The main supervisor of this project is Daniel Axehill, professor at Linköping University. The overall purpose of this project is to combine methods from both artificial intelligence and optimal control into more powerful approaches, combining the strengths of the respective approaches. One of the key challenges of this project is to provide some level of guarantee of safety such that the methods developed are possible to use on a safety-critical platform. The candidate AI approaches to be used are reinforcement learning and generative AI/foundation models. Having access to computational resources would help significantly towards verifying and/our developing our approaches.