Two decades of rocket engineering made it cheap to reach orbit. The returns in the next decade may come from what happens after the rocket lets go.
Katalyst Space, a startup founded in 2020 and based in Flagstaff, Arizona, closed a $12 million round in June 2026 to build NEXUS, a geostationary robotic servicer. On August 14 it was selected by the U.S. Defense Innovation Unit (DIU) for a Deorbit-as-a-Service program supporting the Space Development Agency (SDA). The company's wager is simple to state and hard to execute: servicing, refueling, and upgrading satellites in orbit is becoming repeatable infrastructure, not a string of one-off rescue missions.
TIMELINE: Katalyst Space
────────────────────────────────────────────────────────────
2020 ───── 2025 ───── 2026 ───── 2026 ───── 2027
● ● ● ● ◉ NOW
Founded NASA $12M DIU NEXUS-1
Swift Series Deorbit GEO
SBIR (Jun) as-a- debut
(Sep) Service
(Aug)
Source: Katalyst Space newsroom, NASA, DIU program selection (2020 to 2027)
The paradox of orbital economics
The launch industry spent twenty years driving down the cost of getting mass to orbit. Reusable rockets cut the price of a ride by roughly 90 percent. That achievement created a new problem: once a satellite is up, operators are largely stuck with whatever they launched. Most spacecraft were never designed to be touched, refueled, or upgraded after deployment.
Katalyst argues the money is in fixing that constraint. Its thesis inverts the usual space story. The valuable layer is no longer the rocket. It is the robot that can approach a satellite already in orbit, dock with it, and change what it does. Life extension, hardware installation, inspection, and deorbiting all become services a single vehicle can sell to many customers.
Geostationary satellites sit in one ring at one altitude. A single servicer can reach many customers without large maneuvering costs, so one launch becomes a multi-customer revenue engine rather than a single mission.
One spacecraft, many missions
NEXUS is the vehicle built to test that model. The company describes it as roughly double the power, mass, and maneuvering budget of its precursor LINK, with two payload bays for carrying and installing hardware on orbit. The planned 2027 debut flight bundles three jobs onto one spacecraft: install a space-domain-awareness module on a U.S. Space Force satellite, run additional inspection and rendezvous work for the U.S. government, then dock with a commercial satellite for life-extension work.
The point is not any single task. It is the common architecture underneath them. Rendezvous, capture, maneuvering, and integrated operations are reused across customers, which lets the company pitch a fleet of standardized vehicles instead of a bespoke spacecraft for every requirement. That is the difference between a rescue service and infrastructure.
Capital to build the NEXUS GEO servicer
Funds development of Katalyst's first geostationary robotic servicing spacecraft, targeted to fly in 2027. · Katalyst Space, 2026
LINK and Swift: a proof point with asterisks
Before NEXUS reaches orbit, the company had a nearer demonstration. Under a $30 million NASA contract awarded in September 2025, it built LINK to rendezvous with and reboost the Neil Gehrels Swift Observatory, a 2004 telescope with no propulsion of its own that had drifted toward reentry. LINK was designed, built, and tested in roughly nine months and launched on a Northrop Grumman Pegasus XL in July 2026.
The mission did not go to plan. An attitude-control anomaly meant LINK would not capture and boost Swift as intended. NASA confirmed in an August 19 update that the spacecraft would instead attempt rendezvous and proximity operations to demonstrate the core capability, without the reboost. For an investor, that is the honest caveat: the company proved it can field a sophisticated robotic spacecraft fast and cheap, but the flagship demonstration of value has not yet closed. The rendezvous attempt still matters as technical de-risking for NEXUS.
The field Katalyst is entering
Katalyst is not alone. In-space servicing, assembly, and manufacturing (ISAM) has shifted from research to early operations. Northrop Grumman's SpaceLogistics launched its Mission Robotic Vehicle in July 2026, the first privately owned operational robotic servicing mission, carrying robotic arms to extend satellite life in geostationary orbit. Astroscale is preparing ELSA-M for active debris removal. Starfish Space is flying Otter vehicles for multi-mission servicing. Orbit Fab is building the refueling interfaces that make in-orbit gas stations real.
| Player | Core service | Status |
|---|---|---|
| Northrop Grumman | Life-extension pods (GEO) | MRV launched Jul 2026 |
| Astroscale | Debris removal (ELSA-M) | Demo launch 2026 |
| Starfish Space | Multi-mission Otter | Otter flights 2026 |
| Katalyst Space | Multi-mission NEXUS | NEXUS-1 set for 2027 |
Competitive field in orbital servicing, 2026. Katalyst's differentiator is the standardized, multi-customer vehicle rather than a single service.
What separates the company is the standardized, multi-mission frame. Incumbents tend to lead with one capability, usually life extension. Katalyst is betting that stacking services on one flight is what makes the unit economics work, and that a common robotic platform beats a new vehicle per requirement.
Why the DIU selection matters
The August 14 DIU award is the part most relevant to the investment case. Deorbit-as-a-Service asks the company to show that disposing of dead satellites can be bought as a repeatable commercial service instead of funded as a bespoke spacecraft per target. The program runs in two phases: a design review, then a flight using NEXUS to capture and deorbit SDA satellites not built for servicing.
The acquisition logic is the real shift. Government users stop funding a new spacecraft for every need and start buying the outcome from an existing commercial platform. That model is what lets a small company with a $12 million round sit alongside primes on a national-security program. It also widens the addressable market from life extension alone to the full lifecycle of a constellation.
NEXUS-1 launch and three-in-one mission execution in 2027
Signature of the first commercial GEO life-extension contract
LINK's rendezvous demonstration outcome and any Swift reboost attempt
Competing awards under DIU and SDA deorbit programs
The investor read
The bull case is that orbit is filling up and operators need tools to manage assets across their whole life, not just at launch. Industry market research sizes in-orbit servicing at about $4.9 billion in 2025, growing toward $18.1 billion by 2035 at a 15.9 percent compound rate. As we wrote in August, Star Catcher's $65 million raise for an orbital power grid showed the same logic taking hold: infrastructure, not payload, is becoming the investable layer above the atmosphere. The company is the servicing slice of that same bet.
The bear case is execution and standardization. Robotic docking with satellites never designed for it is hard, as LINK's anomaly showed. Liability and interface standards are still immature, so a servicer often meets a unique customer each time. And the primes, Northrop Grumman above all, have operational vehicles and government relationships Katalyst must beat on cost and speed rather than capability.
For a private investor, Katalyst is an early, small check on a thesis that is larger than one startup. If NEXUS flies its three-in-one mission and the DIU program moves to flight, servicing stops being a rescue story and starts looking like the maintenance layer of the space economy. That is the moment the paradox resolves: the cheap rocket was never the end of the story, only the beginning of the work.
