An artificial intelligence (AI) agent has just scored five straight victories against a human F-16 fighter jet pilot in a digital dogfighting event hosted by the US Defense Advanced Research Projects Agency (DARPA).
A simulated F-16 controlled by an AI “pilot” developed by Heron Systems went 5-0 against its human opponent, after it previously defeated a number of other AI systems.
Heron Systems excelled against AI systems developed by Aurora Flight Sciences, EpiSys Science, Georgia Tech Research Institute, Lockheed Martin, Perspecta Labs, PhysicsAI, and SoarTech to get a chance to battle a top Air Force fighter pilot in the final.
The event was the culmination of DARPA’s AlphaDogfight Trials which aims to demonstrate the feasibility of developing effective, intelligent autonomous agents capable of defeating adversary aircraft in a dogfight.
The final phase of the trials was broadcast live from the Johns Hopkins University Applied Physics Lab via a Zoom webinar. The entire event, with narrative, is available on YouTube. The actual human pilot vs AI battle begins at around 4:38:00 into the video.
DARPA’s AlphaDogfight Trials seeks to advance the state of AI technologies applied to air combat operations. The trials are a computer-based competition designed to demonstrate advanced AI algorithms that can perform simulated within-visual-range air combat maneuvering.
In August 2019, DARPA selected eight technically and organizationally diverse teams to compete in the AlphaDogfight Trials with the purpose to energize and expand a base of researchers and developers applying AI technologies to complex operational problems.
The first of three AlphaDogfight Trials competition events was held at JHU/APL in November 2019. It was an exhibition match with the opportunity for teams to compete against different APL developed adversary agents and test the simulation environment at scale.
The second trial was held in January 2020 and was the first competition where teams were ranked against each other and tested their agents against more challenging adversary agents.
AlphaDogfight Trials is a precursor to the DARPA Air Combat Evolution (ACE) program, which involves AI development and demonstration in three program phases – modeling and simulation, sub-scale aircraft, and full-scale aircraft testing. Ultimately, ACE will be flying AI algorithms on live aircraft to demonstrate trusted, scalable, human-level autonomy for air combat.