The artificial intelligence developer Shield AI has reached another milestone in its endeavor to develop autonomous flight: for the first time, it has succeeded in controlling several aircraft simultaneously.
The San Diego-based company announced on August 21 that it recently completed flight tests with two Kratos MQM-178 Firejet unmanned target drones, each controlled by Shield AI’s Hivemind autonomous agent.
According to Shield AI, the two jets performed coordinated maneuvers and demonstrated “operationally relevant cooperative behavior.”
“This demonstration highlighted the capabilities of future unmanned jets and drones, with all computers running on board and agent communications seamlessly integrated in the air,” the company says.
Two manned jets – an Aero Vodochody L-39 and an L-29 – were used as observation and tracking aircraft during the tests.
Shield AI is best known as the company whose algorithm controlled a specially modified Lockheed Martin F-16 fighter jet during recent air combat exercises with the technology incubator of the US Air Force (USAF) and the Defense Advanced Research Projects Agency (DARPA).
During these tests, Hivemind took control of the X-62 Conduct a variable in-flight stability test and engage conventionally controlled fighter aircraft in within-range combat maneuvers – better known as dogfights.
While DARPA declined to provide details about the human-versus-machine air combat exercises, the secretive agency called the effort a success, with Shield AI’s autonomous agent able to safely conduct at least 21 missions within ten months.
The latest tests mark the second time that Hivemind has taken control of a Firejet, following flights with a single MQM-178 in March. The Kratos is the second jet-powered aircraft and the sixth overall that Hivemind has controlled.
While the technology has reached a new milestone by controlling two aircraft simultaneously, Brandon Tseng, CEO of Shield AI, explains that this number is only limited by budget constraints in testing.
“We limited the number of Firejets to two, but it could have been four, eight, 16, 32, etc.,” he notes.
In particular, Hivemind technology has proven capable of performing complex, dynamic tasks such as fighter aircraft maneuvers within visual range, rather than just simple pre-programmed flight profiles.
“They will improvise and come up with new tactics,” Tseng told FlightGlobal in April at the annual Special Operations Forces Week conference in Tampa, Florida.
The former Marine special forces soldier believes that such technologies will enable the US and its allies to deploy “millions of intelligent, resilient drones” within the next decade.
The Pentagon is in the midst of a sweeping autonomy initiative to address challenges in recruiting troops and China’s numerical superiority in ships and missiles in the Western Pacific.
Unmanned aircraft developer Kratos hopes to benefit from this initiative after ignored for the first round of contracts for the development of a so-called collaborative fighter aircraft for the Air Force.
The company’s XQ-58 Valkyrie was an early pioneer in the autonomous jet space, with several examples acquired by the USAF and US Marine Corps for testing and evaluation purposes.
Shield AI has previously stated that it is working with Kratos to adapt Hivemind for use with the XQ-58.
