What happened in the 15th test
According to a report by IT之家 citing Space.com, SpaceX has taken an unprecedented step with the 15th test flight of the Starship system. The stated goal was to bring both components of the rocket — the Super Heavy booster and the Ship vehicle — back to the launch tower at Starbase, Texas, to be caught by the tower's two mechanical arms, nicknamed the 'chopsticks.'
The booster did not complete the maneuver: on Friday evening, after a low-fuel alarm, it changed its reentry plan and splashed down in the Gulf of Mexico. The ship, however, on Saturday afternoon, 22.5 hours after launch, was caught by the tower's arms. Before that, SpaceX had caught the booster on three occasions but had never attempted to capture the ship.
The difficulty is structural: the ship reenters from higher altitudes and at higher speeds than the booster, making the entire procedure more complex. The comment by SpaceX spokesperson Dan Huot during the livestream — 'we have caught a Starship, from now on everything changes' — is a company statement, not a verification of results.
Why it matters: the operating model and the bottlenecks
Capture by the arms is the heart of the operating model envisioned by SpaceX: booster and ship return, are inspected, repaired and relaunched, aiming for a very high frequency. Elon Musk has set the goal of 10,000 launches per year by 2030, more than one per hour. This scale is motivated by the satellite constellation projects cited in the material: 100,000 satellites for Starlink V3 and one million vehicles for the planned Starmind AI. By comparison, the current Starlink network in low Earth orbit has about 11,000 satellites, already the largest constellation in existence.
The current limit is rapid reusability. SpaceX has reused booster B14 twice (test 7 in January 2025 and test 9 in May 2025) and B15 (test 8 in March 2025 and test 11 in October 2025), but with months between flights. No ship had ever been reused before Saturday's capture; according to the plan described by Musk, the one recovered could fly again.
The 2028 sticking point: crew and orbital refueling
NASA has chosen the Starship vehicle as the first human lander of the Artemis program, with the first crewed landing expected as part of the Artemis 4 mission, scheduled for 2028, near the lunar south pole; the stated goal is a base by 2032. A pressurized cabin and life support systems are needed: according to a NASA report last month, SpaceX is expected to complete by year-end 'Cabin 1,' a full-scale ground prototype.
The second obstacle is orbital refueling: the ship will need to be refueled by dedicated tanker ships at a station built by SpaceX. Estimates of the number of tankers vary: the cited outside experts speak of a maximum of 22, Musk of 6–8. The technology — transfer of liquid oxygen and liquid methane at cryogenic temperatures — has not yet been validated in flight. If SpaceX is not ready by 2028, NASA could turn to Blue Origin's Blue Moon lander, already under contract for Artemis.
