Evaluating Hardware Platforms and Path Re-planning Strategies for the UAV Emergency Landing Problem

Abstract

The present paper evaluates some strategies to apply path re-planning algorithms for emergency landing of Unmanned Aerial Vehicles (UAVs). The strategies proposed are based on executing only one path re-planning algorithm or a combination of them using an ensemble approach. The proposed methods are integrated to the security supervision system of the UAV, namely In-Fly Awareness (IFA), where decision-making algorithms are applying after a critical situation happens. The critical situations considered are battery overheating and engine failure. The performance of the proposed strategies for emergency landing is evaluated in a real world scenario using the Software-In-The-Loop (SITL) technique. Computational results reported show that these strategies are promising based on the success rate to land the UAV safely.