LandSpace Stuck the Landing — Now Comes the Hard Part

Today on Space Stakes: a private Chinese rocket company just pulled off something only two other organizations on Earth have ever done — landing an orbital rocket booster on its own legs. So is the reusable-rocket club about to get a member with real staying power, or is one clean landing still miles from an actual business? Before that, in the headlines: Rocket Lab signs up to test its hardware against the very military network SpaceX is building, NASA welcomes Blue Origin and Firefly onto a new payload-processing contract, a nuclear battery from Zeno Power books its ride to the lunar surface, and SpaceX finally drags a Starship home after more than three weeks adrift in the ocean. Welcome back to Space Stakes, your daily brief on the business of space. It's Wednesday, August 19, 2026. Let's get into it.

First up, Rocket Lab. The company says it'll fly one of its Photon spacecraft in 2027 to test a connection into the Space Force's Space Data Network — think of it as the military's attempt at a data superhighway in orbit, moving information satellite to satellite and down to the ground over laser links instead of radio. That backbone itself is being built by SpaceX, which landed a two-point-two-nine-billion-dollar contract back in May to construct it. So the Space Force turned around and put twelve million dollars apiece into five other companies — Rocket Lab, Amazon LEO for Government, Lockheed Martin, Northrop Grumman, and York Space Systems — as part of a sixty-million-dollar effort to prove that satellites built by SpaceX's competitors can actually plug into SpaceX's network. That's the government explicitly hedging against single-vendor lock-in. SpaceNews reports the specifics here, and we haven't independently confirmed the contract figures beyond their reporting. But the shape of it is clear: the Pentagon wants interoperability, not dependency, and 2027 is when we find out if Rocket Lab's optical comms gear can actually talk to a network somebody else designed.

Next, NASA's ground game. The agency has added Blue Origin, Firefly Aerospace, L3Harris, and All Points Logistics onto its Spacecraft Processing Operations Contract — a mouthful that basically means who gets to prep satellites and payloads before launch, checking them out, fueling them, packing them into the rocket, at sites including Vandenberg Space Force Base. The contract runs through February of 2033, with a ceiling of one hundred million dollars. Now, that ceiling is shared across four companies over roughly seven years, meaning no single one of them is guaranteed anything close to the full amount — it's a maximum they're now eligible to compete for as individual task orders come up. Firefly specifically said this lets it bid on Vandenberg processing work using a payload facility it already owns. My take: NASA's pre-launch infrastructure is opening up to newer players who used to be locked out by the legacy primes, and what actually matters now is how many of those task orders these companies end up winning.

Now a nuclear battery is booking its first trip to the Moon. Zeno Power has signed a deal with Firefly Aerospace to fly a radioisotope heating unit — a small nuclear-powered heat source — on Firefly's CS-8 lunar lander mission, targeting no earlier than 2028. The device is a five-watt unit Zeno's calling the Survive-the-Night Package, and the job is exactly what it sounds like: keeping a payload's electronics alive through the lunar night, fourteen Earth days of darkness where temperatures drop to around minus one hundred seventy degrees Celsius and most hardware just freezes solid. The fuel is Americium-241, which throws off heat as it decays — no moving parts, no sunlight required. NASA actually built a financial incentive into Firefly's CLPS task order for whoever can prove survive-the-night capability, and to collect it, Zeno's unit needs to phone home after a hundred and twenty hours of lunar darkness. Zeno CEO Tyler Bernstein points out the fuel itself has a half-life of roughly four hundred thirty-two years, and the real pitch is years of lunar-surface operation down the road, assuming the lander launches on schedule — this single demo is just the opening proof point.

And SpaceX just pulled off a recovery nobody planned for. The Starship from Flight 13 splashed down in the Indian Ocean back on July 24th, and normally these vehicles tip over and burn up not long after landing — that's just been the pattern on every previous flight. This one didn't. It stayed intact, and a recovery ship spent roughly three and a half weeks towing it hundreds of miles across open ocean to Christmas Island, an Australian territory, where SpaceX confirmed its arrival on Tuesday. Engineers are now heading out to inspect the hardware in calmer water before anyone figures out how to get a rocket over a hundred and seventy feet long back to Starbase in South Texas — more than ten thousand miles away, entirely by sea, because it's too big to fly anywhere. SpaceX's Falcon 9 program showed the value of physically inspecting recovered hardware instead of just relying on sensor data, and getting an intact Starship back means engineers finally get their hands on a ship that's actually flown — real value, but only once SpaceX publishes what they actually find.

A quick and genuinely nice one: ESA astronaut Sophie Adenot became the first Frenchwoman to complete a spacewalk, working outside the International Space Station alongside NASA's Anil Menon to fix a piece of station hardware. The French diplomatic network in South Africa flagged it on X — @FrenchEmbassyZA wrote, quote, 'This is a historic first: Sophie Adenot will become the first French woman to perform a spacewalk,' end quote. The pair got the failed antenna off, but ran out of time to bolt the replacement in, so a follow-up spacewalk is now on the books to finish the job. Small milestone in the scheme of contracts and constellations, sure, but worth pausing on — every one of these firsts is somebody's actual life's work showing up outside an airlock, and France just got one of its own.

Last quick hit, and this one's murkier. The Pentagon is expanding its research funding into technology that can grab dead satellites and pull them out of orbit — publicly framed as debris cleanup, keeping increasingly crowded orbits usable. Breaking Defense reports the expansion, and we haven't independently confirmed further detail beyond their reporting. The same maneuvering hardware that can rendezvous with your own dead satellite to remove it can just as easily rendezvous with someone else's live satellite to disable it — debris removal and counterspace weaponry are, mechanically, close to the same capability wearing different mission patches. That's simply the physics of proximity operations in orbit, and it applies to any program doing this kind of work. The open question, and the one Congress and allied governments will actually want answered as this funding grows, is which legal box this technology gets filed under: civil debris mitigation, or a counterspace capability that happens to also clean up junk. Right now, nobody's published that answer.

Our main story today, and I'm naming it straight: the landing club gets a third member. On Tuesday, a private Chinese company called LandSpace flew its Zhuque-3 rocket for the second time ever — and this time, the first stage came back. It launched out of Jiuquan's Dongfeng Commercial Space Innovation Test Zone in the Gobi Desert, at seven thirty-five p.m. Eastern. A hundred and thirty-seven seconds in, the first and second stages separated, the upper stage carried on toward orbit, and the booster flipped around, fired its engines, deployed landing legs, and touched down vertically about two hundred forty miles downrange, in Minqin County, Gansu Province. Globally, that makes LandSpace the third entity ever — after SpaceX and Blue Origin — to land an orbital rocket booster on its own legs. Now, if you're new to why that specific move matters: landing a booster instead of throwing it away is the entire premise behind cheap, frequent spaceflight. Every Falcon 9 recovery is SpaceX not having to build a brand-new nine-engine rocket stage from scratch, and that saved cost is a big part of why SpaceX now launches more than anyone else on Earth, by a wide margin. A leg landing is hard specifically because you're flying a rocket backward through its own exhaust and asking it to hit a precise spot while slowing from thousands of miles an hour down to zero. Only three organizations on the planet have now managed it. LandSpace failed at exactly this last time. The first Zhuque-3 reached orbit back in December but lost the booster during the final phase of its landing burn. Eight and a half months later, they came back with actual fixes — cutting the number of engines used during the landing burn to simplify the system, adding what they call a predicted landing point to the onboard safety software, and reinforcing thermal protection against reentry heating. Those changes worked on the very next flight. Worth placing this in context, too. This is China's second successful booster recovery in about a month, after the state-owned Long March 10B caught its booster with a net on a drone ship at sea. LandSpace's landing is the first time a Chinese company has done it the SpaceX way — legs, not nets — and it came from a private company, not the state program. The rocket itself is a stainless-steel, methane-fueled, two-stage vehicle that can carry twenty-one metric tons to low Earth orbit expendable, or up to eighteen point three metric tons flying the recovery profile — numbers that put it in the same weight class as SpaceX's Falcon 9. The second stage, meanwhile, delivered a satellite called Honghu 03 into orbit, built by LandSpace's own satellite manufacturing arm, which raised over a hundred ninety-one million dollars just last month — rockets and satellites, both owned by the same corporate family, which is the vertically integrated bet LandSpace is actually making.

So what does LandSpace actually say about what comes next? In its post-mission statement, the company said, quote, 'This mission lays a solid foundation for subsequent recovery, maintenance, reuse flights, and commercial applications,' end quote. Notice the order of those words — recovery, then reuse. LandSpace itself says it's moving from what it calls the recovery technology verification phase into engineering-scale reuse verification. That's a real and specific claim, and it's also, pointedly, not a claim that they've reused a booster yet. Earlier this year, LandSpace said it wanted to fly this rocket in the second quarter and target refurbishing and reflying that same recovered booster by the fourth quarter. In today's statement, they gave no reuse timeline at all — a quiet omission worth noticing. And that's exactly where the skepticism has to live. SpaceX didn't turn booster landings into a business by landing one booster — it took years and dozens of attempts before reflying boosters became routine, and the real economic payoff only showed up once the same hardware was flying again and again without months of turnaround in between. LandSpace has now landed a booster successfully once, after failing once. That's a genuine engineering achievement — fixing a specific failure mode, in under a year, on a company's second-ever orbital flight, is not nothing. But recovery and reuse are different claims, and right now, the company's own language only supports the first one. Zoom out, though, and the bigger context is hard to ignore. This landing came the same week Ars Technica argued the U.S. is underestimating how fast China's space program is closing distance on ground that used to belong to SpaceX alone. I think that argument gets stronger, not weaker, with this specific result — because this wasn't the state-owned prime contractor, CASC, pulling it off. CASC attempted its own booster recovery with the Long March 12A a few weeks after LandSpace's first try, and that one also failed. It's a private Chinese company, one with a pending IPO bid on Shanghai's STAR Market, that just became the first Chinese entity — state or private — to stick a leg landing. That tells you something about where the real velocity in China's launch sector currently sits. This is also a moment to apply a standing view of mine: I don't think this show should root against SpaceX's success — the company earned its position through actual execution. But when one company is the launcher, the operator, and often the only realistic ride for years running, competition showing up anywhere, including from a Chinese private company, is worth taking seriously rather than dismissing as a rounding error. My read on why this matters beyond the scoreboard: Zhuque-3 is explicitly built to support China's two megaconstellation projects, Guowang and Thousand Sails — the country's answer to Starlink. A reusable, Falcon-9-class rocket is the piece of infrastructure that makes launching thousands of satellites financially survivable instead of financially ruinous, which is the exact math SpaceX solved for itself first. I don't think this changes SpaceX's position at the top of the launch market this year, or probably next year either. But it's one more data point that the gap between only SpaceX can do this and multiple actors can do this is shrinking faster than a lot of people in this industry assumed even twelve months ago. Time for the Hype Check. I'm calling this one a six. The landing itself is completely real, technically serious, and a legitimate global first for LandSpace — no argument there. But the marker that actually moves the needle on reusable launch economics is the same one the skeptic case demands: the same booster, refurbished, flying paying customers again, without a turnaround measured in months. LandSpace hasn't shown us that yet, gave no new timeline for when they will, and until that number shows up, this stays a genuine engineering win still wearing the costume of a business breakthrough.

If LandSpace turns this landing into an actual reflight next, the company with the most to gain is LandSpace itself, heading into that STAR Market IPO — and the ones who'll have to answer for it are anyone still calling booster reuse a moat only SpaceX can hold. Follow Space Stakes wherever you get your podcasts so you don't miss where this goes next. 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!