NASA’s plan to rescue the Neil Gehrels Swift Observatory from a slow, atmospheric death spiraling toward Earth got another pause on July 2, after a fault with the Pegasus XL launch vehicle caused the mission to be delayed. The first attempt was slated for 5:09 a.m. EDT (0909 GMT) Thursday from the Marshall Islands, but the launch did not proceed as intended, and NASA said it would set a new launch date after reviewing data from today’s attempt.

At stake is not just one spacecraft. Swift is still scientifically useful, but it has been steadily losing altitude since launching in 2004, and it is rapidly headed for an untimely demise later this year without help. Katalyst Space’s Link mission is designed to rendezvous with Swift, latch on using robotic arms, and then use thrusters to raise Swift’s orbit to about 370 miles (595 km), above the rough pathway of the International Space Station, which orbits roughly 250 miles (400 km). Another delay compresses the timeline and increases schedule pressure to keep Swift above the minimum rescue altitude of 185 miles (298 kilometers) until autumn.

Here’s the setup that makes this mission both clever and stressful. Swift was not designed to be serviced when it launched, which turns what sounds like a straightforward “space tug” job into a logistics puzzle. And NASA only gave the contract to Katalyst in September. The reason timing is so tight is that higher solar activity in recent years ballooned Earth’s upper atmosphere, which unexpectedly accelerated Swift’s fall due to increased drag. That means the spacecraft had to be ready for space in less than a year after the usual punishing requirements of design, build, and testing.

The mission architecture is equally specific. Link, developed by Katalyst Space for $30 million, was set to go up inside a modified Lockheed Martin L-1011 airliner, carried to midair by the plane. From there, a Northrop Grumman Pegasus XL rocket was meant to launch Link into orbit, where it would rendezvous with Swift. Rendezvous is expected to take about a month, after which Link would approach Swift for evaluation, latch onto the observatory with robotic arms, and gradually bring Swift’s orbit to roughly 370 miles (595 km).

NASA is also trying to protect the mission margin with operations changes that effectively trade some science time for survivability against drag. Unless Swift altered its operations plan, it would have been irrecoverable in July. So the plan minimizes Swift’s science so the spacecraft points at targets only if the telescope is put in a “streamlined position” to minimize drag. Power consumption has also been slashed so the solar panels can fly “in a more aerodynamic orientation,” reducing the drag that is driving the fall.

This is why a launch vehicle fault hits harder than it might in a normal mission. Link was built to do a risky “boost the patient” maneuver, not to wait forever. NASA’s agency models suggest these changes will allow Swift to stay above 185 miles (298 kilometers) until autumn. The new launch date will determine how much time remains before that window closes, and how much margin Katalyst has to complete the usual spacecraft commissioning on Link, then execute the rendezvous and latching steps.

The money and the mission philosophy are part of the story too, because executives should notice what NASA is implicitly testing. Swift cost $250 million in 2004, roughly $450 million today when adjusted for inflation, and it is described as a relatively cheap observatory compared with the $10 billion James Webb Space Telescope. Shawn Domagal-Goldman, director of NASA’s Astrophysics Division, called it “a high-risk, high-reward mission” and said NASA has “much to gain” by attempting a boost that is more affordable than replacing Swift’s capabilities, while also advancing the nation’s satellite servicing industry. NASA’s framing is that servicing is the future, and Swift is the proving ground.

If the rescue works, the scientific payoff could still be meaningful even if Swift is only partially “saved.” NASA did not say how much longer Swift would be able to keep observing after reaching the new altitude safely. But the European Space Agency’s figures suggest spacecraft at an altitude of 310 miles (500 km) reenter the atmosphere within about 25 years. That implies that, as long as Swift’s instruments hold out, scientists could have many years of observations ahead.

For decision-makers watching this, the second-order lesson is that satellite operations are becoming a real-time risk management problem, not a one-and-done launch event. When solar activity changes the atmosphere, fall rates change, rescue deadlines move, and launch slips become existential. Swift is a “multitool” for studying the cosmos, described by Swift’s principal investigator S. Bradley Cenko as observing with a wide range of light and rapidly pointing at short-lived outbursts to alert other facilities. But even a multitool needs maintenance windows. Delays like this one test whether servicing strategies can scale in practice, under schedule stress, launch variability, and a planet that does not always behave like the models used to assume.