Moving debris into orbit to several small satellites without conductors: a decentralized method that tolerates breakdowns

Researchers cut the problem into two levels so that a swarm of vessels can carry a rigid object in a fuel-efficient manner.

With the rise of the space economy and the increasing congestion of orbits, the need to move large objects, including uncooperative debris, increases. Entrusting this work to a swarm of small craft rather than a single large tug spreads the risk, but requires coordinating the thrust of each, limited to one cone, with the rotation of the load: a formidable non-convex optimization problem.

Existing approaches oscillate between decentralized control laws that do not respect constraints and centralized planners that reintroduce a single point of failure. The authors propose an intermediate path: an internal, convex level distributes thrust between the machines, while an external level looks for the best torque trajectory to apply to the load, guided by exact gradients.

The method quickly produces fuel-efficient solutions without waiting for complete convergence, gains in performance with the number of agents over short horizons, and is accompanied by theoretical guarantees of convergence. Monte Carlo simulations also show that it recovers in a closed loop after an in-flight failure.