s growing demands, while AI delivers the advanced modelling and control needed to realise commercial fusion energy.
LEAD AUTHOR
Alex F. Savin works across all stages of the patent life cycle from invention capture and drafting, through prosecution, to opposition proceedings at the EPO. He is experienced in helping start-up companies and SMEs build and manage their patent portfolios and has particular expertise in the nuclear sector,including fusion and power generation, and in the field of optics & photonics. He has an MPhys degree in Physics and a DPhil in Atomic & Laser Physics, both from the University of Oxford. His research focussed on designing and implementing high-power laser experiments exploring inertial confinement approaches to nuclear fusion at major international laboratories dedicated to laser-driven fusion in the US, UK and Europe including the Lawrence Livermore National Laboratory, and the Central Laser Facility.