Control method lets robot teams circle a target on different orbits without collisions

Shubham Singh and co-authors including Katsuki Fujisawa have described a safety-critical control method in an arXiv paper for teams of robots that circle a stationary target, each at its own distance. Each robot controls its acceleration and turn rate, measures only its range to the target and shares information only with its two neighbours in the ring.

One control law works for every robot whatever its assigned radius, and a quadratic program with higher-order control barrier functions keeps the robots clear of the target, a geofence and each other. In simulated comparisons, the closest existing multi-orbit design broke both the target standoff and the spacing between robots, even with full state information, while the new controller did not.