The Starliner comes home empty

On 24 August, NASA made the decision it had been delaying for weeks, and announced that the Boeing Starliner that carried two astronauts to the International Space Station in June would come home empty. Butch Wilmore and Suni Williams, who had flown to orbit on 5 June expecting a stay of eight days, were told that they would return instead in a SpaceX Dragon, next February, while the spacecraft they arrived in would be cut loose to fly itself back in early September, an empty shell demonstrating a trajectory it could no longer be asked to perform with people aboard. The agency’s chief, Bill Nelson, said the decision was driven by ‘uncertainty’ over the thrusters; the word, chosen with care, covered a great deal.

The failure is not a mystery, which is what makes it hard. The Starliner’s small reaction-control thrusters had malfunctioned on the approach to the station in June, overheating and losing performance, and the engineers who spent the summer testing them on the ground could not reproduce the problem well enough to guarantee a safe return. The decision was therefore not a dramatic accident but a judgment call, the kind spaceflight has always been full of, and NASA made the judgment that a crewed return carried more risk than it was willing to accept. The courage in the decision is real, and the cost is real too: a spacecraft built to carry people has, for the second time in its history, failed to do so, and Boeing is now two missions deep in a hole it dug with its own engineering.

The broader picture is worth stepping back to see, because the Starliner’s troubles are the exception that proves the rule of the programme. The commercial crew initiative, under which NASA buys rides from private companies, was supposed to create redundancy, and redundancy is exactly what saved Wilmore and Williams: the SpaceX capsule that will bring them home is the second half of the same bet, and it works. The system did what it was designed to do, absorb the failure of one supplier through the existence of another, but the image the public will keep is not the redundancy; it is the two astronauts waving goodbye to the spacecraft that betrayed them, and the image is accurate about the risk underneath the achievements.

The astronauts themselves have handled the news with the professional calm the job demands, and an extended stay of eight months is not a hardship by the standards of the profession; astronauts have lived on the station for a year. What is unusual is the manner of the extension, not its length, and the months of uncertainty, resolved only at the end of August, are the part the official statements cannot quite contain. The station is well supplied and the work continues; the loneliness of the situation is not the distance from Earth but the knowledge of how the ticket home was decided.

The consequences will outlast the mission. The certification of the Starliner for regular service, which was supposed to follow this flight, is now open to question; the station’s schedule, reshuffled to accommodate the rescue, will ripple through next year; and the comparison between the two commercial capsules, the one that failed and the one that rescued, has become a public audit of the two companies the programme was meant to treat equally. Boeing will argue, correctly, that the value lies in having two suppliers, and the argument will meet the obvious reply: a supplier that cannot be relied on is not a supplier. The Starliner will land in the New Mexico desert in September, empty, and the landing will be watched by the two people who should have been inside it, eight months late and grateful that the seat next to them is occupied by a spacecraft that works.