AI Drone Beats Human World Champions on Abu Dhabi Race Course

An artificial intelligence system has flown a racing drone through a championship course faster than three former human world champions.

The system, called MonoRace, was built by researchers at Delft University of Technology's Micro Air Vehicles Lab in the Netherlands. It relies on a single forward-facing camera, an inertial measurement unit and a small onboard computer, with no GPS, motion-capture system, stereo cameras or preloaded course map.

At the 2025 Abu Dhabi Autonomous Racing League drone racing competition, MonoRace beat every rival autonomous team, completing two laps around 11 gates in 16.56 seconds and reaching about 63 mph (102 km/h). The drone weighs 966 grams and measures roughly 11.8 x 11 x 3.9 inches.

The findings were published Oct. 14 in Science Robotics. The model was trained using reinforcement learning in simulation, with researchers varying factors such as weight, voltage, drag and sensor noise so the network could control different drones. Co-author Guido de Croon said the work shows a robot can perceive, decide and act at high speed with limited sensing and computing power, and suggested such networks could one day be tuned for energy efficiency in everyday robotics.

Source: Llive Science