Supaero Reinforcement Learning Initiative

The ISAE-SUPAERO Reinforcement Learning Initiative (SuReLI) is a vibrant group of researchers thriving to design next generation AI. We maintain a constructive, lively environment in a human-sized team that ranges from undergrad students to permanent academic staff, focussed on understanding AI and developping solutions that contribute positive impact on science and society. We work on state of the art algorithms, study their properties mathematically and empirically, and apply them to challenging problems to bridge a gap between theory and real applications. We focus on Reinforcement Learning for sequential decision making problems and its links with other disciplines (e.g. Operations Research, Evolutionary Computation, Planning, Optimization). Our applications include: autonomous vehicles, control of optimization processes, rehabilitation exoskeleton control, video games, aircraft landing scheduling, satellite resources planning, sailboat and UAV planning and control.