Safe path planning for UAV urban operation under GNSS signal occlusion risk
Abstract
Flying UAVs or drones in an urban or peri-urban environment is yet quite a challenge. This presentation introduces a concept of safe path planning for UAV's autonomous operation in an urban environment where GNSS-positioning may become unreliable or even unavailable. Given a geo-referenced environment model, it is possible to predict GNSS signal occlusions and hence to generate a map of local quality/availability of its positioning solution. Motivated from this, our main idea is to utilize such sensor availability map in path planning task for ensuring UAV navigation safety. The proposed concept is implemented by a Partially Observable Markov Decision Process (POMDP) model. It incorporates a low-level navigation system for propagating the UAV state uncertainty in the function of the probabilistic sensor availabilities. A goal-oriented version of the Monte-Carlo Tree Search algorithm is applied for offline POMDP solving. Towards online planning, we explore an approach of learning-based context selection to improve planning efficiency.
Biography
Yoko Watanabe received her Ph.D. degree in Aerospace Engineering in 2008 from Georgia Institute of Technology. She joined the applied control research unit at ONERA/DTIS in 2008, where she has been working on the autonomous navigation and guidance systems design for aerial vehicles and their flight testing. Her research activity covers robust navigation and state estimation by multi-sensor fusion, safe path planning problems under uncertainty, and coordination control of multi drones. She was awarded ICAS John J. Green Award in 2020 for her contribution to coordinating the EU-Japan research cooperation in Aeronautics.