The US government wants to see more of the expensive hardware in space maintained and even built there, rather than back on Earth.
Space activities add billions to the US economy, but the inability to build in orbit limits that contribution. Now, new technologies developed by the US government and private firms are showing what it will take to begin servicing, assembling, and even manufacturing in space. Experts say it is the path toward orbiting factories and long-term habitation on the Moon.
The first step will be rehabbing aging satellites rather than replacing them. NASA is plotting its first mission to refuel a spacecraft. The aerospace firm Northrop Grumman has already flown two missions to extend the life of satellites, and will soon use a new space robot to do the same at scale. The White House released a national strategy for developing these technologies in April, led by space policy advisor Ezinne Uzo-Okoro, an expert in robotic assembly who previously worked for NASA.
“It’s a big deal because it means all of the various departments and agencies within the US government got together and not only decided this was an important issue, but also were able to come to a consensus on how the US government should foster satellite servicing,” Brian Weeden, a space policy expert at the Secure World Foundation, says.
Why we don’t already do this work in space
When the International Space Station was under construction in the first decade of the 21st century, building in space meant humans in space suits out in the vacuum, using hand tools to bolt together trusses and modules pre-assembled on Earth. Robotics and computers used in spaceflight weren’t advanced enough to do the work on their own, and even the Space Shuttle’s capacious cargo bay was too small to fit a fully assembled habitat.
Pamela Melroy, currently NASA’s deputy administrator, was an astronaut who helped assemble the laboratory in orbit, piloting two Shuttle missions in 2000 and 2002, and commanding a third in 2007.
“What we did in low-Earth orbit with astronauts and robotics and spacewalking was amazing,” she said. But we can’t rely on astronauts to make servicing spacecraft routine, she explains. Spacewalks are too dangerous, particularly farther away from the planet where radiation is a bigger threat.
Instead, advanced robots could do the job—if they can find a satellite in orbit, rendezvous with it safely, grab a hold of it, and fix whatever needs fixing.
The challenge facing in-space servicing is that most spacecraft aren’t designed for it. Imagine trying to refuel an automobile that didn’t have a gasoline inlet—it would involve somehow tearing open the propulsion system and closing it back up again.
In a world where rocket launches were infrequent, space hardware was designed to last for decades before being trashed. But, as new rocket companies make orbit more accessible, and the assets there become more important, it makes less sense to simply abandon them—what if they could be refueled, refurbished, or recycled? And what if the tools for those jobs also enabled more ambitious and lucrative orbital projects?
“Instead of shelling out $200 million a pop for an asset, they’d like to keep what they have on orbit and make it useful in a different way,” Trudy Kortes, a NASA technology development manager, says.
But before satellite builders start installing things like fuel ports into their spacecraft, they need to be convinced it’s worth the extra expense. It’s a “chicken and egg problem,” Melroy says, but it’s one that’s edging toward a solution.
Step one: Keep it simple
After her stint as an astronaut, Melroy became an executive at Darpa, the military’s advanced technology hub. The organization had an ambitious plan for refueling spacecraft, one that she said was “way ahead of its time.” Under her leadership, it dialed in on a project called Robotic Servicing of Geostationary Satellites, or RSGS, in partnership with a firm called Spacelogistics, LLC, a subsidiary of Northrop Grumman.
The approach? Keep it simple, says Joe Anderson, Spacelogistics’ vice president. The most expensive satellites (and thus key targets for servicing) fly high above the Earth in geostationary orbits that keep them aligned above a specific point on the planet. More often than not, they are retired not because they are obsolete, but because they run out of fuel to keep them in place.
Spacelogistics’ team realized that though satellites aren’t prepared to be refueled, they are designed to be attached to the rockets that launch them. A vehicle could clamp on to those fixtures, and then use its own propulsion system to keep the satellite in the right spot. In 2020 and 2021, Space Logistics launched two missions with its Mission Extension Vehicles, which successfully clamped onto satellites operated by Intelsat in order to keep them in orbit.
Now, the company is building its Mission Robotics Vehicle. Rather than clamp directly onto a customer’s satellite, the MRV will use a robotic arm developed by Darpa to attach a mini-fridge sized unit called a Mission Extension Pod (MEP) onto the spacecraft. The MEP will have a propulsion system and enough fuel to keep the satellite at work for another six years.
A more ambitious mission is on the way. Earlier this year, NASA signed off on the design of its own robotic servicer, known as OSAM-1. In 2026, the spacecraft will launch into orbit and refuel Landsat 7, an Earth observation satellite operated by NASA and the US Geological Survey. OSAM-1 will use robotic tools to cut away thermal shielding and remove metal wires that secure a cap on the fuel valve, pump in new propellant, then attempt to close everything back up.
After that, a payload on the spacecraft developed by Maxar will attempt to build a working communications antenna and a beam that could be used to frame a space station.
The future of building stuff in space
Satellite servicing is the first step for these technologies because it’s something that obviously adds value now, and people are willing to pay for it. Astroscale, a Tokyo-based company that launched a demonstration mission last year to prove it could use its spacecraft to seize and dispose of a dead satellite, recently won a €14.8 million ($15.4 million) investment to launch a larger version of its debris removal spacecraft.
“We know that sustainability in orbit is big business,” Charity Weeden, Astroscale’s head of public policy, told Quartz last year. “It opens up the door to accessing on-orbit services—tow-trucks, fuel stations, machine shops, and more.”
The same technologies needed to fix or dispose of a satellite in space can be used, for example, to build much larger satellites, telescopes, or space stations in orbit.
“The key thing here, of course, is the ability to build things that are much bigger than will actually fit inside the faring of the rocket,” Melroy says. The James Webb Space Telescope that launched last year required its huge mirror to fold like origami, adding risk and complexity. Cheaper and more powerful telescopes could be built if robots were able to assemble them in orbit.
The same technologies would also lower the cost of manufacturing goods in space. Relying on astronauts and expensive facilities makes such goods cost-prohibitive. Cheaper autonomous infrastructure could change the game for companies that believe unique goods can be manufactured in microgravity, like ultra-efficient fiber optics or novel drugs.
And these ideas even tie into the plan to return to the Moon. Space engineers salivate at the idea of mining lunar ice and turning it into water, oxygen, and rocket fuel, but that, too, needs the kind of autonomous sensing and manipulation that satellite servicing will demonstrate.
At the White House, Uzo-Okoro is working on smoothing that path forward. One key task, she says, will be aligning government investment behind these tools, so companies have a clear signal of demand they can show their investors.
“How do we use government buying power to facilitate industry growth?” she asks, rattling off some potential ideas, from creating a public index of all the ways the government might use space services in the next decade, to pooling funding and technology around specific goals, or selecting specific satellites as targets for life extension. Members of her working group from agencies across the government are figuring that out right now.
“If we as the government can show that it is possible to safely integrate all these technologies, [it will] enable [private companies] to go off and build all these really fascinating new business models,” Melroy says.
Correction: This piece has been updated to report that OSAM-1 is launching in 2026 and will carry a payload developed by Maxar.