Can an AI-controlled racing drone beat the world's best human pilots?

Racing drone

The Drone Racing League's 2018 season is underway, and to celebrate, the league has teamed up with Lockheed Martin to set a challenge: build an AI that's not only capable of racing a tiny quadcopter alongside the world's best human pilots, but can beat them too.

The AlphaPilot Innovation Challenge invites students, industry experts and drone enthusiasts to design an AI framework powered by Nvidia's Jetson platform. The AI must be capable of flying a standard racing drone through several real drone racing courses with no pre-programming or human intervention. The winning teams will receive over $2 million in prizes.

Flight club

The challenge isn't just for fun – AI-controlled drones capable of navigating complex environments at speed would have dozens of real world uses.

"At Lockheed Martin, we are working to pioneer state-of-the-art, AI-enabled technologies that can help solve some of the world's most complex challenges – from fighting wildfires and saving lives during natural disasters to exploring the farthest reaches of deep space," said Keoki Jackson, chief technology officer at Lockheed Martin.

"Now, we are inviting the next generation of AI innovators to join us with our AlphaPilot Innovation Challenge. Competitors will have an opportunity to define the future of autonomy and AI and help our world leverage these promising technologies to build a brighter future."

Entries open in November. For more information and to learn how to take part, check out Lockheed Martin's competition page.

Cat Ellis

Cat is the editor of TechRadar's sister site Advnture. She’s a UK Athletics qualified run leader, and in her spare time enjoys nothing more than lacing up her shoes and hitting the roads and trails (the muddier, the better)