Learning High Speed Flight in the Wild ( Science Robotics, 2021)

Quadrotors are agile. Unlike most other machines, they can traverse extremely complex environments at high speeds. To date, only expert human pilots have been able to fully exploit their capabilities. Autonomous operation with onboard sensing and computation has been limited to low speeds. Stateoftheart methods generally separate the navigation problem into subtasks: sensing, mapping, and planning. While this approach has proven successful at low speeds, the separation it builds upon can be problematic for highspeed navigation in cluttered environments. Indeed, the subtasks are executed sequentially, leading to increased processing latency and compounding of errors through the pipeline. Here we propose an endtoend approach that can autonomously fly quadrotors through complex natural and manmade environments at high speeds, with purely onboard sensing and computation. The key principle is to directly map noisy sensory observations to collisionfree trajectories in a recedinghorizon fashion. This direct