TLDR:
- Artificial Intelligence (AI) flew X-62 VISTA in a simulated dogfight against a manned F-16.
- The breakthrough is part of DARPA’s ACE program and has significant implications for future aerospace AI advancements.
Key Elements:
Last year, the U.S. Air Force Test Pilot School and DARPA conducted a simulated dogfight test where an AI agent piloted the X-62 VISTA against a human F-16 pilot. The test showcased the successful use of AI in within-visual-range engagements, marking a significant advancement in combat aviation capabilities. The X-62A broke barriers in autonomous air-to-air combat, demonstrating the potential for AI-driven autonomy in military applications.
The testing occurred at Edwards Air Force Base, California, and involved over 100,000 lines of flight-critical software changes across 21 test flights. The AI-controlled X-62A was able to perform defensive and offensive maneuvers, getting as close as 2,000 feet at 1,200 miles per hour in high-aspect nose-to-nose engagements. Safety pilots were onboard but did not need to disengage the AI agent during the test flights, showcasing the reliability and safety of AI-controlled flight.
The AI versus human dogfight scenario represents a complex challenge that the X-62A successfully navigated, proving the efficacy of non-deterministic artificial intelligence in aerospace tasks. While dogfighting was the initial testing scenario, the ultimate goal was to develop AI systems that can autonomously handle a variety of tasks in both commercial and defense applications. The lessons learned from this test will shape future aerospace AI advancements and provide a foundation for future programs and research.