An artificial intelligence (AI) algorithm created by Heron Systems defeated a human US Air Force (USAF) pilot 5-0 on 20 August in the final round of the United States Defense Advanced Research Projects Agency’s (DARPA’s) AlphaDogfight Trials.
The event was the culmination of three rounds of AI algorithms created by various contractors controlling simulated Lockheed Martin F-16 Fighting Falcons in aerial combat. The first round had the eight teams flying their algorithms against five adversary AI algorithms developed by the Johns Hopkins University Applied Physics Laboratory (APL).
Screenshot of the DARPA AlphaDogfight Trial final round between a Heron Systems AI algorithm and a human pilot using a F-16 simulator on 20 August 2020. (DARPA)
Teams flew against each other in a round-robin tournament in the second round. In the third round, the top four teams competed in a single-elimination tournament for the right to fly against the human pilot.
The following contractors qualified for the round-robin tournament: Aurora Flight Sciences, EpiSys Science, Georgia Tech Research Institute, Heron Systems, Lockheed Martin, Perspecta Labs, PhysicsAI, and SoarTech. Lockheed Martin-Physics AI was one semifinal while the other was Aurora Flight Sciences-Heron Systems. Heron Systems defeated Lockheed Martin in the final to advance to take on the USAF weapons school-trained human pilot.
Looking to read the full article?
Gain unlimited access to Janes news and more...
