The Robot That Must Become a Service
How licensing, client consent and one shared attachment ring will decide whether GEO servicing becomes repeatable
On 21 July 2026, at about 5:15 pm, a Falcon 9 lifted the Mission Robotic Vehicle from Space Launch Complex 40 at Cape Canaveral Space Force Station, carrying the Robotic Servicing of Geosynchronous Satellites payload and three Mission Extension Pods. The programme that put it there began in 2016, was targeted for launch in 2023, was promised to its first customer for 2024, and flew in 2026. That decade is the load-bearing figure of this report, and it is legible only against the scale of the institutions involved: Space Systems Command manages an annual space acquisition budget of $15.6 billion; NASA’s entire contribution to this vehicle is an estimated $10 million of its own labour over four fiscal years, transferred as no money at all; the European Space Agency’s comparable programme carries a contract value of €119 million. A decade of formation, funded in fragments held on different authorities, produced one spacecraft that must now raise its orbit for roughly a year before it can touch anything. The structural constraint is that a servicing capability is not built by building a servicer. It is built by assembling, for each intervention, a stack of permissions, consents and mechanical assumptions that no single party controls. Whether that stack can be assembled twice at the same cost as once is the question the launch did not answer.
This report proceeds by mechanism rather than by chronology. It sets out the commissioning chain that separates a launched spacecraft from an accepted service, and the vocabulary that keeps the two apart. It examines the transition model through the DARPA programme announcements of 2016, 2017 and 2020, the Maxar termination notice of January 2019, and the executed NASA–DARPA interagency agreement, to establish what the government contributed and what it kept. It reads the two Space Logistics licence applications and the Commission’s grant notice against the ISAM and spectrum rulemakings to establish what the July authorisations actually permit. It traces the mechanical interface on which four independent servicing architectures have converged. It reads the four disclosed customer announcements for what they state and for the one thing they never state. It closes on procurement. It does not value the programme, assess Northrop Grumman as a security, or forecast delivered missions.