Boeing's Starliner aims to fly again - the backup ride NASA can't afford to lose

Space 4 min 14 sources

Two years after its only crewed flight failed, Boeing says its Starliner capsule is close to flying again - as a cargo test, not with crew. It is the second way to the space station that NASA deliberately paid for, and never gave up on.

Key takeaways

Boeing says its Starliner capsule, grounded since its failed 2024 crewed flight, may fly a cargo test late this year - the second astronaut ride NASA deliberately paid for and never dropped.
SpaceX flew Starship a 13th time with its softest splashdown yet, but experts warn its heat shield is a "dead end" for the fast reuse SpaceX wants.
The Roman Space Telescope is fuelled and running nine months early for a late-August launch, and a station crew came home after 241 days.

Boeing’s Starliner capsule has carried astronauts exactly once. That flight, in 2024, failed. Two years later, on a July 28 earnings call, Boeing’s chief executive Kelly Ortberg said the fixes are “going quite well” and the spacecraft could fly again this year [1]. Not with crew - as a cargo run to the International Space Station, effectively another test [2].

It is a strange thing to keep alive. SpaceX’s rival capsule, Crew Dragon, has been flying astronauts reliably for years. Starliner has cost Boeing about 2 billion dollars in losses on a fixed-price contract it signed in 2014, and it still hasn’t earned its keep [1] [2]. So why does NASA still want it? The answer is the quiet logic behind the whole program - and behind a lot of the world you don’t see.

What actually happened with Starliner

Starliner’s one crewed flight, the Crew Flight Test, launched two NASA astronauts to the station in June 2024. On the way up, its propulsion system misbehaved - thrusters failed and helium leaked [2]. NASA judged it too risky to bring the crew home in the capsule. It sent Starliner back empty, and the two astronauts, Butch Wilmore and Suni Williams, stayed in orbit for nine months before riding home on a SpaceX Crew Dragon in early 2025 [1] [2]. This year NASA formally classified the flight as a “Type A” mishap - its label for a serious failure [1].

Since then, Starliner has sat on the ground. Boeing says the redesign of its propulsion problems is nearly done, but a NASA safety panel warned in June that the work is still ongoing and the next flight could be up to a year away [2]. In a July 28 securities filing, Boeing said it now expects an uncrewed flight no earlier than the last three months of 2026, with a crewed mission possible only in late 2027 [2] [1]. Ortberg said he doesn’t expect to take another financial hit before the capsule flies [2].

The workhorse keeps flying

While the backup sits idle, the market leaders keep launching. SpaceX flew its giant Starship rocket on its 13th test on July 24, briefly releasing a first batch of upgraded Starlink internet satellites before the booster and ship came home [3]. The upper stage made what SpaceX called its “softest splashdown” yet in the Indian Ocean [4], and was still floating five days later [5].

But even a good flight came with a caution. Outside experts told Ars Technica that Starship’s current heat shield - the tiles that protect it from the fire of re-entry - is a “dead end” for the rapid, airline-style reuse SpaceX is chasing, and that the underlying materials research has been starved for decades [6]. Impressive engineering, real limits still ahead.

Telescopes, crews and money

NASA’s next big space telescope is running ahead of schedule. The Nancy Grace Roman Space Telescope has been fuelled and is on track to launch on a Falcon Heavy rocket around August 30 - about nine months early [7]. It will sit at a gravitationally stable point about 1.5 million km from Earth and survey huge swaths of sky.

Up on the station, a crew came home. A NASA astronaut and two Russian cosmonauts landed in Kazakhstan after 241 days in orbit, returning aboard a Soyuz capsule [8]. And the money kept moving: satellite maker K2 Space raised 500 million dollars, valuing the California firm at 6.8 billion [9]; Rocket Lab won a 266 million dollar Space Force deal to open a launch site in Alaska [10]; and SpaceX picked up 1.6 billion dollars in orders to fly military satellites [11], while lofting a classified payload for the US’s spy-satellite agency [12].

A quieter picture from deep space

Two smaller items are worth holding onto. NASA’s twin ESCAPADE probes, on their way to Mars, turned around and snapped a “family portrait” of the Earth and Moon together - a reminder of how small home looks from the road [13].

And a warning about how thin our margins can be. In late July, NASA had to evacuate its Madrid ground station - one of only three antenna complexes on Earth that talk to spacecraft in deep space - as wildfires swept toward it. Staff got out safely; the site was spared [14]. The network works because those three stations sit spread across the globe, so the planet’s rotation never leaves a distant probe unheard. Lose one, and there are still two. That spare capacity is invisible until the day the fire arrives - which is exactly the thread running through this week.

02 · Lesson · why it matters

Why NASA keeps paying for a spaceship that won't fly

A system buys safety by paying for a backup it hopes never to use - waste on every ordinary day, everything on the bad one.

The capsule that costs money and does nothing

Boeing’s Starliner has flown astronauts once. It failed. It has sat on the ground for two years and cost Boeing about 2 billion dollars in losses. Its rival, SpaceX’s Crew Dragon, flies reliably and carries the whole load.

On the surface, this is a story about a company that can’t get its product to work. Underneath, it is a story about a choice NASA made on purpose - and kept, even when it looked foolish.

Two rides, built on purpose

NASA did not hire Boeing just to get astronauts to the station. It hired Boeing and SpaceX both - two companies, two capsules, two rockets - to build the same capability twice.

Building a thing twice looks like waste. It isn’t. The point was never to run two rides at once. It was to make sure that if one was grounded, the other could still fly. The spare’s whole job is to exist.

The lean day always wins the argument

Here is the trap. On any ordinary day, the spare is dead weight. Dragon flies; Starliner sits in a hangar burning money and doing nothing.

Every instinct says cut it. That instinct is not stupid - it is efficiency itself. Efficiency means stripping out slack, and a backup is pure slack, three hundred and sixty-four days a year. The case for cutting the spare is easy to make and sounds right every time you make it. That is exactly why the spare is always in danger.

The bad day is the whole point

The margin you pay for is worthless until, one day, it is everything.

Remember what the gap looks like. From 2011 to 2020, after the space shuttle retired, the United States had no way to launch its own astronauts. It bought seats on Russia’s Soyuz - one capability, no backup, no other door. The whole two-provider design exists so that gap never opens again.

You cannot know in advance which day the fire arrives. You can only decide, ahead of time, to keep a door you might never open.

A copy is not a spare

There is a deeper turn. Two of the same thing share the same flaw.

If both capsules ran the same engine with the same hidden fault, a single grounding order would take out both at once. You would have paid twice for one point of failure wearing two hats. That is why NASA wanted a different vehicle on a different rocket - not simply a second Dragon. Real backup is not two copies of one design. It is two things that break in different ways, so no single bad day can catch them both.

The spares you are already standing on

You pay for spares everywhere, mostly without noticing. The second engine on the plane. The generator humming behind the hospital. The spare tyre, dead weight every mile except the one mile it saves you. The extra power feed into the building that holds your photos and your money.

Every one of them is waste on an audit. Every one is someone’s quiet answer to a bad day that cannot be scheduled. And the pressure to cut them - to run lean, to trim the fat - never stops, and never once sounds unreasonable.

The humility is here. We keep margins we cannot yet justify, because the day we could finally justify one is the day it is already too late to build it. Seeing the whole means seeing the spare - and knowing how much of ordinary life rests on backups we will never have to thank.

03 · Lab · your turn

The Spare

Rehearse keeping or cutting a costly backup, and feel why the spare you never use is the one that saves the bad day.

04 · Hope · carry this

People rarely stake everything on a single thread, and that quiet habit of keeping a second way through is one of the wiser things we do. The engineers still patching a stubborn capsule are making the same bet: that a hard thing is worth finishing, not abandoning.