Drone racing with human pilots at the controls has quickly evolved from an underground hobby to a fully fledged professional sport. But for its 2019 season, the Drone Racing League introduced a new type of competition, calling on developers of autonomous drones to pit their creations against one another in a bid to claim a US$1 million cash prize.
The Drone Racing League’s (DRL) Artificial Intelligence Robotic Racing contest is a four-event series that takes place within its regular racing season. Nine teams from around the world took part in the inaugural edition, each developing their own AI software to autonomously pilot the same standard racing drone around each of the four circuits.
Team MAVLab from Delft University unveiled the world’s smallest autonomous racing drone ahead of the competition earlier in the year, and it has now produced some big results. Using the DRL's standard RacerAI drone, the team aircraft clocked a finishing time of 12 seconds in Austin, Texas on Monday, 25 percent faster than the second-placed drone and enough to secure the Artificial Intelligence Robotic Racing (AIRR) Circuit championship for 2019.
In a bonus round, the Team MAVLab drone was then put to the test against one of the Drone Racing League’s professional pilots, Gabriel "Gab707" Kocher. The robot finished second in this instance, trailing Kocher by five seconds, but with the AI-powered drones only in their first season, you suspect we haven’t seen the last of this rivalry.
And this is about more than just bragging rights. Advances in AI software for autonomous aircraft could have benefits extending far beyond drone racing, with these types of vehicles also playing a role in search and rescue operations, package delivery and surveillance, just to name a few applications. Technology that enables autonomous drones to travel faster and better navigate obstacles would be a huge boon to them all.
Source: Drone Racing League via PRNewswire
Update December 11: This article has been updated to clarify that Team MAVLab won the AIRR championship using DRL's standard RacerAIDrone.