The cost of putting humans into space to carry out on-orbit assembly tasks can severely limit satellite design and functionality, requiring costly engineering solutions (or no solution at all) to the challenge of fitting everything inside a payload fairing, and little to no on-orbit repair capability.
A key step towards building more complex vehicles and structures in space is mastering the art of robotic assembly in orbit, sidestepping a number of problems in design and construction. If this technology can be adapted and improved, it can have applications beyond satellites to stations, vehicles, and ultimately more ambitious projects like orbital factories and refineries.
DARPA’s mission, generally speaking, is to pursue high-risk, high-payoff technology development projects for the Pentagon. The agency has spent over a decade developing on-orbit satellite-servicing and robotics technologies. Currently, DARPA has two such programs underway: Phoenix, ultimately aimed at salvaging useful components from old satellites, and the Robotic Servicing of Geostationary Satellites, intended to establish a robotics operation in geosynchronous orbit to perform servicing tasks.