Why NASA Needs the Capsule It Just Watched Fail

Today on Space Stakes: after years of thruster leaks and NASA's own 'Type A' mishap classification, why is Boeing's Starliner suddenly getting more business instead of less? Before that, in the headlines: European startups crash the party at a Paris space summit, Impulse Space piles more money onto an already massive round, and Starship's first shot at reaching orbit slides another week. Welcome back to Space Stakes, your daily brief on the business of space. It's Friday, September 18, 2026. Let's get into it. And if you think this is a story about a comeback, wait until you hear who's actually footing the bill for it.

Let's start in Paris, where World Space Business Week just handed European space startups a genuine moment. The European Space Agency awarded ICEYE — the Finnish radar-imaging company — and legacy prime Leonardo parallel contracts to study architecture for the EU's future Earth Observation Governmental Service. The catch: that program doesn't start building anything until 2028, and Airbus, Thales, and Leonardo — who are themselves merging their space divisions — remain heavily favored once real construction money shows up. Now, the direction is the story. Aerospacelab walked away with a two-point-four-billion-euro contract to build two hundred sixty-four satellites for IRIS-squared — the EU's already-approved secure connectivity constellation, which locked in a signed implementation agreement in August, with Airbus holding contracts of its own as the program moves into full-scale deployment, with first launches expected in 2029. Eutelsat partnered with French ground-station startup Skynopy, and Safran teamed up with propulsion startup Exotrail. As Mark Boggett of Seraphim Space put it at a WSBW panel, European defense organizations are 'rewriting the rules as they're going along.' My read? The appetite for risk changed before the delivery track record caught up — 2028 is when you find out if it sticks.

Now, sticking with money moving fast: Impulse Space just tacked another three hundred eight million dollars onto its Series D round. That brings the round's total to eight hundred eight million dollars. The in-space mobility company — think of it as a tow-truck fleet for satellites, moving payloads from one orbit to another after the rocket drops them off — announced the first five hundred million back in June at a four-point-two-six-billion-dollar valuation. Where the number stands now is genuinely unsettled: a source close to the company told Reuters five-point-four billion, while Reuters' own wire copy, syndicated elsewhere, put it at five-point-six billion. Either way, you're looking at serious growth in a matter of months. Founder and CEO Tom Mueller told Payload the extension reflects 'the demand we're seeing from both our customers and investors, and the momentum across the business even in the last few weeks.' The company says it's sold out its 2028 Caravan rideshare missions for its Helios kick stage, and it's stacking government work too — a Space Force upper-stage contract and a twenty-eight-million-dollar extension for VICTUS SALO missions. The company also brought on its first chief financial officer, Adam Townsend, previously CFO at smart TV maker VIZIO. When a rocket company needs a finance chief, that usually means the spreadsheet finally got serious.

Over at SpaceX, the target for Starship's first real trip to orbit just slipped again. Flight 14 is now targeting no earlier than Monday, September 28 — pending regulatory approval — pushed back from a prior date of September 22. This is planned as Starship's first orbital mission, which means the current-generation ship would actually complete a real orbit or several rather than the suborbital hops you've watched across thirteen straight test flights. Now, if that flight goes well, Flight 15 becomes the first realistic opening for a ship catch, and in the best case, a double catch — catching booster and ship together. Flight 16 is the best-case candidate for the first Starship launch from historic Launch Complex 39A at Kennedy Space Center — the same pad that sent Apollo 11 to the Moon and hosted most of the Space Shuttle program. Workers there have been chilling down ground equipment and testing the tower's arms, the unglamorous sign that construction is turning into an operational pad. Worth remembering: this date has already slipped once, from a prior target of September 22, and remains pending regulatory approval, so treat September 28 as a target, not a locked date.

One more item from the unglamorous side of the business — ground infrastructure. Orange and Telesat announced they've switched on the first European ground gateway for the Telesat Lightspeed constellation, a low-Earth-orbit network still being built out. The gateway sits inside Orange's teleport near Paris, and Orange will carry that traffic onward, over its own network, to Telesat's planned point of presence in Paris — the hub where Lightspeed data is supposed to reach Orange's terrestrial customers. As @Telesat put it on X, this was 'another landing station milestone' with partners at Orange. Now, Telesat's chief commercial officer, Glenn Katz, said the company is aiming for Lightspeed — a planned constellation of two hundred twenty-five satellites — to reach global service by the first quarter of 2028. Here's the thing to hold onto: one ground gateway switching on tells you almost nothing about whether the satellites themselves are on schedule. It's plumbing, not proof — and this business runs on plumbing you rarely hear about until it breaks.

Our main story today: why is NASA about to hand more business to a capsule the agency itself flagged with a 'Type A' mishap — its own safety classification for the most serious failures? Call this one 'The Redundancy NASA Can't Afford to Give Up.' Here's the setup, in plain terms. Multiple sources tell Ars Technica that as early as next week, NASA will announce it's ordering two additional crewed missions from Boeing's Starliner capsule, kicking in money to help fix its thruster problems, and backing support to certify that capsule on a new rocket. Treat that as reported, not settled — this is sourced to Ars's own reporting, and every other outlet covering it so far, from Starlust to ufofeed, is simply relaying what Ars found rather than confirming it independently. There's no NASA press release yet, and no signed contract sitting on the record as of taping. So why now? Because SpaceX has told NASA officials, in recent private discussions, that it intends to retire the Crew Dragon capsule by 2030, if not sooner. Dragon is currently the only spacecraft NASA has for routine astronaut trips to low Earth orbit — the region a few hundred kilometers up where the International Space Station lives. If Dragon goes away before Starliner's ready, NASA doesn't have a backup ride for its own astronauts. NASA calls having two totally different vehicles able to do the same job 'dissimilar redundancy' — the idea being that a problem grounding one capsule shouldn't strand American astronauts on the ground. Right now, that's a phrase NASA is chasing, not a fact already on the books, because only one of the two vehicles has actually been flying crew reliably. This isn't an abstract scheduling problem, either — NASA rotates astronaut crews to and from the space station on a fixed cadence, and if there's only one working vehicle in the fleet, a single grounding investigation on that one vehicle would leave the station without a way to swap crews at all. And here's what should stop you: Ars Technica itself describes Starliner's spaceflight record as, quote, 'decidedly checkered,' end quote — this is a capsule that just earned NASA's most serious mishap classification, and now NASA wants to buy more of it. There's a mechanical complication layered on top, too. Starliner's Atlas V rocket is retiring, so beyond fixing the thrusters, NASA now has to help certify Starliner to fly on an entirely different vehicle. Certifying a capsule carrying humans on a new rocket isn't paperwork you rubber-stamp — it's flight testing, safety reviews, and margin analysis, the same slow process that delayed Starliner's crewed debut in the first place. And here's the money question nobody's answered yet: how much new funding is NASA actually putting toward those thruster fixes and that rocket certification? Neither NASA nor Boeing has published that figure. Until it surfaces, you can't tell whether this is a smart insurance policy or an expensive way to avoid admitting the alternative didn't work. So does the math actually hold up — is Starliner worth what NASA's about to spend keeping it alive? That's where the reaction gets loud, and where I've got a read of my own.

So let's hear the reaction, because this split people right down the middle. The most pointed comment didn't come from a critic — it came from SpaceX's own president. This week, on the All-In podcast, SpaceX president and chief operating officer Gwynne Shotwell was asked whether the company would keep flying Crew Dragon longer if demand called for it. She wouldn't commit. 'We're not retiring it today, for sure,' she said. Then she turned the question back on Boeing: 'There is another provider. Boeing has been paid, I think, probably more than we have been paid to develop their human capsule. So we're just gonna let them have some business. They should fly their capsule. The government paid a lot of money for that.' That's a striking thing for a company president to say about a competitor, live on a podcast — not gloating exactly, but pointed. Her framing, boiled down: you already paid for this capsule, so go use it. Now, not everyone's convinced that's the right call. @RossBishop77, an aerospace systems engineer, posted on X: 'NASA are in tight spot when it comes to crewed spaceflight from next decade. It makes sense that they take the lead on an alternative to Dragon, but further funding for Starliner feels like a reward for failure and doesn't sit well with me.' That's the tension in a single sentence — the strategic logic is sound, and the optics are still ugly. Even Ars Technica's own framing of its story, posted on X, couldn't hide the irony. The headline reads: 'NASA to order more Starliner missions and help certify a new rocket to keep it flying.' The fix for a capsule with a checkered safety record is, apparently, more of that capsule. Here's my read. I think NASA's underlying logic is defensible on its own terms. Dissimilar redundancy isn't a luxury when you're launching human beings — it's the entire point of funding two competing capsules in the first place, and getting stranded with a single working vehicle because the other company decided to retire theirs is a genuinely bad position for a space agency to be in. I don't fault NASA for wanting a second option. But defensible strategy and defensible spending are two different questions, and right now, you only get an answer to the first one. Nobody has published what this actually costs — not the price tag on fixing the thrusters, not the price tag on certifying Starliner for whatever rocket ends up carrying it next. And there's a second question worth asking once NASA actually names that rocket: does helping certify Starliner on someone else's vehicle also hand that company a human-rating credential it didn't have to earn entirely on its own? That's a real competitive wrinkle for future NASA crew contracts, and it's not one anyone in this reporting has raised yet. And there's a cost-sharing question buried in this too: NASA helping pay to fix problems Boeing's engineering created isn't inherently unreasonable — dissimilar redundancy is a public good — but it does mean taxpayers, not just Boeing's shareholders, are absorbing some of the price of Starliner's mistakes. That's the kind of detail a program this size should put in a press release, not leave for reporters to chase down. My standard here — and it's mine, not NASA's — is that any company holding this much of the taxpayer's crew manifest owes the public a straight number on what the fix costs and when it actually flies. That standard applies to Boeing exactly as much as it applies to SpaceX. Until NASA and Boeing put a real dollar figure and a real date on this, what you're looking at is a plan, not a program. So if this holds the way it's trending: Boeing keeps a business its own capsule nearly lost for good, NASA locks in a backup it may genuinely need by 2030, and taxpayers are the ones who find out later, after the money's already spent, whether the backup was worth the price. Time for the Hype Check. I'm putting this one at a five. The strategic case for redundancy is real, and Shotwell's comment adds genuine texture to why it's happening now — but there's no funding number, no signed contract, and the whole thing is still sourced to reporters' sources rather than a NASA announcement. Real stakes, real tension, half the paperwork.

If you're the kind of listener who wants the receipts, not the press release, that's the whole reason this show exists — follow Space Stakes wherever you get your podcasts, and if today's episode helped you make sense of the Starliner situation, send it to the one person in your group chat who still thinks this is just a rocket story. This has been Space Stakes, an AI-voiced podcast, created and built by a real human using today's cutting-edge technology. Nothing you heard on this show is financial advice. I'm Brian Lampert, and I'll catch you all tomorrow — take care!

I also host Concrete Compute: a daily briefing on the AI buildout. The datacenters, the megawatts, and who actually pays for them. Find it wherever you get your podcasts.