A Fault-Tolerant Automated Flight Path Planning System for an Ultralight Aircraft

The development and integration of fault-tolerant systems has considerably increased flight safety over the years. One of the research areas that has made this improvement possible is the development of more advanced flight guidance systems, that are able to compute feasible flight trajectories in an automated manner, even under non-nominal conditions. However, such highly automated systems are normally not available for low-cost ultralight aircraft, which are usually piloted by non-professional pilots, who may not react properly under adverse circumstances. In this paper, we propose a model-based flight path planning system that uses an automated AI planner. By leveraging the flexibility of the AI planner to adapt to different planning problem models, we integrate "fault-tolerant" capabilities into the planning system. Therefore, optimal control parameters learned for various non-nominal flight conditions can be considered too. Finally, extension tests were performed under a selected number of scenarios to validate the feasibility of the plans.