Developing intelligent Unmanned Aerial Vehicles (UAVs) entails more than ensuring safe navigation; it also involves equipping them with high-level decision-making abilities through AI planning. Simulation-based testing is a necessary gateway in transitioning research works into operational real-world UAV missions. We demonstrate Dynamic REAP1, a simulation environment that considers accurate physics of the vehicle, while providing georeferenced semantic segmentation and tools to include dynamic background characters, removing thereby the assumption of a static world when used for validating planning and acting methods. A system demonstration based on a Search and Rescue (SAR) scenario can be found at: https://www.youtube.com/watch?v=mthOS7hksZQ     «

Developing intelligent Unmanned Aerial Vehicles (UAVs) entails more than ensuring safe navigation; it also involves equipping them with high-level decision-making abilities through AI planning. Simulation-based testing is a necessary gateway in transitioning research works into operational real-world UAV missions. We demonstrate Dynamic REAP1, a simulation environment that considers accurate physics of the vehicle, while providing georeferenced semantic segmentation and tools to include...     »