Human space flight

Flying people to the Space Station is SpaceX’s biggest deal yet

It ought also to end NASA’s old way of commissioning hardware

IT WAS ONE of those neat bits of symmetry that history seems to enjoy. On May 30th, at 15.22 local time, Douglas Hurley, an American astronaut, blasted off from the Kennedy Space Centre at Cape Canaveral, Florida, on his way to the International Space Station (ISS). The last time Colonel Hurley went into space was in 2011. On that occasion he was the pilot of the 135th—and final—Space Shuttle mission. Since that mission’s craft, Atlantis, returned to Earth, America had lacked the capacity to launch its own astronauts into space.

Colonel Hurley’s lift-off raises the curtain on a new era. His ride to the ISS, which he shared with another former Shuttle crew member, Robert Behnken, is in a Dragon space capsule made by SpaceX, a firm founded in 2002 by Elon Musk. It was propelled into orbit at the second attempt (the first, scheduled for May 27th, was called off because of bad weather) by a Falcon 9 rocket built by the same firm, and docked successfully with the Space Station 19 hours later. There it will remain for between one and four months before returning to Earth with Colonel Hurley and Colonel Behnken aboard. Assuming that all goes off without a hitch, this flight will be the biggest feather yet to grace SpaceX’s well-festooned cap.

Changing the rocket equation
It will also, however, be something else—the biggest endorsement so far of the new approach to procurement that NASA embarked on a decade and a half ago. For it will be the first time anywhere in the world that astronauts have reached orbit in a craft operated by a private company rather than a government agency.

For NASA, SpaceX’s rockets have many selling points. One is simply their Americanness. Since the last flight of Atlantis, America has had to pay Russia to fly its astronauts to the ISS. That has been an embarrassment. Dragon is also more capable than Russia’s veteran Soyuz system, in that it can carry seven people to Soyuz’s three. And, even were that not so, the aphorism “don’t put all your eggs in one basket” argues the advantages of having a second means of reaching the space station.

The third selling point, and the most transformative, is cost. The Space Shuttle programme was ruinously expensive. Definitive numbers are hard to come by (space-flight accounting sometimes seems far more complex than mere rocket science). But using NASA’s own figures the Planetary Society, a space-exploration advocacy group, reckons that the total cost of developing the Shuttle orbiter—just the spaceplane itself, in other words, ignoring the booster rockets that helped it into orbit—was $27.4bn in 2019 dollars. By the society’s reckoning the Dragon programme cost NASA just $1.7bn, making it the cheapest human-rated spaceship ever developed in America (see chart).

Dragon is cheap for two reasons. One is that SpaceX’s focus has always been on driving costs down. The firm was founded to pursue Mr Musk’s desire to establish a colony on Mars. Cheap access to space is the sine qua non of that ambition. The firm takes an iterative approach to design, learning from each launch and making appropriate tweaks—but has also made radical, money-saving innovations. In particular, other firms’ rockets are discarded after they have done their job. SpaceX’s Falcon machines are partly reusable. Their first stages are designed to fly back to Earth and land on ships at sea, whence they can be returned to shore and flown again.

SpaceX’s low costs have given it a big chunk of the commercial launch market. Jeffries, an American bank, reckoned last year that the firm’s per-launch prices were below those of all competitors. Its reward was to have sold $2bn-worth of launches in 2018—more than any of those competitors.

The second reason Dragon is cheap is NASA’s procurement shake-up. The old method was to award tightly specified contracts to build rockets and spacecraft to incumbent aerospace giants, who were then guaranteed by those contracts a profit on top of the costs they accrued. There was little competition involved in these arrangements, and few incentives to keep costs down. In the mid-2000s Mike Griffin, one of Mr Bridenstine’s predecessors, began experimenting with a new approach. The agency started to award fixed-price contracts, and to include ambitious, unproven startups, as SpaceX then was, in the list of competitors. Instead of specifying what a rocket would look like, NASA stated what it wanted it to do (take cargo to the ISS, for instance) and then left the competitors to work out the details for themselves.

Ticket to ride
Despite scepticism from military-industrial grandees, and from within NASA itself, this approach has been vindicated. Until now, its biggest success had been the provision of unmanned cargo runs to the space station. These have been flying since 2012 and have been shared between SpaceX and Orbital Sciences, a veteran space company that also bid for the new-style contracts. SpaceX’s share of those contracts helped pay for the development of both the Falcon rocket and the cargo-only predecessor of the current Dragon spacecraft.

Crewed spacecraft were the next step. Not everything has gone to plan. The crewed version of Dragon is late, having been supposed to fly in 2015. An explosion during testing in 2019 destroyed one of the spacecraft entirely. But the fact remains that NASA’s new, cost-saving approach has added a human-rated spaceship to its list of accomplishments. It may yet add a second—for Boeing, one of aerospace’s oldest dogs, has been trying to teach itself the tricks for success under this new regime. An uncrewed launch of the firm’s Starliner capsule, in December, however, went badly. No crewed flight is expected until 2021. Moreover, Starliner is likely to be more expensive than Dragon. NASA reckons that launching astronauts in a Dragon will cost it $55m per seat while for Starliner the figure will be $90m.

The question is how far the lean-and-agile approach can go. NASA has been instructed to return Americans to the Moon. The current target date for doing so is 2024, though history suggests that is likely to slip. To get them there it has commissioned a new, superheavy rocket called the Space Launch System (SLS), the procurement of which has exemplified the old approach. Work has been doled out among the constituencies of powerful politicians. Those same politicians have micromanaged the rocket’s specifications by insisting that it recycle old Space Shuttle technology, leading some to dub the SLS the “Senate Launch System”. NASA reckons it will have cost at least $17bn by the time of its first mission, in 2021. The Orion spacecraft that will sit on top of it has cost $23.7bn to develop.

Some other parts of the return-to-the-Moon mission are already being contracted out to keen private-sector bidders. Last month NASA awarded three new-style contracts to design a lunar lander. One went to Blue Origin, an upstart rocketry firm owned by Jeff Bezos, the founder of Amazon; a second went to Dynetics, a veteran space-and-technology firm based in Alabama; and the third went to SpaceX itself.

But why not go further? SpaceX is already developing a new, fully reusable rocket-and-spaceship combination, called Starship. This will be a lot bigger and more powerful than Falcon 9. It will both cut the cost of taking things into space and be capable of carrying heavier payloads there, which is crucial to Mr Musk’s goal of colonising Mars.
If and when it flies, Starship could offer a new and much cheaper route to the Moon. Thanks to its reusability, Mr Musk is aiming at a cost per launch of Starship of around $2m. That is hugely ambitious. One analyst thinks $10m is the lowest SpaceX could go. But even that would be an extraordinary achievement, considering that the firm’s existing Falcon-rocket launches cost $62m a pop. However, NASA estimates that the SLS will cost between $500m and $900m to launch. Mr Musk could miss his target by two orders of magnitude and still undercut the competition. The SLS, in other words, risks ending up as a white elephant.

To infinity, on a budget
A privately run Moon mission really would be new territory. Starship’s development has not been entirely smooth. SpaceX’s iterative approach to design has seen four prototypes destroyed in accidents of various sorts (the most recent on May 29th, when one exploded after an engine test). And Mr Musk has much on his plate already. He runs Tesla, an electric-car firm, the Boring Company, a startup intended to transform the ancient art of tunnelling, and Neuralink, which plans to link brains directly to computers. Besides Starship, SpaceX itself is working on a plan called Starlink to beam broadband internet transmissions to and from a constellation of thousands of low-cost satellites, an idea that has never worked properly before, and has bankrupted many who have tried. But SpaceX also has a record of doing what it has said it would—albeit not always on time. Adding a Moon mission to the list of pending jobs would be a tall order. But people once said that about flying astronauts to the ISS.