Novel AI-Based Navigation System Could Repair, Maintain, and Refuel Satellites, Spacecraft In-Orbit

A new navigation system for robots is under development at the University of Cincinnati, in the hope of creating machines that can fix satellites and spacecraft in orbit — without crashing into them.

"We have to provide a reliable collision-avoidance algorithm that operates in real time for autonomous systems to perform a mission safely," Daegyun Choi, doctoral student at UC College of Engineering and Applied Science, explains. "So we proposed a new collision-avoidance system using explainable artificial intelligence."

The algorithm's first mission, though, was about as far away from spaceflight as you can get: Navigating around the two-dimensional spaces of a supermarket. "This scenario presents many of the same obstacles and surprises that an autonomous car sees on the road," co-author Donghoom Kim, assistant professor, explains.

"We can see unexpected human behaviors there and learn how well we can actually predict their follow-on motions. Likewise, we can test how we can operate those robotic platforms autonomously without causing collisions."

With two-dimensional navigation handled, the team is now expanding it to three-dimensional navigation in the frictionless vacuum of space — attempting to build a system, which would allow uncrewed robots to dock with satellites and spacecraft to effect repairs and maintenance - without running the risk of crashing into them.

Choi, Kim, and colleagues aren't the only ones working on the problem of extending the operational lifespan of satellites with automated repair, maintenance, and refueling missions: A team led by scientists from Carnegie Mellon University are looking into the same problem. "No one knows how to refuel spacecraft such as satellites and telescopes," claims principal investigator Howie Choset. "If we're successful, we will."

"Servicing satellites in orbit will soon become vital to the operations of government and commercial missions," predicts Andy Kwas, a Northrop Grumman fellow and engineering systems architect. "Our goal is to develop and transition critical concepts from this consortium to further revolutionize technologies for both government and commercial use."

Choi, Kim, and Anirudh Chhabra's conference paper on their collision-avoidance system is available under open-access terms on ResearchGate.