It's better to get it up there than to delay for another 1, 2, 3, 4 percent of reliability.
In the thin upper atmosphere above Earth, a twenty-one-year-old spacecraft that has spent its life chasing the universe's most violent explosions is slowly falling home. A small Colorado startup called Katalyst Space Technologies has taken on the task of catching it — not because the tools exist, but because they are building them in time. The mission asks whether commercial urgency can accomplish what institutional caution never quite managed: to reach out and save something irreplaceable before it is gone.
- Swift's orbit is decaying faster than NASA predicted, compressing what was once a comfortable rescue window into a hard deadline of June 1 — after which the satellite drops too low for a safe docking attempt.
- Katalyst's forty engineers are soldering, assembling, and improvising around the clock in a Broomfield factory, swapping vendors and building components themselves to meet a timeline that would make traditional aerospace programs flinch.
- No one has ever attempted to capture a satellite this old, this unmodified, and this poorly documented — engineers cannot even confirm what Swift's sun-degraded exterior looks like where the robotic arms must grip.
- The rescue spacecraft, Link, will ride a Pegasus XL rocket — likely the program's final flight — launched from an airplane over the Marshall Islands to match Swift's unusual equatorial orbit.
- If the mission succeeds, it validates a new commercial model for satellite servicing; if it fails or slips, a unique astrophysical instrument burns up and a once-promising industry proof-of-concept burns with it.
Inside a factory in Broomfield, Colorado, a team of roughly forty engineers is building a spacecraft against a deadline that leaves almost no room for error. Their target is Swift, NASA's gamma-ray burst observatory, which has been hunting the universe's most violent explosions since 2004. Unusually high solar activity has expanded Earth's upper atmosphere, accelerating Swift's orbital decay. Without intervention, the spacecraft will reenter somewhere between late July and October.
Katalyst Space Technologies, founded in 2020, won a thirty-million-dollar NASA contract last September to build a rescue vehicle called Link — equipped with three robotic arms designed to clamp onto Swift and push it into a higher, more stable orbit. It would be the first time anyone has attempted to service a satellite never designed to receive visitors. When the contract was signed, engineers believed they had until late 2026 or early 2027. The accelerated decay changed everything.
The June 1 launch deadline is unforgiving. By late summer, Swift will fall below two hundred miles in altitude — too low for two large spacecraft to dock safely. CEO Ghonhee Lee acknowledges the company has traded the pursuit of perfect reliability for the chance to actually fly. Engineers are working nights and weekends, and in some cases fabricating parts themselves when subcontractors couldn't keep pace.
Among the deepest uncertainties is what Link will actually find when it reaches Swift. The satellite's exterior has endured twenty-two years of ultraviolet radiation and atomic oxygen, which can render insulation as brittle as glass. Detailed photographs of the underside — where Link must grip — are scarce even in NASA and Northrop Grumman archives. The robotic arms are being designed for maximum versatility, capable of finding purchase on whatever edges or protrusions remain intact.
Katalyst chose the Pegasus XL rocket — an air-launched vehicle likely making its final flight — over a Falcon 9, because Swift's near-equatorial orbit made a dedicated Falcon launch prohibitively expensive. The rocket will be released from an aircraft over the Marshall Islands and climb directly into Swift's orbital plane.
Lee draws a pointed contrast with NASA's canceled billion-dollar robotic servicing program, which advanced technology but never flew because it lacked a real customer. For Katalyst, the customer is real, the clock is running, and the stakes are existential. If Link docks successfully, Swift gains years of additional science and the industry gains a proof of concept. If the window closes, a singular observatory — one no other American satellite can replace — will fall silently back to Earth.
In a nondescript factory in Broomfield, Colorado, a small team of engineers is racing against the calendar to save a piece of NASA history from burning up in Earth's atmosphere. The Swift Observatory, a twenty-one-year-old spacecraft that has spent more than two decades hunting the universe's most violent explosions, is falling. Without intervention, it will reenter the atmosphere sometime between late July and October. Katalyst Space Technologies, a startup founded in 2020, has nine months to build, test, and launch a robotic rescue mission—a feat that would mark the first time anyone has attempted to service a satellite that was never designed to receive visitors.
Swift launched in November 2004 on a mission to detect gamma-ray bursts, the most powerful explosions known to science. These cataclysmic events occur without warning when massive stars collapse into black holes or when neutron stars and black holes collide. The aftermath fades in seconds to hours, which is why Swift's defining trait—the ability to rapidly pivot toward a burst and observe it before the light dies—made it invaluable to astrophysicists. For two decades, the spacecraft has done exactly that, identifying and locating bursts for follow-up study by other observatories. No other American satellite fully replicates what Swift can do.
But age and orbital mechanics are catching up. Swift orbits in the thin upper reaches of Earth's atmosphere, where residual air resistance slowly drags satellites downward. The Sun has been unusually active lately, triggering geomagnetic storms that expand the atmosphere and increase drag. When NASA awarded Katalyst a thirty-million-dollar contract last September to rescue the spacecraft, officials expected Swift would have until the end of the year or perhaps early 2027. The accelerated decay changed that calculus. Now the window is closing fast.
Katalyst's solution is a spacecraft called Link, equipped with three robotic arms and grippers designed to clamp onto Swift and push it into a higher, more stable orbit. The challenge is that no one has ever attempted this before—not on a satellite this old, not on one that was never meant to be captured, and not under such crushing time pressure. When Ars visited the company in late February, technicians were soldering components, assembling solar panels, and preparing parts for testing. For a traditional government space program, a project at this stage would typically be years away from launch. Katalyst started this mission five months earlier.
Ghonhee Lee, Katalyst's founder and CEO, is acutely aware of the stakes. "This is really technically ambitious," he said. The June 1 launch date leaves almost no margin for error. By late summer, Swift will drop below two hundred miles in altitude—too low for Katalyst to safely control two large spacecraft docking together. "It's a lot of drag with two big spacecraft docking together," Lee explained. "Originally, we thought we had more time." The company is trading the pursuit of perfect reliability for the chance to actually fly. Engineers are working nights and weekends, swapping vendors when subcontractors couldn't meet the timeline, and in some cases building components themselves. About forty employees are focused on the rescue, all working within shouting distance on the factory floor.
One of the biggest unknowns is what they'll actually find when Link reaches Swift. The satellite was never designed to be captured, and detailed photographs of the spacecraft's exterior—particularly the underside where Link will attempt to grab it—are scarce or nonexistent. Engineers have combed through NASA and Northrop Grumman archives without finding the documentation they need. The outer insulation has been exposed to twenty-two years of ultraviolet radiation and atomic oxygen, which can cause the material to become brittle and shatter like glass. A radiator panel was painted before launch and has likely cracked and peeled. "As soon as you initiate fracture on it, it will just propagate," one Katalyst manager said. The robotic arms are being engineered to be as versatile as possible, capable of finding and gripping whatever edges or protrusions might be grabbable.
Katalyst's choice of launch vehicle also reflects the constraints of the mission. Rather than booking a ride on SpaceX's Falcon 9, the company selected the Pegasus XL, an air-launched rocket that hasn't flown since 2021. The decision might seem odd until you consider Swift's orbit. The spacecraft flies near the equator, between twenty degrees north and south latitude, to minimize exposure to the South Atlantic Anomaly, a weak spot in Earth's magnetic field where radiation can damage satellites and contaminate observations. A dedicated Falcon 9 launch from Florida would cost sixty-five to seventy million dollars. Pegasus, built for exactly this kind of unusual inclination and responsive mission profile, was the right tool. The rocket will be flown to Kwajalein Atoll in the Marshall Islands, released from an airplane, and climb into the specific orbit Swift occupies. It's likely the last Pegasus mission ever—the program is winding down—which adds a certain poignancy to the rescue attempt.
Lee is under no illusions about what lies ahead. "We realized that you can't get 100 percent guaranteed success on this," he said. But he also believes this approach—accepting risk, moving fast, focusing on a real customer need rather than abstract capability development—represents a better way forward than the government's previous attempts at satellite servicing. NASA spent 1.5 billion dollars on a now-canceled robotic servicing project that advanced key technologies but never flew. A DARPA program faced similar delays. "The problem with those programs is that they never really had a customer or a use case," Lee said. "The danger of that is you get massive amounts of scope creep." Commercial companies, by contrast, have to fly or they don't survive. "For us, these things have to fly, otherwise our company doesn't deserve to exist," he said. If Katalyst launches in June and everything works as planned, Swift will get a new lease on life, and the space industry will have proven that rescuing aging satellites is not just possible but economically viable. If the launch slips, or if Link cannot find a suitable grip on Swift's weathered hull, the observatory will fall to Earth, and a unique capability will be lost. The clock is ticking.
Notable Quotes
This is really technically ambitious. It's a lot of drag with two big spacecraft docking together. Originally, we thought we had more time.— Ghonhee Lee, Katalyst founder and CEO
We realized that you can't get 100 percent guaranteed success on this. It's better to put together a functional spacecraft that has a realistic shot pulling off the mission than delaying by two or three months to ensure another 1, 2, 3, 4 percent of reliability.— Ghonhee Lee
The Hearth Conversation Another angle on the story
Why does Swift matter so much that NASA is willing to try something this risky?
Because there's nothing else like it. Gamma-ray bursts happen without warning, and Swift can turn and look at them faster than anything else we have. For twenty-two years it's been the first to see these events. Once it's gone, we lose that ability until someone builds a replacement—which costs hundreds of millions and takes years.
But Swift is old. Why not just let it go and build something new?
That's the obvious question, but it's more complicated. Swift still works. It's still making discoveries. And building a new observatory takes a decade or more. In the meantime, we're blind to these events. Plus, the rescue mission itself is the point—NASA wants to prove that you can service old satellites, that they don't have to be disposable.
Katalyst has never docked with another satellite before. How confident are they?
Honest answer? They're not confident. They're accepting that they might fail. But they're also accepting that waiting for perfect confidence means Swift falls before they can launch. So they're choosing speed over certainty. Lee said it plainly: it's better to get a functional spacecraft up there with a realistic shot than to delay and gain another few percentage points of reliability.
What's the hardest part of actually grabbing Swift?
They don't know what they're grabbing. There are no good pictures of the underside of the satellite. The insulation is brittle from decades in space. Paint has cracked. A radiator they might want to grab is probably peeling. So they're building robotic arms that can handle multiple scenarios, and they're hoping they find something solid to hold onto when they get there.
Why use a rocket that hasn't flown in five years?
Because it's the only rocket that makes sense for this orbit. Swift flies near the equator to avoid radiation. A Falcon 9 would cost seventy million dollars and be overkill. Pegasus was built for exactly this—unusual orbits, responsive missions, the right payload capacity. It's probably the last Pegasus ever, which is bittersweet.
If this works, what changes?
Everything. It proves you can rescue satellites that weren't designed to be rescued. It opens up a whole new market for companies like Katalyst. And it keeps Swift alive, which means we keep seeing gamma-ray bursts that we otherwise would have missed.