DARPA seeks artificial intelligence tools to automate tracking of satellite sensor data

BAE Systems has won a $7 million contract from the Defense Advanced Research Projects Agency to develop AI tools to automate the monitoring of data collected by military satellites. On May 25, the company announced that it will collaborate with artificial intelligence experts OmniTeq and AIMdyn on a DARPA project known as Oversight.

DARPA also selected Apogee Research and Systems & Technology Research for the first phase of the program.

“Traditional space-based intelligence, surveillance and reconnaissance (ISR) data collection can be disjointed and time-consuming,” DARPA states. Said.

Lyle Rudd, surveillance program manager for DARPA’s Office of Tactical Technology, said the existing reliance on individual ground station operators “significantly increases latency and minimizes the tactical benefit of satellite sensor data.” The Space Force and Space Agency said they will need tools to autonomously monitor relevant targets as they prepare to deploy large sensor satellite suites.

DARPA’s control program will develop AI-enabled software that autonomously maintains continuous “protection” of multiple targets tracked by satellites. The three companies selected for the 15-month Phase 1 surveillance will identify monitoring requirements, resource requirements, and interfaces between emerging technologies. In the second phase, the software will be transferred to the orbiting spacecraft.

Software for migration to SDA satellites

John Grimes, manager of BAE Systems’ small satellite division of FAST Labs, said the company’s tracking control software will be applied to many types of sensors, including electro-optical and radio frequency. If the UAE is chosen for Phase 2 and beyond, the software could be deployed on the Space Agency’s sensor satellites tracking hypersonic missiles, Grimes said.

Grimes described a hypothetical scenario for how autonomous target containment would work. For example, a satellite detects a signal from an unexpected ship because it is in a controlled area of ​​the ocean. The onboard system updates information about other satellites in the network and assigns this ship a priority for storage.

Instantly, an electro-optical satellite takes an image and identifies the boat as a fishing vessel. This location and ID are then sent to the nearest Coast Guard vessel for review.