Lockheed demos AI battle management system that can improvise if the unexpected occurs

Defense prime Lockheed Martin has operated a battle management system able to instantly provide choices to a drone operator if fuel runs low, the company’s innovation arm Skunk Works said today.

The flight demonstration involved integrating Lockheed’s mobile artificial intelligence technology with a Stalker XE Block 25 unmanned aerial vehicle and a modified Alta X 2.0 drone, according to a news release.

During the trials out of Fort Worth, TX, Lockheed simulated a variety of interruptions in fuel availability between the drones to showcase how the battle management autonomy can make decisions on the fly and reassign tasks or mission planning to account for unexpected problems quicker than humanly possible.

Once a fuel-related emergency was recognized, Lockheed said “the ground command-and-control system's AI analyzed the situation, generated re-plan options and displayed them to the operator” within seconds.

The operator then selected the best option, automatically reassigning the Stalker drone’s tasks to the Alta X and directed the Stalker to return to base, Lockheed said.

“By letting the AI agent handle the unexpected contingency, the operator could resolve the problem quickly while staying focused on other critical mission duties,” according to the news release.

The same ground-based C2 node that the Stalker drone fed its data to was controlling a separate unmanned ground vehicle miles away in Kansas at the same time, Lockheed said, proving how one AI system “can mutually support and direct multiple geographically separated drone meshes.”

It comes as the Pentagon is moving fast on its “drone dominance” initiative which seeks to buy more than 200,000 small commercial drones by 2027.

The flight test “demonstrates how AI can automatically adjust an unmanned mission when an unexpected problem occurs, advancing the shift towards more autonomous unmanned operations,” Lockheed wrote.