NASA ends its far-side moon-lander deal as delays push the landing toward 2031

Space 4 min 80 sources

NASA and Draper mutually cancelled a $73 million contract to land science instruments on the moon's far side, after redesign after redesign pushed the target from 2025 to 2031 — and the agency is already spreading the same science across other landers.

Key takeaways

NASA and Draper cancelled a $73 million contract to land science instruments on the moon's far side after eight years of redesigns pushed the landing from 2025 toward 2031.
NASA paid $43 million for milestones the company did complete, and plans to fly the finished instruments on a future lander — part of its strategy of buying many cheap commercial moon rides and dropping the ones that stall.
Elsewhere: SpaceX aborted its 13th Starship flight when engines failed to ignite, and satellite investment hit a record $8.1 billion in the first half of the year.

NASA has cancelled one of its moon landers before it ever left the ground. The agency and the defence contractor Draper have mutually agreed to end a task order — designated CP-12 — to fly three science instruments to the far side of the moon [1]. Draper’s Japanese-backed subcontractor, ispace-U.S., revealed the termination in a statement on July 14; Draper confirmed it on July 15 and pointed questions to NASA [1][2].

The reason NASA gave is plain: the lander kept being redesigned, and the finish line kept moving. When the deal was signed in July 2022, the launch was set for 2025 [1]. A 2023 redesign to fit NASA’s payloads pushed it to 2026. A new engine in May 2025 pushed it to 2027. In March, ispace changed engines again and merged its American and Japanese lander designs into one — pushing this mission to 2030, behind two other flights [1]. NASA said future milestones “were projected to take years to complete, leading to a landing date in 2030 or 2031” [1].

What NASA paid for, and what it walked away from

The contract was worth $73 million. NASA had paid Draper $43 million of it — money released only as the company completed specific milestones, not handed over up front [1]. That structure matters: when NASA ended the deal, it stopped paying for a lander that would not arrive for five or six more years, and kept what the earlier payments had already bought.

The mission itself was scientifically real, not a throwaway. It was headed for the Schrödinger Basin on the moon’s far side, carrying three experiments: a seismometer package to feel moonquakes, a probe to measure heat flowing up from the interior, and an instrument to read the moon’s faint electric and magnetic fields [1]. The far side is prized because Earth’s radio noise never reaches it — it is the quietest listening post in the inner solar system.

NASA says it still wants that science. It plans to fly the already-built instruments on a future lander “at the earliest opportunity” [1]. But there is a catch: on June 30, NASA awarded nearly $600 million for four new lander missions to Astrobotic (now Voyager Lunar Systems), Firefly Aerospace and Intuitive Machines — and at least some of those are going to the near side, which cannot host a far-side experiment [1].

Why a cancelled lander is not a failed programme

The Draper mission was part of CLPS — Commercial Lunar Payload Services, NASA’s approach of buying rides to the moon from private companies at a fixed price instead of building one government lander itself [1]. The idea is to place many small, cheap bets and accept that some will not pay off. Firefly’s Blue Ghost lander touched down successfully; others have stumbled. Draper’s is not the first to end before launch — Orbit Beyond dropped out in 2019, and Masten Space Systems went bankrupt in 2022, both after winning task orders [1].

Seen inside that model, a termination is not the system breaking. It is the system doing what it was built to do: cut a stalled bet loose and move the money and the science to landers that can fly sooner [1].

The rest of the week off Earth

SpaceX called off its 13th Starship flight at the last second on July 16, when some of the Super Heavy booster’s engines failed to start and an automatic abort kicked in [3]. Elon Musk said two Raptor engines would be removed and replaced, with the next attempt “early next week” [4]. This flight — the second of the upgraded Starship V3 — is carrying 20 working Starlink satellites it will deploy in space, not the dummy weights of earlier tests [3].

The money keeps flowing in. Investment in satellite companies hit $8.1 billion in the first half of 2026, already beating every previous full-year total tracked by the firm Space Capital [5]. The Finnish radar-satellite maker Iceye led the quarter after a $1.2 billion raise to meet NATO demand [5].

And a reminder that the sky still reclassifies itself: NASA’s Jet Propulsion Laboratory reported that a near-Earth object long catalogued as an asteroid is in fact a comet [6]. It looked like a rock, but its motion — tugged by gas quietly venting off its surface — gave it away [6]. Nothing about the object changed; only what we had measured did.

02 · Lesson · why it matters

The savings live in the cancellation

Spreading your effort across many small bets only pays off if you can bring yourself to stop the ones that stall — the upside and the cull are the same decision.

A contract dies quietly

NASA just cancelled a moon lander it had been paying for since 2022. Three science instruments, bound for the far side of the moon, now have nowhere to ride. The agency had already spent $43 million of the $73 million deal, and the landing had slipped from 2025 all the way to 2031.

The easy read is failure — money spent, nothing flown. But the more useful read is how ordinary this was. The programme it belonged to was built to make exactly this decision easy. The cancellation isn’t the system breaking. It’s the system working.

The whole point was to be able to do this

NASA buys moon landings two ways. The old way: build one government lander, pour everything into it, and because everything is riding on it, it can never be allowed to fail — which means it can never be cancelled either.

The new way, called CLPS, is the opposite. Pay several private companies a fixed price to each try. Expect some to make it and some not. One firm, Firefly, landed cleanly. Others went bankrupt or dropped out. Draper’s is simply the latest to end before launch.

The promise of the second way is that it’s cheaper, because the risk is spread across many small bets instead of loaded onto one big one. But that cheapness is a promise you only collect at one moment: when you close the bets that stall. A portfolio you never prune costs just as much as the single giant project you were trying to avoid — you’re just paying for the giant in installments across a dozen names.

The two things we keep treating as separate

Here is the part almost everyone gets wrong. The upside of spreading bets — cheap failures, many shots on goal, no single point of collapse — and the discipline of killing the laggards are not two things. They are one thing seen from two sides.

You cannot bank the first without paying the second. The moment you refuse to cancel a stalled bet, the whole logic quietly inverts: you’re back to the giant project, funded piece by piece, with none of the savings you signed up for.

Yet we adopt portfolios for the upside and treat the cull as a separate event — optional, painful, always postponable. That splitting is the mistake. Most human trouble comes from treating connected things as if they were separate, and this is a clean case. The saving and the cancellation are the same act. Skip the second and you never had the first.

Why the cull is the hard part

If cancelling is the whole point, why does it feel so brutal? Because $43 million already spent screams finish it. Walking away sounds like waste.

Look at how NASA built the deal, though, because the design is doing quiet work. It paid Draper only as the company hit specific milestones — not a lump sum up front. So when NASA ended the contract, it had already received what those payments bought, and it owed nothing for the years of work that would never happen. The remaining $30 million simply stayed in NASA’s pocket.

That structure isn’t a law of nature. Someone chose it, precisely so that stopping would be cheap and clean instead of a fight over a refund. It serves NASA, which gets a painless exit. It serves the companies, which keep what they earned. It serves the science, which can look for a faster ride. A cancellation only feels ordinary when the arrangement underneath it was built to make cancellation ordinary.

Most of us aren’t so lucky, because most of our portfolios have no milestone structure to lean on. A hospital keeps a stalling drug trial alive because so much has gone into it. A company nurses a project everyone privately knows is dead. You keep the side thing you’ve “put too much into to quit now.” And as a taxpayer, your money sits inside this exact space portfolio, funding both the landers that fly and the ones that get cut. We are all holding laggards we can’t bring ourselves to close, for the same reason NASA had to design its way around.

What the seat can’t see

There’s one more thing, and it should keep us honest. Cancelling a stalled bet is never obviously right — it only looks right from far enough away.

The person signing the termination sees the slipping dates and the milestones. What they can’t see is whether the lander they just killed would have been the one that worked, or whether the freed-up money lands somewhere better. And there’s a real cost hiding in this particular cull: the far-side instruments are now stranded, because the landers NASA is funding next are going to the near side and can’t carry them. The quiet science of moonquakes and buried heat waits, indefinitely, for a ride that fits.

So the cull is not wisdom applied to fools. It’s a bet against a bet, made in fog. The people who can’t let go of a stalling project aren’t simply weaker than the ones who can — they may be weighing a loss the crisp rule doesn’t show. Seeing the whole here doesn’t hand you a clean answer about when to quit. It hands you a reason to hold your own answer a little more loosely.

03 · Lab · your turn

The Cull

Run a portfolio of moon landers and feel how cancelling the stalled bets — a decision made in fog — is what frees the money to land one.

04 · Hope · carry this

One lander stopped, but the instruments it was carrying still exist, waiting for a ride that fits. A way of reaching the moon that expects some attempts to fail is, in the end, a way that keeps trying until one lands.