Defense contractors are working on artificial intelligence-powered sensors and other partially autonomous machines that could help the U.S. Army process images and identify targets, Breaking Defense reported Thursday.
Vern Boyle, vice president for advanced capabilities at Northrop Grumman, said companies are developing sensors that can identify features and share data with other systems without requiring “a lot of command and control back into physical systems.”
An example of a weapon system that can see, share and record data is the Ripsaw robotic tank demonstrator from Textron Systems, Howe & Howe and FLIR Systems. This combat vehicle features a Skyraider quadcopter drone and a ground robot.
The quality of image processing by sensors and other machines relies on the quality of collected data and industry executives said companies should train algorithms on weird images and data to ensure their accuracy in target identification.
“Should we bias training data towards the weird stuff?,” said Patrick Biltgen of Perspecta. “If there’s a war, we’re almost certain to see weird things we’ve never seen before.”