This paper introduces Swift, an autonomous drone racing system that achieved world-champion-level performance against three human champions in head-to-head races. Swift combines deep reinforcement learning (RL) in simulation with real-world data to overcome the challenges of high-speed flight and onboard sensor limitations. The system's success represents a significant milestone for mobile robotics and machine intelligence.
Publisher
Nature
Published On
Aug 30, 2023
Authors
Elia Kaufmann, Leonard Bauersfeld, Antonio Loquercio, Matthias Müller, Vladlen Koltun, Davide Scaramuzza
Tags
autonomous drones
drone racing
deep reinforcement learning
mobile robotics
machine intelligence
high-speed flight
sensor limitations
Related Publications
Explore these studies to deepen your understanding of the subject.