Abstract: Artificial intelligence and machine learning are rapidly gaining traction and making an impact in fusion energy research. However, the use of AI more widely in fusion is often prevented by challenges such as the lack of skills and trust. digiLab, a UK-based small company is aiming to address these challenges by using simple, lightweight, and trustworthy probabilistic methods such as Gaussian Processes. Though digiLab works in fusion, they also apply these methods to other data-sparse technical industries which place an importance on trust; including renewables, nuclear, and aerospace. In this talk Dr Cyd Cowley will present work from digiLab, covering a wide range of the organisation's work in fusion, from surrogate modelling of turbulence in spherical tokamaks to AI-enhanced optimisation of integrated diagnostics in fusion power plants. Much of this work stems from collaborative research between digiLab and scientists at the UK Atomic Energy Authority. 

Sponsorship of an event does not constitute institutional endorsement of external speakers or views presented.