This is like giving each satellite in a large flock its own smart autopilot that talks to its neighbors, so the whole flock flies in formation safely and efficiently—without needing one giant, slow central brain on the ground.
Coordinating and controlling large swarms of satellites in real time is computationally expensive and brittle if done from a single central controller. This work proposes a distributed, computationally efficient model predictive control (MPC) scheme that lets each satellite compute its own maneuvers while ensuring the overall swarm behaves safely and optimally.
Proprietary control algorithms and proofs of stability/optimality tailored to satellite dynamics and swarm coordination constraints; potential integration with operator-specific orbital regimes and mission concepts makes the implementation sticky once embedded in flight software and operations tooling.
Classical-ML (Scikit/XGBoost)
Time-Series DB
High (Custom Models/Infra)
On-board compute and communication bandwidth/latency constraints for running distributed MPC in real time across many satellites.
Early Adopters
Focuses specifically on making distributed model predictive control computationally efficient for satellite swarms, rather than generic centralized MPC or heuristic formation control, enabling practical deployment on resource-constrained space hardware.