Inside the Room: How Starship Went From No-Go to Orbit

"On anomaly two, we just got confirmation that we are no-go for orbit."

That was the call inside SpaceX mission control a few minutes into Starship's first orbital flight. One of the ship's six engines had shut down early. The plan written before launch was clear: if the vehicle is unhealthy, do not commit to orbit, and dump the ship in the Indian Ocean. The 26 Starlink V3 satellites in the payload bay would go down with it.

Minutes later, Starship was in orbit anyway.

I posted SpaceX's film about this flight on my channel. When I covered the webcast back in September, all we could see was the hosts going quiet and then coming back with a go. This film puts you in the room where that call flipped, and it is the clearest look I've seen at how SpaceX actually makes risk decisions.

The engine that almost ended it

Before launch, the mood was simple. "This is our first orbital mission," one engineer said. The first big objective was "delivering useful payloads to orbit," real Starlinks that would join the constellation, not dummy mass.

Liftoff was clean. The film’s loop reports all 33 Super Heavy engines still running at T+1 minute; SpaceX’s recap says one later shut down on ascent and 31 relit for boostback. On the ship, engine 6, a vacuum Raptor, shut down early. A controller in the film said it appeared a controller reset drove that shutdown.

The film is honest about what that meant. "The moment we lost the engine, we all kind of knew this was the end of the mission for the ship." There are only about 17 minutes in that first coast to do anything. In the Starlink room, people were already saying it would be "the end of the show."

One question from Elon

Here is the part I keep rewatching. The team had done the conservative thing. They wrote a rule ahead of time and they were following it. Then, in the words of one engineer, "Elon really questioned the requirements and asked us, 'Is there anything actually prohibiting us from going forward?'"

That is the SpaceX operating model in one decision: question the requirement, then check the vehicle. SpaceX’s flight recap says the team assessed the ship and determined that the three sea-level Raptors needed for the later burns were healthy before giving the go for orbit.

Another engineer laid out the logic. "A development program has a very different risk concept to a production program." On launch day they collect risks from engines, avionics, computers, plumbing, the pad, the tower and the catch arms. "But the goal is always to take the risk in the place where you can learn the most."

According to an engineer interviewed in the film, Jake was the final go for engine systems and recommended to Elon that they keep going. On the loop, the team requested that the hold be lifted so the ship could proceed to orbit with the engine-out condition.

The film describes a final go/no-go poll shortly before the orbital insertion burn. Stage 2, go. Raptor 2, go for orbit. Burn two startup. Burn two shutdown. "No alerts or alarms on the board. We're in orbit."

The payload that makes Starship a business

This is why the call mattered. All 26 satellites deployed, and all 26 connected to the network.

A Starlink engineer in the film called each V3 satellite “basically 10 times as performant” as a V2 and described a hypothetical V2-to-V3 fleet replacement as a network 10 times as capable. The precise published comparison is capacity, not a blanket speed or performance multiplier: about 1 Tbps of downlink capacity per V3, roughly 10 times V2, and 160 Gbps of uplink capacity, roughly 22 times V2. The engineer also projected that about 12 Starship launches a year could match Starlink’s best year of network growth, and said the technology could support “on the order of half of all the internet traffic that is on Earth today.” Those are the speaker’s scale-up projections, not capacity demonstrated by this flight.

“It’s not just a test ship now,” one of them said. “It’s a true mission.” The engineer described future launches as generating real value. This flight demonstrated a first operational payload delivery, a milestone toward the production-and-reuse cycle SpaceX is trying to build.

Chile or Hawaii

Getting to orbit was only half the job. The mission had two planned ocean splashdown zones: the primary off Chile and an early contingency north of Hawaii. An engineer in the film called the primary a “nine-hour mission”; SpaceX’s published preflight timeline actually ran to about 9 hours 50 minutes. The film’s team discussion describes having about 30 minutes after payload deployment to choose the early-return option.

Their worry list was blunt. Another engine shutdown. Helium. Reentry. And the one that scares every engineer: the real risk, as one put it, is missing something because nobody is looking at the right thing.

Starship's coasts are long for an orbital stage, and its propellants are cryogenic. Heat from the environment keeps trying to boil them off, so every extra hour is liquid you might not have for the deorbit and landing burns. They committed to the early Hawaii contingency.

Everything rides on the deorbit burn

Most upper stages burn up when they come back. Starship is meant to survive reentry, which means it cannot be left in orbit to come down wherever it wants, possibly over land. "The worst case scenario is that we'll stay in orbit," one engineer said. "None of that mattered if you can't get it back."

The film includes an engineer describing Raptor Stand Bravo, where two booster modules reproduce a slice of the vehicle’s aft end so engines can be installed and tested as they are on the vehicle. Elsewhere in the sequence, engineers describe testing the conditions needed for a reliable ship deorbit-engine relight. The narration does not establish that the booster-module rig is that deorbit test setup. They also recount starting the first Raptor tests horizontally to avoid building expensive flame-trench infrastructure before knowing their engine would work. One engineer says that, after repeated cold flows and spin primes, they finally lit it the evening Elon came to town.

Then the real thing. "Deorbit burn startup. Deorbit burn shutdown. Nominal descent." Someone in the room yells. "We're coming home."

The quiet win

Two more details point toward reuse. An engineer in the film says two tiles were removed from Ship 40 to fly again on Ship 41, describing this as Starship’s first tile-reflight demonstration. He says he would have bet a year’s salary against that prospect on the day Ship 40 flew. That is reuse of two tiles, not a reflight of the complete ship or proof that the entire heat shield is reusable.

And right at the end, the team ran a rehearsal poll for a ship return to launch site. Flaps, go. Structures, go. Aero, go. "If this had been an RTLS mission, the vehicle would continue to return to launch site."

That is a step toward the next attempt. SpaceX now says Flight 15 is planned to return both Super Heavy and Starship to the launch site for catch, subject to vehicle health and regulatory approval. Flight 14 demonstrated orbital payload delivery and a controlled deorbit; it did not demonstrate recovery and reflight of the complete rocket.

"After this," one of them joked, "the robots will have to take over."