Katalyst Space's LINK robotic servicing vehicle burned up in Earth's atmosphere on September 25, concluding an 85-day mission that fell short of its primary objective: capturing NASA's Swift observatory and lifting its deteriorating orbit. The spacecraft never made contact with the telescope, and the planned rescue was abandoned.
Yet the mission's final report offers more than a simple pass/fail verdict. According to Katalyst's post-flight analysis, an electrical power-system fault knocked out two of LINK's three reaction wheels, forcing engineers to rely on propellant-hungry thrusters for attitude control. Aviation Week independently verified the reentry and the mission's end.
LINK did manage to regain partial control, close to within 12–15 kilometers of Swift, and exercise its three robotic arms. It also validated key power, propulsion, navigation, and communication subsystems on orbit. Those accomplishments, however, represent only a fraction of what a full servicing mission requires.
What LINK Actually Achieved
Launched July 3 on a Pegasus XL rocket, LINK is a 400-kilogram spacecraft equipped with three robotic arms, precision navigation, and electric propulsion. Its target was exceptionally challenging: Swift, a 2004 observatory, lacks any servicing port or grappling fixture.
During commissioning, LINK entered a multi-axis spin and temporarily lost reliable communications. Katalyst traced the anomaly to a power fault that rendered two reaction wheels inoperable. Engineers combined thruster firings with the remaining wheel and uploaded new control software, reducing the body rate from 9 degrees per second to 1.47 degrees per second—an 84% reduction. Hall-effect thrusters played a key role in that recovery.
While the stabilization was technically impressive, it altered the mission's fuel budget. Reaction-control thrusters consume propellant with every attitude correction, leaving insufficient reserves for a safe capture and orbit raise. Katalyst and NASA therefore converted the remaining flight into a proximity and robotics demonstration.
Why 15 Kilometers Is Not Docking
A 12-to-15-kilometer pass is close by orbital standards, but still far for robotic servicing. At that range, LINK collected what Katalyst calls "unresolved imagery." Space.com reported that all three electric thrusters fired and all three arms deployed—demonstrating subsystem functionality, not capture readiness.
The missing kilometers contain the hardest part of the sequence. A servicer must estimate the target's position, attitude, and rotation while closing safely, remain within approach corridors, stop at hold points, and retreat after any anomalous reading. Near contact, centimeter-scale pose errors and small closing velocities matter far more than cruise accuracy. Then the robot must touch without imparting a tumble, requiring synchronized arm motion, acceptable contact loads, and a secure grapple before any orbit-changing burn.
LINK performed kinematic, lifetime, and gripper tests in free flight, but did not validate those functions against Swift's actual geometry and motion. That creates a clear ladder of evidence: commissioning proves individual systems can operate in space; a flyby adds relative navigation and trajectory control; docking adds close-range sensing, contact dynamics, and structural restraint; reboosting adds coupled guidance for two spacecraft moving as one. LINK reached the second rung, with partial tests of hardware needed for the third.
Swift's Added Complexity
Swift offered no docking fixture, navigation marker, or purpose-built grapple point. NASA had validated controlled rendezvous using a full-scale ground model, but flight introduces changing sunlight, communications delays, sensor noise, and two independently moving bodies.
Time pressure came from orbital physics, not a failed instrument. Higher solar activity heated and expanded the upper atmosphere, increasing drag on the observatory. Drag removes orbital energy; falling lower exposes a spacecraft to denser air and faster decay. NASA adjusted Swift's pointing strategy to reduce its cross-section, but that could only delay reentry.
The Fault Exposed a Spacecraft-Level Trade
Reaction wheels orient a spacecraft without spending propellant. Losing two wheels removed that efficient three-axis control. Thrusters supplied external torque and enabled recovery, but every correction reduced the fuel reserved for rendezvous and boosting. The workaround therefore consumed the resource needed for the primary job.
Katalyst attributes the unavailable wheels to an electrical power-system fault. It has not published a full root-cause report, detailed telemetry, or an independent review. That leaves important questions open: whether one initiating fault crossed redundant channels, and how future vehicles will isolate it.
The nine-month build schedule is also evidence, but not proof of a repeatable service. Rapid development got hardware into Swift's low-inclination orbit before the telescope fell too far, but it also compressed design, integration, environmental testing, and operations preparation. Public evidence does not yet separate schedule pressure from unrelated component failure. Flight data could answer that question, though most remain private. Useful disclosures would include wheel fault timing, power-bus behavior, and thruster duty cycles.
