The Drone Racing League and its technology partner Lockheed Martin have selected nine teams from top-tier universities and aerospace institutes to participate in its first autonomous racing circuit.
Those picked for the AlphaPilot Innovation Challenge are tasked with developing an artificial intelligence that can pilot high-speed racing drones. At stake is a $1 million cash prize. The teams, comprised of 69 students, drone technologists, aerospace engineers, and coders from around the world, will be pushing the boundaries of autonomous flight.
DRL will provide the drone hardware, while the teams will design an AI framework capable of piloting the drones through aerial courses without any GPS, data relay, or human intervention. Teams will showcase their skills as part of DRL’s first autonomous drone racing series, the Artificial Intelligence Robotic Racing Circuit, which takes off later this year.
The four-event racing circuit, sponsored by Lockheed Martin, will additionally include an additional $250,000 reward for the first team whose autonomous drone can beat a human-piloted drone.
“The combination of robotics and AI will fundamentally change the future of sports,” said DRL CEO and Founder Nicholas Horbaczewski. “Until now, AI had yet to enter the realm of real-life sports as a competitor, and we are thrilled to watch history be made as the AlphaPilot Teams introduce AI pilots as racers for the first time in AIRR. Witnessing a moment when AI will defeat a human pilot will be a paradigm shift in how we think about the role of AI in our lives.”
Lockheed Martin Chief Technology Officer Keoki Jackson said the company has been accelerating AI-enabled autonomous technologies that help astronauts, military service personnel, and first responders do demanding and sometimes dangerous jobs more safely and efficiently.
“We look forward to working with these teams at the boundaries of AI and fully autonomous flight, and digitally transforming the way we fight wildfires, orchestrate disaster-and-recovery operations and support our front-line service personnel,” he said.
The ICARUS team is affiliated with Georgia Tech’s School of Aerospace Engineering. Team members said the challenge enables it to bring the department’s existing AI work with autonomous car racing into a three-dimensional aerial-racing space.
“We’re not all new to it because we do terrestrial racing, but racing in the air is a whole new issue,” said ICARUS Captain Manan Gandhi, a Ph.D student at Georgia Tech. “It’s a challenging problem. It requires aggressive technology and fast decision making, but we’re fairly familiar with this and have nice algorithms to accomplish these tasks.”
Other teams include Formula Drone (affiliated with UCLA) from Los Angeles, KEF Robotics from Pittsburgh, MAVLab (Delft University of Technology) from Delft, Netherlands, TEAM USRG @ KAIST (Korea Advanced Institute of Science and Technology) from Daejeon, South Korea, Team Puffin, comprised of four team members from the U.S., Sweden, and Australia, UZH Robotics and Perception Group (University of Zurich) from Zurich, Switzerland, the Warsaw MIMotaurs from Warsaw, Poland, and XQuad (Federal University of Minas Gerais) from Minas Gerais, Brazil.
The AlphaPilot Challenge launched in November with 424 teams from 81 countries. Teams had to compete in a series of qualification tests this spring. A panel of industry experts then evaluated their technical strategies and abilities to develop image-classification algorithms and perform in simulated racing environments.
Gandhi, from ICARUS, said his team thought the challenge would be a great opportunity to apply their lab research in a real world setting. Georgia Tech has already produced one drone racing pilot, Nick Willard, aka “Wild Willy,” who competed in the 2018 racing season is studying Aerospace Engineering at Georgia Tech.
“The team is fairly young but we have a lot of drive and energy to solve as many problems as we can quickly and absorb as much as we can,” Gandhi said.