PhysicsAI Machine-Learning Technology Controls F-16 Aircraft in Real-world Dogfight Trials
23 Julio 2024 - 6:50AM
PhysicsAI’s autonomous artificial intelligence agents have
successfully controlled the X-62A VISTA aircraft in real-world
dogfight test engagements. Test flights of the AI-powered fighter
aircraft began in December 2022 and include Live Virtual
Constructive (LVC) dogfights, live testing of AI agents dogfighting
against expert pilots, and a test flight with U.S. Secretary of the
Air Force Frank Kendall in May 2024.
PhysicsAI achieved these milestones as a member of the X-62A Air
Combat Evolution (ACE) Team, which is composed of U.S.-based
companies, government organizations and academic institutions.
PhysicsAI’s algorithms controlled a specially modified F-16 test
aircraft known as the X-62A, or VISTA (Variable In-flight Simulator
Test Aircraft), at the Air Force Test Pilot School at Edwards Air
Force Base, California. ACE team’s accomplishments have introduced
a fundamental paradigm shift in autonomous aviation that opens the
door to radical leaps in military and commercial aviation
performance capabilities going forward.
“PhysicsAI is honored to be one of the companies that competed
in the original AlphaDogfight Trials in 2019-2020 and continues to
support the ACE program to this day,” said Dr. John M. Pierre,
principal of PhysicsAI. “Our team of talented scientists and
engineers have worked hard over the years to create and refine our
technology to help advance ACE program goals, which include
increasing trust in combat autonomy by demonstrating
machine-learned autonomy in the demanding and safety-critical
dogfighting environment.”
Dr. Pierre added, “Our machine learning-based AI agents use
end-to-end deep neural networks to learn effective and efficient
control policies in applications such as aerial combat between
highly dynamic and unpredictable adversaries. ACE’s achievements
and the technologies developed through the program, including
PhysicsAI agents, should serve as a foundation for future advances
in safer, more reliable, more efficient aviation – both for
military and commercial domains.”
The Defense Advanced Research Projects Agency (DARPA) launched,
leads, and funds the ACE program. The X-62A ACE Team was one of
four finalists for the National Aeronautic Association 2023 Collier
Trophy. The PhysicsAI team performing on the ACE program included
BAE Systems FAST Labs and HRL Laboratories, LLC.
About PhysicsAIPhysicsAI is a research and
development company that builds novel technology for advancing the
state-of-the-art in robotics, autonomous vehicles, and intelligent
sensors. PhysicsAI specializes in designing innovative deep neural
networks that can perceive the real-world, make complex decisions,
and adapt to changes in real-time. In addition, PhysicsAI has
developed proprietary technology for large-scale training of
autonomous agents and creating synthetic training data using
generative AI. PhysicsAI’s customers have included DARPA, the US
Army, and the DoD Strategic Capabilities Office. Learn more about
PhysicsAI at www.physics-ai.com and follow PhysicsAI on
LinkedIn.
Contact:
Kevin Wilson
kevin@wilsonpr.com