Swift isn't dying of neglect: nobody ever built it to be fixed

It's tempting to read Swift's end as a story of a great telescope left to rot. It's a tidy thesis: we launch gleaming observatories, applaud, and then refuse to pay for upkeep. The trouble is that the record says something messier. NASA did try to save Swift. It spent real money, moved fast, and failed. That's a more uncomfortable story than neglect, and in our view a more useful one.

What happened: a rescue that needed rescuing

The Neil Gehrels Swift Observatory launched on November 20, 2004, to hunt gamma-ray bursts, the most powerful explosions in the cosmos. It has no propulsion of its own, so atmospheric drag slowly pulled it down from its original orbit of roughly 600 kilometers, and intense solar activity sped that up. In September 2025 NASA gave Katalyst Space Technologies a $30 million contract to build a robotic tug, LINK, and launch it in under a year.

LINK launched on July 3, 2026. By late July it was in an uncontrolled multi-axis spin, and two of its three reaction wheels were out. Engineers got the spin down to 1.47 degrees per second, but on August 19 NASA and Katalyst concluded that a capture was not possible because of attitude control problems. Swift is now expected to reenter later this year; Infobae reports an expected reentry in early November, though other outlets only say late 2026.

The real failure was in the design, not the budget

Here is where we part ways with the "nobody paid for maintenance" argument. Swift was never meant to be maintained. Per SpacePolicyOnline, it wasn't designed to be reboosted and has no grappling fixtures. The plan depended on LINK finding "capture locations" where one of its grippers might attach. That is improvisation on an object that never offered a handle.

So the missing maintenance budget was not a line item cut in 2026. In our reading, it was a choice made two decades earlier, when a $364 million observatory (the figure Infobae gives) was built without any way to be serviced. Asking for a reboost budget after the fact is like asking for a lifeboat after the ship has sailed without one.

The timing also tells a story. The orbit was already at 451 kilometers in April 2025, according to Daniel Marín at Naukas, and a feasibility study with a $150,000 award came only in August 2025. We think the lesson is less "NASA didn't care" than "NASA reacted late, to a decay that could be computed in advance." Orbital decay is not a surprise. It is physics with a calendar attached.

A $30 million gamble is not the scandal, and not the answer

We'd defend the attempt. Administrator Jared Isaacman argued that NASA "should be willing to move quickly and take smart risks when the potential return is worth it." Fair enough. A first-of-its-kind servicing demonstration on a satellite that was going to burn anyway is, in our view, about as cheap an experiment as exists. And it was not a total loss: Infobae reports that controllers brought LINK to within 12 to 15 kilometers of Swift, where it photographed the telescope, and that Katalyst tested its robotic arms before LINK reentered the atmosphere in late September. Katalyst says it will use what it learned in future satellite-recovery work. Whether that turns into real value, we'll see; for now it's a stated intention, not a result.

But we'd also resist the spin that "the lessons were worth it" settles the matter. Swift has been running on two gyroscopes since one failed in 2023, according to Naukas, so even a successful boost would probably have bought a fragile telescope a couple of years, not a new lease on life. NASA ended science operations earlier this year to slow its descent, then, as Xataka reports, turned instruments back on once the rescue collapsed, so at least some data comes out of the final stretch. That is pragmatic, not a victory.

What should change before the next one

Xataka notes that Hubble is also slowly decaying and that its managers had doubts about the profitability of a LINK-style mission. Swift's outcome gives those doubts more weight. Our reading, and it is only an opinion: nobody should conclude that servicing is hopeless from one rushed, nine-month spacecraft. But nobody should plan around rescue-by-improvisation either.

The practical recommendation is dull and therefore unpopular. Build servicing interfaces into new observatories, such as grapple points and a plan for the end of life. Fund an orbit-keeping or disposal option at launch rather than after the alarm is raised. And treat solar-cycle drag as a scheduled cost, not a surprise. Launches get the ribbon-cutting; this is the part that decides whether a $364 million instrument serves twenty-two years or thirty.

Swift has given us two decades of discoveries. The most honest tribute is to stop building telescopes that can only be admired, never fixed.