NASA hired a startup that has never flown to rescue a falling telescope with a robot called LINK

The Neil Gehrels Swift Observatory is sliding closer to the atmosphere every day, and NASA's plan to stop it is to send a robot, built by a company that has never flown one, to grab it from orbit.

In the next several weeks, an Arizona startup called Katalyst Space Technologies is set to launch its first payload, a robotic servicer named LINK, on a mission to catch the 22-year-old Swift telescope and push it to a higher, safer orbit. NASA framed the operation as a salvage in its mission materials and told reporters this month that final launch preparations are underway. The agency has set an October 2026 deadline for the boost, because beyond that point the risk that pieces of Swift survive reentry and reach the ground starts to rise.

The bet behind the bet is bigger than one aging telescope. NASA is treating the LINK rescue as a proof of pattern: a deliberate decision to buy speed and commercial agility from a vendor with no flight heritage, rather than wait for a slower, more conservative path. Katalyst built LINK in roughly nine months from contract award and arrived at NASA Goddard for environmental testing on April 14, 2026. Satnews reported on June 29, 2026 that launch preparations are now finalized.

That cadence is the news. NASA usually procures orbital work on multi-year timelines and tends to lean on providers with extensive flight history. Here, the agency accepted a first-flight risk on a vendor whose only on-orbit track record is the April 2025 acquisition of Atomos Space, a small in-orbit servicing company whose March 2024 Quark-LITE/Gluon demo exercised orbital transfers and docking with anomalies. Katalyst's own corporate site describes its NEXUS servicing bus at TRL-6 and its on-orbit robotics at TRL-9, but those are company self-descriptions, not independent validation.

The mechanism is unusual enough to describe. LINK will not ride a conventional rocket. It is scheduled to fly as a payload on a Northrop Grumman Pegasus, an air-launched booster dropped from a high-altitude carrier aircraft over the Pacific. Once in low Earth orbit, LINK will use robotic arms and grippers to capture Swift, then fire high-thrust thrusters to raise the observatory from its current roughly 210-mile perch toward a longer-lived orbit.

The scale of that handshake is what makes the mission hard. Swift is moving, tumbling gently, and running on systems designed in the late 1990s for a different era of orbital operations. Capturing it without damaging the structure or the science instruments, then firing thrusters without destabilizing the stack, is the part the procurement decision cannot buy. NASA's preview materials are blunt about the risk. The Swift Boost mission page describes the attempt as a salvage operation rather than a guaranteed rescue, and Satnews reporting on the final preparations frames it as salvage rather than a guaranteed success. The constructive read and the risk critique coexist: NASA is opening a commercial-agile lane for time-critical orbital work, and it is also gambling an irreplaceable 22-year-old science asset on a vendor's first flight.

If LINK works, the agency gets more than extra years of Swift. It gets a working template for buying fast, high-risk in-orbit servicing, including the kind of deorbit-boost work that an aging fleet of low Earth orbit satellites will increasingly need. If it does not, Swift burns up on schedule and NASA absorbs a public setback on its first commercial orbital rescue.

The next test is the launch itself, then the rendezvous. Watch for the Pegasus drop window, the first LINK telemetry, and whether the company confirms a clean capture before the agency commits to firing the boost thrusters. That sequence, more than any single press event, will tell the reader whether the commercial orbital-rescue lane is real or aspirational.

In a separate timeline, NASA is preparing the Nancy Grace Roman Telescope, its next flagship astrophysics mission, for shipment to Florida in late August 2026. Roman and LINK are unrelated missions, but they will run within weeks of each other, with the agency staging a new commercial-services bet and a flagship science mission side by side.