NASA Loses Swift Rescue and Starling Scores a First: Two Lessons in Space Autonomy

One mission proved a new way to navigate. The other failed to save a telescope. Both landed on NASA's desk within a month.

On 17 August, NASA announced that its Starling mission had used a new system to fix a satellite's position in orbit by referencing other objects in space instead of a navigation network. The FALCON technology demonstration, a joint experiment between NASA and the Stanford spinout EraDrive, matched objects seen by Starling's cameras against a public catalog of roughly 20,000 space objects maintained by the U.S. Department of War. NASA says the spacecraft then used those verified objects as reference points to calculate its own orbit. Over three days, FALCON refined the known orbits of more than 200 objects without help from the ground.

Two days later, the picture darkened. Scientific American reported that NASA's Neil Gehrels Swift Observatory, launched in 2004 to study gamma-ray bursts, will reenter Earth's atmosphere and burn up later this year after a rescue attempt failed before it truly began.

A telescope without thrusters

Swift has no thrusters of its own, so it cannot raise its orbit. The observatory began losing altitude fast during a period of intense solar activity that started in 2024. Ars Technica explains the mechanism: increased solar activity heats Earth's atmosphere, which increases drag on satellites in low Earth orbit and accelerates orbital decay.

In 2025, NASA awarded a $30 million contract to Katalyst Space Technologies, an Arizona company, to build a rescue spacecraft called LINK. The plan, as Ars Technica described it, was for LINK to fly up to Swift, capture it with three robotic arms, and boost it high enough to escape reentry. NASA gave Katalyst nine months to design, build and launch a first-of-its-kind satellite, a process that usually takes several years.

LINK launched on 3 July on Northrop Grumman's Pegasus XL rocket. According to NBC News, mission managers established communications after launch, but less than two weeks into the flight teams began reporting problems with the probe's orientation. In a 15 July blog post, NASA said Katalyst engineers had addressed early communications and attitude control issues. The software patches appear to have failed. On 28 July, NASA said LINK's attitude control systems had malfunctioned, causing the spacecraft to spin and resulting in sporadic communications.

Ars Technica reported that two of the satellite's three reaction wheels stopped working, and that cold gas thrusters used for finer pointing control also had issues. That left three low-impulse plasma thrusters as the only means of regaining control of a spacecraft moving at nearly 5 miles per second. NASA said LINK was rotating at about 9 degrees per second. Thrusters slowed the spin this month, but the underlying attitude control problems persisted. On Wednesday, NASA and Katalyst announced they were giving up on the rescue.

"This is not the outcome we were working toward, but it does not change why this mission was worth attempting," NASA Administrator Jared Isaacman said in a statement. "We are going to learn everything we can from LINK's rendezvous attempt and put those lessons to work on the missions that follow."

Katalyst CEO Ghonhee Lee said in a statement that the company took on a "high-risk, high-reward challenge" and was proud of the milestones reached. "We have already learned a tremendous amount, and now our job is to turn those lessons into something durable," Lee said, describing a plan to build "a repeatable playbook to inform future rendezvous and proximity operations and satellite servicing."

What Starling actually showed

The contrast with Starling is sharp. Starling's FALCON payload did not need new hardware. It reused the spacecraft's star-tracker cameras, standard instruments that identify bright objects to determine a spacecraft's orientation, and combined them with EraDrive's Era-Core flight software and an onboard catalog of known satellites.

"FALCON is yet another success for the Starling demonstration mission," said Roger Hunter, program manager for NASA's Small Spacecraft and Distributed Systems program at Ames Research Center, in the agency's announcement. "The results from FALCON can have far-reaching implications for on-orbit space-traffic monitoring, collision avoidance, and alternative navigation."

NASA says the self-orbit determination is a first for a spacecraft using optical cameras to navigate by relative position to other objects. In separate experiments, the catalog-update capability produced better object position predictions onboard Starling than those provided by ground stations.

Traditional satellite navigation depends on GPS signals, which are unreliable or unavailable in lunar or deep space. NASA frames FALCON as groundwork for lunar satellite swarms, distributed science missions and human exploration. Later this year, Starling, which launched in 2023, will extend the experiment so its four-spacecraft swarm can share tracking data and refine positions collectively.

The two cases are not a clean morality tale about old and new approaches. LINK's failure was mechanical: reaction wheels, thrusters, attitude control. FALCON's success was computational. But both point at the same pressure, which is the cost of relying on ground networks and human operators for decisions that spacecraft could increasingly make themselves.

Swift's loss makes the stakes concrete. Astronomers still depend on the observatory's combination of multi-wavelength instruments and agile pointing to locate gamma-ray bursts for follow-up by other telescopes. "We were all hoping for more science from Swift," said Shawn Domagal-Goldman, director of NASA's astrophysics division, in a statement. "But we knew the takeaways from this mission would be worthwhile either way."

Katalyst is not abandoning LINK. The spacecraft is still alive, and the company says it will try to approach and rendezvous with Swift to demonstrate close-in navigation technology. NASA's astrophysics division called the mission "a first-of-its kind attempt, developed on an unprecedented timeline driven by the Sun's activity."

That timeline is the part worth watching. NASA compressed years of development into nine months and accepted technical risk to do it. Whether the agency repeats that trade, and what it changes next time, will matter more than the loss of one 22-year-old telescope.