A taxonomy for aerospace collision avoidance with implications for automation in space traffic management

As space becomes increasingly congested, the collision detection, evaluation, and avoidance maneuver planning problems become increasingly intractable for an unassisted human-centric response. Today spacecraft collision avoidance maneuvers are planned by human subject matter experts and conducted days to hours in advance. Automated techniques for spacecraft maneuvering can respond faster and at greater scale for tasks such as collision avoidance and station keeping. Before automated systems may be used, they must be rigorously verified by evaluating the maneuver decision and control systems against system requirements and safety properties. However, no comprehensive list of requirements for automatic spacecraft maneuvering exist in standards, regulations, or research literature. The Air Traffic Management community’s regulations and today’s aircraft collision avoidance systems serve as inspiration in the space domain. In this work, four categories of collision avoidance are proposed with descriptions and examples of each for the air and space domain. Aircraft domain examples in the most automated of the four categories are evaluated in depth to provide insight into possible spacecraft collision avoidance system requirements.