This week, DARPA is conducting its third AlphaDogfight Trials. The winning AI will be pitted against a human pilot in a human vs. algorithm aerial showdown.
This week, Defense Advanced Research Projects Agency (DARPA) will conduct the last of three planned AlphaDogfight Trials. As part of the challenge, artificial intelligence (AI) algorithms will pilot simulated F-16 fighter jets in air-to-air combat scenarios. The winning team’s algorithm will then face off against an Air Force fighter pilot in an F-16 simulator.
Originally, the final competition was scheduled to take place in Las Vegas at AFWERX featuring Air Force Weapons School pilots located at Nellis Air Force Base, however, the event is now being held virtually due to the coronavirus pandemic.
“We are still excited to see how the AI algorithms perform against each other as well as a Weapons School-trained human and hope that fighter pilots from across the Air Force, Navy, and Marine Corps, as well as military leaders and members of the AI tech community will register and watch online. It’s been amazing to see how far the teams have advanced AI for autonomous dogfighting in less than a year,” said Col. Dan “Animal” Javorsek, program manager in DARPA’s Strategic Technology Office in a press release.
SEE: TechRepublic Premium editorial calendar: IT policies, checklists, toolkits, and research for download (TechRepublic Premium)
Johns Hopkins Applied Physics Laboratory will stream the virtual event via a multi-channel ZoomGov webinar Aug. 18-20. On the first day of the event, teams will deploy their flying algorithms against “five APL-developed adversary AI algorithms.” Day two features a round-robin style tournament where teams will compete against one another. On the final day of the event, the four top-ranked teams will compete for the AlphaDogfight Trials Championship in a single-elimination bracket. The eventual champion will then face off against a human pilot in an F-16 simulator to analyze “to test the AI’s abilities against a human.”
“Prime time” viewing is scheduled from 1:30 to 3:30 p.m. EDT on Aug. 20. This block will include commentary from DARPA officials (including Javorsek) as well as recaps of the first two days of the event and the culminating human vs. AI fighter pilot dogfight.
“Regardless of whether the human or machine wins the final dogfight, the AlphaDogfight Trials is all about increasing trust in AI,” Javorsek said. “If the champion AI earns the respect of an F-16 pilot, we’ll have come one step closer to achieving effective human-machine teaming in air combat, which is the goal of the ACE program.”
SEE: Natural language processing: A cheat sheet (TechRepublic)
The first two trials were held at Johns Hopkins Applied Physics Laboratory (APL) in Laurel, Md., in November 2019 and January 2020. Originally, eight teams were selected to compete in the AlphaDogfight Trials with the objective being “to demonstrate advanced AI algorithms capable of performing simulated within-visual-range air combat maneuvering.”
These trials were created to expand DARPA’s Air Combat Evolution (ACE) program AI developer base. The ACE program is focused on automating aerial combat as well as building “human trust in AI as a step toward improved human-machine teaming.”