The Pentagon’s Defense Advanced Research Projects Agency (DARPA) completed the first-stage virtual testing of advanced artificial intelligence (AI) algorithms designed for conducting air combat maneuvering, colloquially known as an aerial dogfight. The video with AI performance is available at the DARPA’s website.
The tests were carried out within the framework of the Air Combat Evolution (ACE) program. ACE aims to develop AI dogfighting programs capable of taking over all types of air-to-air combat.
The ACE program presupposes that AI is being trained in the rules of aerial dogfighting in a similar way to how new fighter pilots are taught. It should behave as a student pilot first and then gradually improve its skills to become a professional.
ACE seeks to automate air-to-air combat and build human trust in AI as a step toward improved human-machine teaming. DARPA’s vision is that with trusted AI able to manage lower-order operations, pilots could focus on higher-order strategic challenges such as orchestrating teams of unmanned aircraft across the battlespace
The trials took place at the Johns Hopkins University Applied Physics Laboratory. Eight competing AIs were conducting virtual air-to-air combat, as well as participating in dogfights against each other.