The satellite servicing industry just became real — and almost nobody noticed

For most of the last fifteen years, the phrase "in-orbit satellite servicing" was a press-release term — a way of describing work that engineers were thinking about, occasionally simulating, and almost never actually doing. A SpaceX Falcon 9 lifted off from Cape Canaveral at 5:15 p.m. ET on July 21 carrying Northrop Grumman's Mission Robotic Vehicle and three Mission Extension Pods. That launch, combined with a market forecast published by Novaspace on August 14, is the moment the conversation changes. The industry has gone from aspiration to line item. Almost no one is talking about it.

The numbers that landed this week

Start with the Novaspace report. The space consulting and market intelligence firm published its latest In-Orbit Services Markets report on August 14, 2026, forecasting that emerging in-orbit satellite services will generate approximately $3 billion (€2.59 billion) in cumulative service revenues over the coming decade. (Source: satnews.com, "Novaspace Projects $3 Billion Cumulative In-Orbit Servicing Market Over the Next Decade," August 14, 2026.) The forecast segments into three operational capabilities: satellite refueling at $1.2 billion over the ten-year window, "backpacking" life-extension at $860 million, and active debris removal plus orbital transfer making up the remainder. (Source: same Novaspace report via satnews.com.) The largest single segment by revenue is refueling, which is driven primarily by defense agencies adopting what the industry calls Dynamic Space Operations — high-maneuverability satellites that can change orbit on demand. The commercial market is the second tier, and it is real.

Then there is the launch. Northrop Grumman's SpaceLogistics subsidiary sent up the Mission Robotic Vehicle (MRV) — a robotic-armed servicing platform built in partnership with the U.S. Naval Research Laboratory and NASA — alongside three Mission Extension Pods (MEPs). (Source: SatelliteToday.com, "SpaceX Launches Northrop Grumman's First Mission Robotic Vehicle," July 21, 2026.) The MEPs are propulsion "jet packs" that attach to aging geostationary satellites and add roughly six years of operational life each. The MRV itself, with two robotic arms, can perform refueling and more complex servicing work. Northrop Grumman has already signed three customers for the MRV mission — two satellites from Intelsat and one from Optus. (Source: SpaceNews.com, "Northrop Grumman eyes 2026 launch of robot-armed satellite servicer," 2026.)

"The toolkit that we have right now when there are problems in space is very, very limited," Ryan Tintner, vice president and general manager of Northrop Grumman's Space Superiority Division, said in May ahead of the launch. "It's about to be widely expanded." (Source: SatelliteToday.com, July 21, 2026.) That is the most understated way I have heard a five-year strategic shift described.

The history that made this launch possible

The MRV mission did not emerge from nowhere. It is the third generation of Northrop's commercial GEO servicing program, and the prior generations are worth understanding because they explain why the new vehicles are credible.

The Mission Extension Vehicle program traces back to a 2011 proposal by ViviSat, a 50/50 joint venture between US Space and Alliant Techsystems. (Source: Wikipedia, Mission Extension Vehicle, citing Aviation Week 2011.) ViviSat dissolved in April 2016 after the partners fell out, and Orbital ATK continued the program as a solo project before being acquired by Northrop Grumman in 2018. The FCC approved the first MEV servicing plan in December 2017. MEV-1 launched in October 2019 and docked with Intelsat 901 in February 2020 — the first time a telerobotically operated commercial spacecraft had serviced another commercial satellite on orbit. (Source: Wikipedia, Mission Extension Vehicle; CNBC, April 17, 2020.) MEV-2 launched in August 2020 and docked with Intelsat 10-02 in geostationary orbit on April 12, 2021, while the target satellite remained operational — a first for the industry. (Source: fodnews.com, "MEV-2 Conjunction Alert," 2026; SpaceNews.com, April 12, 2021.)

MEV-2 has now been docked to Intelsat 10-02 for over five years. In May 2024, Intelsat and Northrop Grumman extended the servicing contract by four additional years, nearly doubling the original five-year service window. (Source: businesswire.com, "Intelsat to Extend Life of Additional Satellites with Mission Extension Vehicles," May 23, 2024.) That extension is the most concrete signal yet that the market treats life extension as a recurring revenue service, not a one-off demonstration.

The other half of the in-orbit servicing story is Japan-based Astroscale. Astroscale launched ELSA-d (End-of-Life Services by Astroscale — demonstration) on March 23, 2021, with a servicer and a client satellite stacked together. (Source: Astroscale.com, "ELSA-d Mission," official mission page.) The mission completed magnetic capture demonstrations in August 2021, complex rendezvous operations in May 2022, and finalized deorbit operations in January 2024, marking the world's first commercial end-of-life demonstration mission. (Source: same Astroscale.com mission page.) Astroscale's next mission, ELSA-M (End-of-Life Services by Astroscale — Multi-client), is slated to launch in 2026 and is contracted to remove a defunct OneWeb satellite from a 1,200-kilometer orbit — the first commercial contract to remove a multi-client prepared satellite using magnetic capture. (Source: space.com, "Astroscale's space debris removal demo mission funded for 2026 launch," 2026.)

So when I say the industry became real this summer, I am not saying it started this summer. I am saying the operational and commercial pieces finally aligned within the same calendar quarter. MEV-1 has been flying for six years. MEV-2 has been docked to a customer satellite for over five. Astroscale has been operating the only commercial end-of-life demonstration ever flown. But the MRV launch — a vehicle explicitly built to do both refueling and life extension at commercial scale, with three paying customers already on the manifest — is the line that distinguishes "early demonstrators" from "operational industry."

Why this is bigger than the launch

There are about 500 satellites in geostationary orbit today, and roughly 10 or more of those reach end of life each year because they have run out of station-keeping propellant. (Source: SpaceNews.com, citing SpaceLogistics President Rob Hauge, 2026.) A geostationary communications satellite that costs $250–400 million to build and launch and that has a planned service life of 15 years can now be extended for a small fraction of that cost using a Mission Extension Pod. That math is the foundation of the $860 million backpacking life-extension segment of the Novaspace forecast.

The refueling math is different and, in some ways, more interesting. A satellite that can be refueled in orbit is a satellite whose operational life is no longer bounded by its launch-time propellant budget. That changes how operators design constellations. For commercial operators, it changes the cost equation: a satellite built to be refuelable may have a higher upfront cost but a much longer revenue-generating life. For defense operators, it changes the operational equation: a refueled satellite can change orbit on demand, which makes it much harder to track and much harder to target. The "Dynamic Space Operations" concept Novaspace flagged — high-maneuverability satellites supported by on-orbit servicing — is fundamentally a military capability that happens to have commercial spinoffs.

What I think is genuinely new is the convergence. The MRV can refuel. The MEPs can extend life. Astroscale's ELSA-M can remove debris. None of these capabilities is, on its own, a new idea. What is new is that all three are now flying, with paying customers, in a single calendar quarter — and that the market forecast treats them as a coherent industry rather than as three separate demonstrations.

What this means for the rest of us

For the amateur and observer community, the practical takeaway is that the geostationary belt is about to get more crowded and more active. GEO satellites do not streak across the sky like the ISS — they appear nearly stationary because they orbit at the same rotational rate as the Earth. A modest tracking telescope can hold a single GEO satellite in the field of view while you watch it. If you have never pointed a telescope at a geostationary satellite, the DWARFLAB Dwarf 3 smart telescope is the most accessible way to do it — its tracking mount is designed exactly for this kind of slow, steady hold.

For hobbyists interested in commercial spaceflight as an industry rather than a hobby, the most useful single reference I have found remains the Wiley volume on the space economy. It is not breezy, but it is the one book I would hand a smart friend who wants to understand what is happening in orbit commercially.

For watching GEO passes with a wider field of view — the moment a satellite exits Earth shadow into sunlight, or the moment an old satellite begins a graveyard orbit transfer — a 15x70 binocular like the Celestron SkyMaster is the standard amateur recommendation. It will not resolve the satellites themselves, but it will show you the brightness transitions and the visible maneuvers that are the most user-visible signals of the in-orbit servicing industry becoming real.

The deeper view is that the satellite servicing industry just completed a fifteen-year transition from press release to line item. The ViviSat proposal that started this in 2011 was a 50/50 joint venture with no announced customers. MEV-1's 2020 docking was a world first that nearly no one outside the industry saw. The MRV launch in July was a SpaceX flight with three paying customers and a market forecast to back it. The Astroscale ELSA-M mission flying later this year will be the first commercial contract to remove a real satellite from orbit. None of this is speculative. All of it is flying.

The next year of this industry is going to be quieter than the launch visuals suggest. What we are watching is the operational phase — first refuelings, first pod attachments, first deorbits — and these are events that look like routine service work on a telemetry screen rather than dramatic moments on a launch pad. They are going to happen anyway. They will define whether the Novaspace $3 billion forecast is conservative or aggressive.

That's the essay for this week. Next Saturday: a closer look at the lunar surface power question — fission, RTGs, and beamed solar — and which architecture is most likely to actually win.

— Atlas Renner, Editor-in-Chief, SpaceOrbitals

Atlas Renner

Editor-in-Chief · Publisher