AI algorithm project for autonomous dogfights gets DARPA green light
NewsNovember 12, 2020
ARLINGTON, Va. The Defense Advanced Research Projects Agency (DARPA) awarded contracts to five entities to develop artificial intelligence (AI) algorithms enabling teams of manned and unmanned combat aircraft to conduct aerial dogfighting autonomously.
Under the DARPA contract, Boeing, EpiSci, Georgia Tech Research Institute, Heron Systems, and physicsAI were tapped to develop air-combat maneuvering algorithms for individual and team tactical behaviors under Technical Area (TA) 1 of its Air Combat Evolution (ACE) program. Each team will develop AI agents that expand one-on-one engagements to two-on-one and two-on-two within-visual-range aerial battles. These algorithms, when completed, will be tested in three program phases -- modeling and simulation, subscale unmanned aircraft, and full-scale combat representative aircraft -- that are scheduled for 2023.
“The TA1 performers include a large defense contractor, a university research institute, and boutique AI firms, who will build upon the first-gen autonomous dogfighting algorithms demonstrated in the AlphaDogfight Trials this past August,” said Air Force Col. Dan “Animal” Javorsek, program manager in DARPA’s Strategic Technology Office. “We will be evaluating how well each performer is able to advance their algorithms to handle individual and team tactical aircraft behaviors, in addition to how well they are able to scale the capability from a local within-visual-range environment to the broader, more complex battlespace.”
DARPA's ACE program intends to boost trust in combat autonomy by initially working on human-machine teamed dogfighting, which will then be an entry into developing complex human-machine teaming for defense. According to Javorsek, in Phase 1 the different technical areas will be participating in increasingly realistic scrimmages as the tech is matured and risk is reduced; the transition to subscale unmanned aircraft could be less than a year away, he adds.
