“Artificial Intelligence techniques for GNC design, implementation and verification

Background and justification: ​Artificial intelligence (AI) techniques have developed into a transformative force across industries. In GNC systems for space applications, it presents significant untapped potential. The AI4GNC project is a explorative study consisting of research (knowledge generation) and preliminary industrialization feasibility assessment (knowledge integration) of some artificial intelligence techniques for Guidance, Navigation, and Control systems.​

Achievements and status: ​The field of AI technologies is vast and this activity focused on three major directions and techniques ​Automated GNC tuning with Bayesian optimization (BO) ​AI techniques allow for efficient automated exploration and black-box optimization of GNC software using Monte Carlo style simulation.​Guidance through robust trajectory optimization ​On-board optimization of guidance trajectories allow for significant performance improvements while increasing design flexibility and transparency. Robust control techniques can furthermore be used to effectively robustify trajectories on guidance level allowing for safe operation using system models learned from data.​AI-enhanced system identification ​Generate high-performance system models outside the range of validity of linear models.​