SpaceX flies Starship a 13th time - the one rocket built to do many different jobs

Space 3 min 80 sources

Starship's 13th test flight deployed 20 next-generation Starlink satellites and tested the design NASA has picked to land astronauts on the Moon. It was a good day for a machine meant to do several unrelated jobs at once, and a reminder that one rocket now carries internet, a lunar landing, and a Mars ambition on the same development bill.

Key takeaways

SpaceX flew Starship a 13th time on Friday, deploying 20 next-generation Starlink satellites while the booster splashed down harder than planned.
The same Starship design is being adapted to land NASA astronauts on the Moon, so one rocket now carries internet, a lunar landing, and a Mars ambition on a single development bill.
A busy week also saw China's Gravity-1 launch from a ship at sea, India's Skyroot reach orbit for the first time, and a 30-year-old "asteroid" revealed to be a comet.

A rough splashdown, a clean deployment

SpaceX flew its Starship rocket for the 13th time on Friday, and called it a mostly successful test [17][18]. The giant Super Heavy booster separated from the upper stage, flipped, and flew back toward Earth [30]. Its landing burn went wrong: only 5 of the 13 engines needed appeared to fire, and the booster hit the Gulf waters hard, its cameras cutting out [30]. The upper stage did better. On its arc over the ocean it opened a slit-like door and pushed out 20 next-generation Starlink satellites, one after another, before the ship itself came down [30][45].

The satellites did not stay up. Flight 13 flew a suborbital path on purpose, so the 20 Starlinks fell back to Earth after deploying [30]. This was a rehearsal, not an operational launch. SpaceX had tried to fly on July 16 but aborted at the last second over engine trouble, then waved off a July 23 attempt for weather [30][21]. “Starship is intact, floating in the ocean,” Elon Musk said afterward [30].

One rocket, several unrelated errands

What makes Starship unusual is not a single flight. It is that SpaceX is building one machine to do jobs that have almost nothing to do with each other.

The 20 satellites it carried Friday were Starlink Version 3: large, flat craft with laser links, built to replace SpaceX’s existing internet constellation [30]. Starship is designed to loft 60 of them at a time through that slit-like door, “like a giant PEZ candy dispenser” [30]. That is job one: launch the company’s own internet business.

Job two is the Moon. NASA has chosen a version of Starship as the lander that will carry astronauts to the lunar surface on its Artemis III mission [15][19]. The same basic vehicle that flings out internet satellites is being adapted to set humans down on another world.

The most complicated mission NASA has flown

Artemis III shows how much now rides on that one design. NASA calls it the most complex mission it has ever attempted: four astronauts, three spacecraft, and about two weeks in orbit [15]. A crew will launch aboard the Orion capsule on NASA’s own SLS rocket, then rendezvous separately with two commercial landers, SpaceX’s Starship and Blue Origin’s Blue Moon [15]. Only one crewed mission in history, the Soviet Soyuz T-15 in 1986, has flown people to meet more than one crew-capable spacecraft in a single trip [15].

NASA says the mission is on schedule, though Blue Origin is repairing a launchpad in the meantime [19]. On the flight, astronauts plan to beam 4K video home from Orion using SpaceX’s Starlink laser links [31]: the internet network, again, doing a second job.

A busier, more varied launch week

Starship was not the only vehicle flying. Off the coast near Shanghai, China’s Gravity-1 rocket lifted 9 satellites to orbit from a ship at sea, a record for a solid-fueled launcher of its kind [2][62]. India’s Skyroot Aerospace, a private company, reached orbit for the first time with its Vikram-1 rocket [56]. And a Chinese Long March 3B was struck by lightning seconds after liftoff yet still delivered its payload, in a photo that traveled widely [53][23].

Different countries, different companies, different rockets, all chasing the same thing: cheaper, more frequent ways to reach space.

The asteroid that was hiding a tail

The quiet find of the week came from a second look at an old object. A near-Earth body watched for nearly 30 years turned out to be a comet, not an asteroid [1]. Astronomers noticed its path drifting in a way gravity alone could not explain. They traced the drift to a faint tail of escaping gas, the tell-tale sign of a comet gently pushed off course by its own venting [48]. The method matters for planetary defense. An icy object that quietly nudges itself is harder to predict than a dead rock, and spotting that difference early is how you know what you are really tracking [1].

02 · Lesson · why it matters

Why one machine that does many jobs beats many machines that each do one

One costly thing that can do several different jobs spreads its price across all of them, so each job comes cheap.

Friday’s rocket ran three errands at once

Look at what Starship actually did, and what it is being built to do. On Friday it flung out 20 internet satellites for SpaceX’s own Starlink network. The same design is being fitted out to land NASA astronauts on the Moon. The company says the whole point of it, eventually, is Mars. Launching internet, landing on the Moon, reaching another planet: three jobs that have almost nothing in common, all resting on one machine.

That is not how spaceflight used to work. For decades, a hard new job meant a new vehicle, designed and paid for from scratch. Here, one development bill is being asked to cover many unrelated tasks.

The trick is not making more, it is doing more kinds

There is a familiar way costs fall: make the same thing over and over, and each copy gets cheaper. That is not what is happening here. Starship is not cheap because SpaceX makes a lot of Starships. It is cheap per job because one Starship is asked to do many different jobs.

Economists call this economies of scope. Building the machine costs a fortune once. After that, every new kind of task it can take on rides on money that is already spent. The engines, the factory, the trained crews, the launch towers were paid for to loft internet satellites. Pointing that same capability at a Moon landing does not mean paying for all of it again.

Why NASA is buying instead of building

You can see the saving from NASA’s side. NASA needs a lander to put people on the Moon. Developing a brand-new deep-space lander from nothing would cost a staggering sum and take many years.

So NASA is not building one. It is buying a lander that is really a variant of a rocket SpaceX built for its own reasons. The hardest, most expensive part, the vehicle itself, already exists. NASA pays for the adaptation, not the invention. A job that would have been ruinous alone becomes affordable because someone else already paid the big bill for a different purpose.

You are already riding this bill

This is not a story about one company. Follow the thread outward and a lot of separate-looking things sit on the same foundation. Think of the astronaut who lands on the Moon, the person streaming Starlink internet in a remote valley, the operator whose satellite hitches a ride, the taxpayer funding Artemis. All of them draw on one shared pile of sunk cost. Even the weather forecast and the map on your phone lean on cheaper access to orbit that this kind of sharing helps deliver. Watch it stack up. On the Artemis flight, astronauts plan to send 4K video home from Orion over SpaceX’s Starlink laser links. The internet network built to sell you broadband will double as the relay carrying a Moon crew’s pictures to Earth. One asset, quietly doing a second job again. The reader is not watching from outside. You are one of the many jobs the bill is spread across.

The saving has a shape, and the shape has an owner

Notice what this does to the field. Once a company owns the one machine that does everything, a rival who must build a separate vehicle for each task cannot match the price. The all-purpose owner is spreading its costs; the single-purpose builder is not. So the work quietly gathers around whoever got there first.

That is the arrangement under the good news. Cheaper access to space is real, and it genuinely helps everyone downstream. It also means NASA, the military, and commercial customers increasingly reach for the same few providers, because those providers alone can offer the shared-cost price. The efficiency and the concentration are the same fact seen from two sides. Neither is a villain; both are true.

What seeing the whole leaves you with

It is tempting to file these as separate stories: a rocket test, a Moon program, a satellite internet company, a Mars dream. They are not separate. They are one bet, seen from four angles, held up by one bill that no single one of them could carry alone.

That is worth holding loosely. When many different things come to rest on one shared foundation, they rise together and they lean together. No one seat, not NASA’s, not a subscriber’s, not even the company’s, can see the full weight resting on it.

03 · Lab · your turn

Spread the bill

Rehearse economies of scope: one shared asset only beats dedicated tools when enough different jobs ride on it.

04 · Hope · carry this

There is quiet encouragement in a machine asked to do many things at once: the same effort that carries a stranger's internet may also carry a crew back to the Moon. When our hardest goals come to rest on shared ground, a step forward in one of them rarely stays in one place for long.