The Drone Racing League (DRL) bills itself as the sport of the future, so it’s no surprise it plans to help steer technology into tomorrow.
In September 2018, the DRL and Lockheed Martin announced the AlphaPilot Challenge, a competition pitting teams of engineers, students, coders and drone enthusiasts against each other to develop artificial intelligence (AI) technology to be utilized in an autonomous drone.
“Since the beginning of the DRL we have known sports are an interesting proving ground for technology,” DRL CEO and founder Nicholas Horbaczewski said. “Sports are an incredible crucible, pushing boundaries of speed and performance on a regular basis but doing so in a controlled environment. It allows you to really focus on the innovation of technology, and gives a platform to test and refine it.”
Similarly to how Formula 1 has been integral with the advancements of automobiles and Formula E is a testing ground for electric vehicles, the Drone Racing League is using its platform to improve, test and evolve drone technology and systems, having already seen advancements in small-performance drones, tracking systems and other related tech.
More than 400 teams representing approximately 2,300 innovators from 79 countries around the globe have signed up for the AlphaPilot Challenge since registration began in mid-November 2018. In conjunction with the challenge, the DRL announced the four-race Artificial Intelligence Robot Racing (AIRR) Circuit, which will debut in fall 2019.
Nine teams comprised of up to 10 members each will showcase their autonomous drones—flying without human intervention or navigational pre-programming—through three-dimensional courses with the circuit’s winning team taking home $1 million.
Not only will these teams be competing against one another for the top AI drone, but they will be pitting their technology against human pilots. The first team to beat a human pilot head-to-head will be awarded $250,000.
This wouldn’t be the first time an autonomous drone attempted to defeat a human pilot. In 2017, researchers at NASA’s Jet Propulsion Laboratory developed three autonomous drones equipped with preloaded course maps to race against DRL pilot Ken Loo, who maneuvered the course faster on average than the AI drones.
Unlike that challenge where the autonomous drones were preloaded with course maps, the AIRR drones won’t have that luxury, making the task more difficult, but the payoff that much greater.
“We know AI programming is going to be a key skill set in the future because it’s going to touch many aspects of our lives,” said Dr. Robie Samanta Roy, vice-president of Technology, Strategy and Innovation at Lockheed Martin. “It’s important we get the future workforce involved and work on things that are going to have a very positive development in this field.
“… We’re doing things with a purpose and thinking about our future. Artificial intelligence is going to have a significant impact at the global level.”
The end goal for the AlphaPilot Challenge and AIRR Circuit is to take the tried-and-tested AI technology and use it in real-world applications—especially in lieu of human-intensive tasks—including search and rescue, transportation, delivering aid and relief, reconnaissance, construction and more.
“All of these things now become economical, feasible and will get levels of data and access you can’t get without autonomous systems,” Horbaczewski said. “If you think of all the ways autonomous cars can change the world, autonomous drones can deliver that same flexibility and support.”