The U.S. Air Force’s XQ-58A Valkyrie combat experimental drone has flown under artificial intelligence control for the first time. On July 25, 2023, the robotic aircraft made a three-hour combat flight at the Eglin Test and Training Range in Florida.
The latest test is a continuation of Skyborg’s Vanguard program to develop an artificial intelligence/machine learning capability for safe flight in tactically relevant missions. The AI was trained in a simulator where it honed its skills during millions of hours of training sorties on the X-62 VISTA using algorithms developed by the U.S. Air Force’s Autonomous Combat Operations (AACO) team.
A joint project between Kratos Unmanned Aerial Systems (KUAS) and the U.S. Air Force Research Laboratory (AFRL), the XQ-58A Valkyrie aims to develop the necessary technology to create an autonomous reactive combat drone capable of operating independently without human intervention, in cooperation with other drones and as part of a human/machine team.
“The tests validated the robustness of the layered safety system on the AI-controlled drone and demonstrated the AI’s ability to successfully solve tactically relevant problems and deal with ‘problematic challenges’ encountered during air operations,” said Col. Tucker Hamilton, chief of the U.S. Air Force’s AI Pilot Test Division.
“This technology has the potential to create AI pilots who will have all the skills necessary to conduct air combat in today’s environment, in both defensive and offensive air missions.”
The idea is to reduce the ever-increasing costs of manned aircraft by giving the US Air Force a robotic alternative that can accompany and work with fighter jets and also be sent into highly dangerous areas.
With a wingspan of 6.7 meters and a top speed of 1,050 km/h, the Valkyrie has an operating ceiling of 13,715 meters and a range of 3,941 km. It is capable of carrying up to eight weapons.
“Artificial intelligence will be a key element of future military combat and it will provide us with the speed we need to make decisions and react quickly to the evolving combat situation in the combat operations zone,” said Gen. Scott Kane, AFRL commander.
“Artificial intelligence and human-machine collaboration continue to evolve at an unprecedented rate, and we need the coordinated efforts of our government, academia and industry partners to keep pace,” Gen. Scott Kane added.