DARPA’s AI-Controlled F-16s Work as a Team in Virtual Dogfights

DARPA’s goal of integrating AI into air-to-air dogfighting took a good step forward last month as it carried out tests where two simulated F-16 fighter jets worked together to fight off an enemy jet. 

The successful tests mark the halfway point of DARPA’s Air Combat Evolution (ACE) program’s Phase 1, which is exploring how air-to-air combat can integrate more AI and machine learning into its systems. 

The goal of the program, as DARPA explains, is “to develop trusted, scalable, human-level, AI-driven autonomy for air combat by using human-machine collaborative dogfighting as its challenge problem.”

What DARPA’s simulated dogfights demonstrated

The program kickstarted last year, and the first simulated fight took place in August 2020, which saw simulated F-16 fighters in one-to-one aerial dogfights. These recent tests saw two to one aerial simulated dogfights, and more weapons added into the mix. 

In these tests, two simulated F-16 fighters (in blue) worked collaboratively against an enemy fighter (in red). 

Now, DARPA’s “biggest focus at the end of Phase 1 is on the simulation-to-real transition of the AI algorithms as we prepare for live-fly sub-scale aircraft scenarios in late 2021,” said Col. Dan “Animal” Javorsek, program manager in DARPA’s Strategic Technology Office.

“Managing this transition to the real world is a critical test for most AI algorithms. In fact, prior efforts have been brittle to just these types of transitions because some solutions can be over-reliant on digital artifacts from the simulation environment.”

So far, DARPA’s ACE program has hit a number of milestones, including, “advanced virtual AI dogfights involving both within visual range (WVR) and beyond visual range (BVR) multi-aircraft scenarios with updated simulated weapons; live flights of an instrumented jet to measure pilot physiology and trust in AI; and initial modifications to the first full-scale jet trainer scheduled to host an onboard AI ‘pilot’ in Phase 3 of the program.”

The team plans on carrying out Phase 2 of the program by the end of the year, which will see live subscale aircraft dogfights take place. In anticipation of Phase 3, the DARPA team has started modifications of a full-scale L-39 Albatros jet trainer, which it hopes to fly in dogfights by 2023 or 2024. 

The ultimate plan is to integrate AI and machine learning in unmanned drones, to work alongside manned fighter jets. These drones would be able to carry out unmanned dogfights, all the while the fighter pilot in his jet would be primarily focusing on managing the main attack or defense.