A Starship has never come home. Every single one that has flown has ended as debris in an ocean. Something shifted at SpaceX this week. On the surface, it was quiet. No static fire, no rollouts, but three separate stories when you line them up all point at the same thing. A drone ship reassigned, three Starships being built at once, and orbital launch mount starting to come together. SpaceX is getting ready for something it has never done before, and it is closer than anyone thinks. My name is Felix. Welcome to What About It. Let’s dive right in. Starship updates. Let’s start with the biggest story at Starbase this week because the production pipeline is in a genuinely remarkable state. Ship 39 is done with testing. 60-second static fire complete, back in Megabay 2, ready for flight 12. Ship 40 is fully stacked. Thermal protection system tile work looks complete. From everything we can see, this vehicle is almost ready to roll out
[00:01:02] to Massey’s for its own cryo test campaign. And ship 41 is in active stacking right now. The forward dome is being integrated in Megabay 2 this week. That is the piece that separates the tank section from the payload bay. Three ships, three different stages of readiness, three distinct flights lined up. Now, here is where this gets interesting. Ship 39 is flying flight 12 and will do a soft water landing somewhere in the Indian Ocean, just like many ships before [music] it. That is expected. But, look at what comes after. Ship 40 flies flight 13. According to the FCC filing we covered a few weeks ago, flight 13’s ship will fly an orbital trajectory and may or may not return to the landing site. That language matters. May or may not return is not the language of another water landing. That is the language of a possible catch without a catch. If ship 40 returns, and if SpaceX decides to attempt a catch, ship 40 becomes the
[00:02:01] first Starship ever to come home to the tower. And if ship 40 doesn’t make it, ship 41 is already being built, waiting right behind it. Here’s what’s remarkable about this. For years, SpaceX has been iterating on ship designs, losing vehicles to re-entry, losing them to in-space anomalies, losing them to static fire accidents. And now, on the cusp of version 3 flying for the first time, they have a production line running at a pace where three ships are simultaneously in different stages of flight readiness. That is an operational program trying to emerge. It just depends on whether ship 39 makes it to the Indian Ocean intact and according to plan. If everything works out, one of these ships will finally return home. And the data SpaceX collects from the first caught ship reshapes the entire program. For the engineers, one thing is more important
[00:03:00] than a data stream during flights. Analysis will be way more efficient that way. Now, here’s a story that surprised a lot of people this week, and on the surface, it might not sound that exciting. But, stay with me because what this actually signals is huge. SpaceX officially announced that after 156 successful Falcon 9 landings, Just Read the Instructions, one of SpaceX’s two East Coast autonomous drone ships, is being fully dedicated to Starship operations going forward. So, here’s the first question. Why would SpaceX make this change? Let’s walk through the options. The obvious first thought is that Just Read the Instructions is being repurposed to be a Starship landing site at sea. After all, that’s what it does for Falcon 9, right? But, with Starships, that is not that easy. You cannot catch a Starship on a drone ship. The catch requires MechaZilla chopsticks mounted on a tower, and you cannot fit a tower that tall on a drone ship barge.
[00:04:01] The structural loads alone would be impossible. Catch infrastructure lives on land, or at least the barge would need to be way bigger to change that. The oil rig idea SpaceX had and still has is a different thing altogether. It is not a barge, it’s an oil rig, literally a huge difference. So, if it’s not for landings, what is it for? The answer is straightforward. Just Read the Instructions is joining You’ll Think Me Later to support the transport of ships and boosters from Starbase in Texas to Kennedy Space Center in Florida. Two barges moving finished vehicles from the factory to the coast by sea. And we have this directly from SpaceX. Kiko Donchev, SpaceX’s VP of launch, posted a clarification that explains the whole thing. With Pad 39A becoming a primarily Falcon Heavy and Starship pad, SpaceX does not need two operational drone ships on the East Coast to maintain the Falcon manifest. A shortfall of Gravitas can support a 4-day Falcon 9 launch
[00:05:01] cycle alongside RTLS missions, which is exactly how the West Coast already operates with one drone ship. So, Just Read the Instructions becomes available. And instead of retiring it, SpaceX is putting it to work on the biggest logistical challenge still standing between Starbase and Cape Canaveral. But, why wouldn’t they just manufacture ships and boosters at the Cape? Robert’s Road? No. Since SpaceX started building at Robert’s Road, it has looked way more like a refurbishment and maintenance center to me. The Robert’s Road infrastructure is way smaller than what SpaceX has at Starbase. It might just be that, at least initially, the idea is not to have two full production sites for Starships. Later, that might follow, but for now, it looks like a lot of the work will happen at Starbase, Texas. And that’s what you need the second transport barge for. Back and forth feeding two launch sites with Starships. This is the infrastructure layer of the Starship launch network becoming real.
[00:06:00] Two barges running continuously carrying 70-m boosters and 50-m ships from the factory in Texas to Florida. For the Starship operational cadence SpaceX wants to hit, this capacity has to exist. And now it does. One drone ship goes from catching Falcon 9 boosters to ferrying Super Heavies and ships across the Gulf. That’s the quietest and most significant strategic announcement of the week. And to make all this even more real, over at Starbase, more and more parts for the new Pad 1 orbital launch mount are showing up on site. Flame bucket pipe segments, deluge water manifolds, hardware being staged for what is very clearly a parallel OLM build. Soon, SpaceX will begin assembling the new Pad 1 OLM square at the Sanchez production site. The refurbished Pad 1 is going to be nearly identical to Pad 2. We already know what it will look like. So, the real question is, how long will this take? Some solutions are extremely good and done in
[Sponsor: Incogni — data-broker removal service. Code FELIX, 60% off annual.]
[00:08:00] Take your personal data back with Incogni. Use code Felix at incogni.com/felix, and get 60% off an annual plan. Link is in the description. All right, back to Pad 1 and its timeline. For the Pad 2 OLM, the answer was about 7 and 1/2 months from the first corner section arriving at Sanchez in late September 2024 to the OLM being lifted into position on May 12th, 2025. Now, here is why I think Pad 1 goes faster. First, this is not a first-time build. SpaceX has done this before. Every process, every jig, every sequence the crew already knows. Second, there is no design iteration. Pad 1’s OLM is a near identical copy of Pad 2. No experimenting, no adjusting, just build. Third, the parts are already being staged ahead of the assembly phase, which means the supply chain is ready when the welding torches fire up. The Pad 2 OLM took 7 and 1/2 months. I expect Pad 1 OLM to take about 6 months, possibly less. If SpaceX begins assembly
[00:09:02] in the next two to four weeks, which is mid-May to early June, 6 months later lands in mid to late November. This also shows us that there is almost no chance that Pad 1 will be operational this year. Q1 2027 looks realistic, but this isn’t the only pad SpaceX is building for Starship. Fresh satellite imagery of the Space Coast pads is now available from Harry Stranger and his website spacefromspace.com. The images taken on April 18th reveal things we simply cannot see from our helicopter because we are not allowed to fly directly overhead. And what the satellite images show at SLC-37 is genuinely fascinating because the real construction doesn’t match the environmental impact statement build plan at all. Take a look at this. The tower segments for the first of two planned MechaZilla towers are staged in a location that does line up with the EIS plan. Tower number one will likely end up approximately where it was
[00:10:01] supposed to go. Tower number two is a different story. The EIS plan put tower two in a location that’s currently a large stormwater pond. And from the imagery, that pond doesn’t look like it’s going to be relocated anytime soon. Okay? In fact, both current stormwater ponds on the site are in different positions than what the EIS plan showed. Yeah, that alone tells you this construction is deviating from the original plan and there is something else interesting. A graded foundation square, roughly the size you’d need for a launch pad, sits smack in the middle of the site. Piles appear to have been driven into the ground to stabilize it. In the EIS plan, this spot had nothing but roads crossing through it. Now, it’s clearly being prepared for something substantial. Just above that, four more prepared foundation spots. My best guess is tank farm infrastructure, but that is speculation. Two more foundations appear
[00:11:00] to be on the west side of the site and honestly, I can’t tell what they are for. Yeah. And then there is the stumpy concrete building at the north tip of the pad that we talked about in the last episode, right next to the large crane. From the satellite angle, it looks substantially enough to be a tower foundation and its position makes sense for the second tower opposite to where the first tower is being built. In the EIS plan, this exact spot was supposed to be a storm surge pond. In reality, it is a building. Here’s the bigger lesson. The EIS plans we see published publicly are at best examples of what a site could look like. They are not blueprints. What SpaceX actually builds evolves as the engineering gets refined, as the regulatory approvals come through and as they learn from pad two. Now, here is a very rare and special closer look. The close-up images we got of the Raptor 3 engines during the April 17th booster lift reveal something that looks incredible. Visible 3D printing layers
[00:12:02] on the engine surfaces. Actually visible to the eye in focused photographs from ground cameras. This is the direct result of SpaceX taking 3D printing all the way. Almost the entire engine is printed now. That is what enables Raptor 3 to have cooling channels integrated directly into the outward facing parts of the power head, so the engine itself can handle re-entry heat without the heavy external shielding that Raptor 1 and Raptor 2 required. We already talked about the minimal shielding a couple of episodes back, but the visible print lines are the visual proof of how it was achieved. And remember what Musk said in the Everyday Astronaut interview with Tim Dodd. With Raptor 3, repairs literally require cutting the engine open. The number of separate parts, flanges, fittings, anything that used to be bolted together has been so drastically reduced that the engine is essentially one continuous printed structure. You cannot unbolt it. You
[00:13:01] have to cut. That’s the direction every Raptor is now going. Simpler, lighter, printed as one piece.
[00:14:00] [Stoke Space segment begins]
2026 has already proven to be an amazing year for us space flight enthusiasts, but there is one thing that’s still never been done. Nothing short of the holy grail of reusability. Fly to orbit, land both stages, refuel and do it again. If you are a Y regular, you likely have a clear picture in your mind. One of SpaceX catching both Starship stages with Mechazilla towers. Gwynne Shotwell raising her arms in slow motion. SpaceX, you’re incredible. But a Washington state startup is working on the same goal. Stokes Space. Granted, Stokes Space is not building a giant. Nova is just a medium-lift rocket, but it’s filled with innovation to bring home
[00:15:02] both stages intact and fly them again. To understand why Nova matters, you have to understand the next milestone in the reusability revolution. Recovering a first stage is the easy part. Well, relatively easy. A second stage re-enters at incredibly higher speeds, much higher than those experienced by the booster. The problem here is that high speeds translate into massive amounts of energy that must be dissipated upon re-entry from orbit. Physics. We’re looking at 10 to 15 times more energy per kilogram than a first stage return. At that speed, there is no easy solution. This is why SpaceX is working so hard to find the right heat shield solution for Starship. It is a problem that needs to be solved before full reusability can be achieved. Today, only three vehicles in history have ever survived an orbital re-entry and flown again. The first is the space shuttle.
[00:16:00] Its reuse was limited heavily by its famously fragile tile system that required tens of thousands of labor hours between flights. The second one is SpaceX’s Dragon capsule. This one also uses an ablative heat shield that has to be refurbished and at least partially replaced after each flight. And then there is the X-37B, a highly classified space plane that has landed, undergone refurbishment and been flown again. That’s it. Starship itself, so far it’s proven that it’s able to survive re-entry with varying degrees of damage. It still has a good way to go before it’s rapidly reusable. But here is the problem. Throwing the upper stage away each flight is roughly like scrapping a Boeing 737 after every New York to London flight. Stokes Space was founded in 2019 specifically to solve this. Its idea sounds compelling but outlandish at first. Why not build an engine that’s also the heat shield? Because that’s
[00:17:00] undoubtedly the biggest technical hurdle. The heat shield. What if it could be helped? Let’s take a closer look at this remarkable rocket. Nova’s upper stage is about 4 m wide and tapers toward the top, vaguely capsule-shaped. Its base is a curved metallic dome ringed with 24 small thrust chambers arranged like numbers on a clock face. And here comes the twist. Together, they form a single engine called Andromeda 2. It’s burning liquid hydrogen and liquid oxygen, the same high-performance propellant combination that powered the upper stages of Saturn V, the space shuttle and SLS. And what’s the intriguing part? That metal dome at the base of the stage serves two purposes. In flight, it’s a nozzle system. On the way home, it becomes the heat shield. Embedded inside the metal are tiny [cooling channels]. Before the cryogenic hydrogen ever reaches the combustion chambers, it
[00:18:00] flows through those channels. Like water through a radiator, it’s soaking up re-entry heat. The hydrogen boils and expands, then goes on to spin the turbopumps and feed the thrusters. Effectively, the same propellant that will power the engine first cools the heat shield. And the heat itself becomes fuel for the pumps. It’s self-regulating. The more heat is absorbed, the more fuel will be supplied to the thrusters. During Stokes flight tests, the cooling was so effective that frost visibly formed on the outside of the shield while the thrusters fired directly underneath it. Nova’s second stage, in effect, is a stainless steel skillet pan cooled by its own fuel holes. There’s another elegant consequence of the 24-chamber ring. Because the thrusters surround a central dome, there is nothing to gimbal. It eliminates heavy, leaky swivel joints. To steer, Nova simply throttles some thrusters harder than others. It’s the same principle that lets a quadcopter
[00:19:00] drone hover without any thrust vectoring. One rotor spins faster and the drone flies a curve. That makes the whole system lighter, simpler and built for hundreds of cycles. How about the lower part of the rocket? Nova’s booster is no slouch, either. It’s powered by seven Zenit engines, each producing 45 tons of thrust burning methalox, the propellant combination of choice for the new generation of reusable boosters. What makes Zenit remarkable isn’t its size, but its plumbing. It runs a full-flow staged combustion cycle, the most efficient rocket engine cycle ever developed. At this point, only one engine on Earth, SpaceX’s Raptor, has ever flown this cycle. Here’s why this is so significant. Every rocket engine needs to spin turbopumps to force propellant into the the chamber. Older engine designs waste fuel to do this, venting smoke overboard. Full flow throws everything, both the oxidizer side exhaust and the fuel side exhaust,
[00:20:01] back into the main fire. Nothing is wasted. The propellant arrives as gases. They mix like a gas stove instead of a splash of lighter fluid, and the turbines run cooler and live longer. That last part is the whole point. Zenit is built to be flown over and over without rebuilding. Fully stacked, Nova stands about 40 m tall. That’s far smaller than Falcon 9’s 70 m. It can put about 3 tons into low Earth orbit in its fully reusable mode, or around 7 tons if flown expendably, with 2.5 tons to geostationary transfer orbit and 1.25 tons to the moon. While this payload capability feels small, there are customers who simply don’t require more. In September 2023, on a dusty test pad in Moses Lake, Washington, Stoke briefly hovered a prototype of its upper stage about 10 m off the ground, translated sideways, and landed it on three spindly legs, all inside 15 seconds. Today,
[00:21:03] Stoke is preparing for the first full flight test. Nova’s home base is Slick 14 at Cape Canaveral. That’s the historic Mercury pad, the birthplace of crude US space flight. Stoke broke ground in October 2024. Nearly 4,000 truckloads of soil laid the original Mercury era concrete has been recycled into new foundations. A 37-m umbilical tower is up. Propellant farms are installed, and the horizontal integration facility is built. By mid-February 2026, Stoke posted, “Launch pad complete.” A first orbital flight originally targeted for 2025 has slipped to early 2026, but realistically, we’re looking at later in 2026. The debut will not attempt a booster landing. It will fly into a heliocentric orbit carrying, among other things, a Celestis deep space memorial payload. Landing and reuse get tested on later flights. If
[00:22:00] Nova flies as advertised, it can quickly become a lucrative launch vehicle. The demand is real. Modern satellites are often smaller, and constellations are generating a structural need for frequent, cheap, medium-lift launch vehicles. And that’s it for today.