Photorealistic artist rendering of a Mars colony. Several Starship rockets stand upright near a complex of buildings. The sun, smaller than it appears from Earth, is setting over the landscape, giving the scene a dusky purple color.

Prime mover

Starship will be the biggest rocket ever. Are space scientists ready to take advantage of it?

Go to content
Rocket company SpaceX envisions using fleets of Starships to establish human outposts on Mars, depicted in this artist’s concept. SpaceX
issue cover image
Table of contents
A version of this story appeared in Science, Vol 377, Issue 6607.Download PDF

NASA’s Lunar Crater Observation and Sensing Satellite mission was brutish and short. It began on 9 October 2009, when the hull of a spent Centaur rocket stage smashed into Cabeus crater, near the south pole of the Moon, with the force of about 2 tons of TNT. And it ended minutes later, when a trailing spacecraft flew through and analyzed the lofted plume of debris before it, too, crashed. About 6% of the plume was water, presumably from ice trapped in the shadowed depths of the crater, where the temperature never rises above –173°C. The Moon, it turned out, wasn’t as bone dry as the Apollo astronauts believed. “That was our first ground truth that there is water ice,” says Jennifer Heldmann, a planetary scientist at NASA’s Ames Research Center who worked on the mission.

Today, Heldmann wants to send another rocket to probe lunar ice—but not on a one-way trip. She has her eye on Starship, a behemoth under development by private rocket company SpaceX that would be the largest flying object the world has ever seen. With Starship, Heldmann could send 100 tons to the Moon, more than twice the lunar payload of the Saturn V, the workhorse of the Apollo missions. She dreams of delivering robotic excavators and drills and retrieving ice in freezers onboard Starship, which could return to Earth with tens of tons of cargo. By analyzing characteristics such as the ice’s isotopic composition and its depth, she could learn about its origin: how much of it came from a bombardment of comets and asteroids billions of years ago versus slow, steady implantation by the solar wind. She could also find out where the ice is abundant and pure enough to support human outposts. “It’s high-priority science, and it’s also critical for exploration,” Heldmann says.

The stainless steel Starship rocket, standing upright, photographed from near the top so that its height is accentuated by the perspective.
SpaceX expects to launch the 120-meter-tall Starship on its first orbital test flight in the coming months.SpaceX

When SpaceX CEO Elon Musk talks up Starship, it’s mostly about human exploration: Set up bases on Mars and make humans a multiplanetary species! Save civilization from extinction! But Heldmann and many others believe the heavy lifter could also radically change the way space scientists work. They could fly bigger and heavier instruments more often—and much more cheaply, if SpaceX’s projections of cargo launch costs as low as $10 per kilogram are to be believed. On Mars, they could deploy rovers not as one-offs, but in herds. Space telescopes could grow, and fleets of satellites in low-Earth orbit could become commonplace. Astronomy, planetary science, and Earth observation could all boldly go, better than they ever have before.

Of course, Starship isn’t real yet. All eyes will be on a first orbital launch test, expected sometime in the coming months. Even if it is a success, no one knows whether SpaceX will be able to achieve its vision of launching the rockets daily and reusing them many times. Also unsettled is whether a market will materialize for a rocket that could put so much into orbit. But scientists need to prepare, Heldmann says. “We on the science side need to be ready to take advantage of those capabilities when they come online.”

So do NASA centers such as the Jet Propulsion Laboratory, which designs and builds many space science missions, says Casey Handmer, a former JPL software engineer. In a series of provocative blog posts with titles like “Starship is still not understood,” he has argued that Starship will upset the traditional way of doing space science—spending billions of dollars to make one-of-a-kind instruments that work perfectly. If the NASA centers don’t find ways to take risks and make more stuff more cheaply, he says, they will find themselves displaced by companies willing to do so. “The writing is on the wall,” Handmer says. “And all the NASA centers should be thinking really carefully.”

On a balmy night in February, Musk strode onto a stage in Boca Chica, Texas, home of SpaceX’s Starbase launch site, for a public update on the status of Starship. Towering behind him, bathed in lights, was the latest prototype, about 120 meters tall: the Starship vehicle, which carries people or payloads, resting on top of a Super Heavy booster. The prototype wasn’t flight ready, nor had the Federal Aviation Administration (FAA) given SpaceX permission to launch it from Starbase—but it was still a spectacular backdrop, packed with coiled purpose. After welcoming the crowd of faithful rocket geeks, Musk launched into an impromptu lecture on the philosophy propelling him and his company beyond Earth. “Why build a giant, reusable rocket? Why make life multiplanetary? I think this is an incredibly important thing for the future of life itself.”

SpaceX’s workhorse rocket, the 70-metertall Falcon 9, has already shaken up the aerospace business. With that rocket, SpaceX pioneered reusability, employing retrorockets and steerable fins to guide the first stage to a landing after it reenters the atmosphere. Today, SpaceX routinely slaps on a fresh coat of paint and launches it again; in June, the company flew one of these “flight tested” stages a record 13th time. Another record is on the horizon: The company is on track to launch more than 50 Falcon 9 and Falcon Heavy rockets this year, or about one per week on average. The dependable reuse and rapid launch cadence are two of the reasons why SpaceX can charge $67 million for a Falcon 9 launch, much less than its competitors. But Musk wasn’t satisfied.

In 2016, at an International Astronautical Congress in Mexico, Musk sketched out plans for a rocket to colonize Mars, one he would soon be calling BFR (Big Falcon Rocket, in family-friendly terms, but you get the joke). The concept evolved into Starship, but the focus remained on affordability and reusability—making launches as dull and routine as FedEx cargo flights. The body of the rocket is stainless steel, heavier than the aluminum alloys of most rockets, but cheaper and more easily manufactured. The 33 Raptor engines crammed into the backend of Super Heavy burn methane rather than the traditional kerosene-based rocket fuels, not only because it is cheaper, but also because it could be harvested on Mars by combining carbon dioxide and water. The booster is designed to return to the launchpad after its 6-minute ride; the company believes it can be refueled and ready to relaunch in an hour. Starship is also reusable. The goal is to be able to launch each vehicle three times a day.

Once in orbit, a loaded Starship could be gassed up by a “tanker” version of the vehicle—enabling it to take its 100 tons of payload on to the Moon or Mars. At the February event, Musk explained how a single Starship, launching three times per week, would loft more than 15,000 tons to orbit in a year—about as much as all the cargo that has been lifted in the entire history of spaceflight. Musk has claimed the price of each launch might eventually be as low as $1 million, or $10 per kilogram to low-Earth orbit. The only rocket close to Starship in its capabilities is NASA’s Space Launch System, set to fly for the first time this month. Earlier this year, the agency’s auditor found each launch would cost about $4 billion, or nearly $60,000 per kilogram.

Heavy lifters

Starship would be the largest rocket the world has ever seen. It could also offer the cheapest launches. To achieve that, SpaceX will need to launch fleets of Starships daily—and it will need customers for the immense cargo capacity.
Line illustrations for five rockets, accompanied with their debut year, height, and cost per kilogram to low-Earth orbit in 2021 dollars. Starship is the cheapest by far, with a cost per kg of $10 instead of thousands of dollars. Detailed data: Space Shuttle: Debut, 1981; Height,  56 m; cost per kg, $65,400. Falcon 9: Debut, 2010; Height, 70 m; Cost per kg, $2600. Space Launch System (Block 1): Debut, 2022; Height, 98 m; Cost per kg, $58,000. Saturn V: Debut, 1967; Height, 111 m; Cost per kg, $5400. Starship: Debut, 2022; Height, 120 m; Cost per kg $10.

Long hauler

By refueling in orbit, each Starship can carry its full 100-ton payload to Mars. On Mars, water, carbon dioxide, and electricity can be used to make methane fuel.
1 Booster lifts ship and returns to launch pad.
2 Ship reaches Earth orbit.
3 Tankers refuel ship and return to Earth.
4 Refueled ship travels to Mars.
5 Ship refueled using martian resources.
6 Ship returns directly to Earth.
V. Altounian/Science

Pierre Lionnet, a space economist at Eurospace, an industry trade group, is skeptical SpaceX can achieve such a low price point. It may not correctly account for the costs of developing and building the rocket, for example. “When I look at Starship, I’m looking at what seems to be a very expensive device.” To achieve profitability with such high capital costs, SpaceX will have to attain its ambitious launch rates, which means it will need paying customers to soak up all that cargo capacity. SpaceX hopes to develop new markets in space mining, tourism, or other activities not yet dreamed of, but Lionnet is not so sure the heavy lifter will whet that appetite all by itself. “If you’re vegetarian, and I’m offering you a burger, I can offer it at the cheapest possible price, and you don’t eat it.”

The debate will soon graduate beyond the theoretical. In May 2021, after several spectacularly explosive failures, a Starship upper stage flew 10 kilometers up into the atmosphere. After landing, it briefly caught fire, but the company deemed the suborbital flight a success. Since then, SpaceX has built out Starbase, constructing a launch tower that can catch returning boosters with two robotic arms the company calls “chopsticks.” It has refined its rocket assembly line, which can now build four Raptor engines per week. And in June, FAA gave SpaceX approval to launch from Starbase, provided it takes steps to minimize the impact on the environment.

At the February event, Musk said he was confident Starship would make its first orbital attempt this year. For Musk, the sci-fi dreams are tantalizingly within reach. “Let’s make this real!” he exhorted the crowd, pumping his fists.

Science has mostly been an afterthought for Musk. But Heldmann has been surprised that, for many planetary scientists, Starship has also been an afterthought.

In 2020, she and a team of researchers and industry insiders submitted a white paper touting the benefits of Starship to the “decadal survey” in planetary science, an influential community exercise that helps NASA and Congress set long-term priorities. “It’s a good time to try and get this idea in the consciousness of others,” she says. Heldmann and her colleagues suggested NASA create a dedicated funding line for missions relying on Starship.

Three small groups of workers stand on platforms extending out from the booster base as the Starship rocket is lowered onto it.
In 2021, workers stacked a Starship upper stage on a Super Heavy booster.SpaceX

The survey embraced the ideas. In its April report, the survey committee explicitly mentioned Starship and cited ideas in Heldmann’s paper. The committee recommended a funding line relevant to Starship’s specs and said NASA should plan to capitalize on the rocket’s potential. “Both cargo and crew flights to Mars offer significant potential science opportunities,” the committee said.

The benefits wouldn’t be limited to the Moon and Mars, points out Daniel Baker, director of the Laboratory for Atmospheric and Space Physics at the University of Colorado, Boulder. For instance, Mercury, with its weak gravity, has been a tough target because of the extra fuel required to slow a spacecraft enough to get into orbit; the Sun’s heat is another problem. But Starship wouldn’t mind the big gas tank or the sunshade needed to keep the spacecraft from melting. Baker also envisions faster missions to the outer planets that don’t require time-consuming gravitational assists from other planets. Even farther afield, Interstellar Probe, a proposed mission to follow in the footsteps of NASA’s famed Voyager mission, could carry more capable instruments aboard Starship—and get a faster ride to interstellar space.

Some astronomers also have Starship in their eyes. “There’s no way to talk about it without resorting to cliches, but ‘best rocket engine ever,’ probably, by most metrics,” says David Rubin, a cosmologist at the University of Hawaii, Manoa. He wonders how much simpler the $10 billion James Webb Space Telescope (JWST) might have been if its 6.5-meter-wide segmented mirror hadn’t had to fold up to fit on its rocket. Engineers could have built a monolithic mirror and launched it as is within the 9-meter-wide Starship fairing, which encloses a volume about half as big as a hot air balloon.

Rubin also dreams of using Starship to construct a giant telescope—say 30 meters—in space. Limbed robots could precisely lay down mirror segments on a scaffolding, forming a giant mirror that could pick out the universe’s first galaxies and look for signs of life in the atmospheres of Earth-like exoplanets. “The science gains scale really quickly as you build larger and larger telescopes,” Rubin says.

quotation mark
If thousands and thousands of tons are going to orbit, someone’s going to figure out how to put a telescope up there without NASA.
  • David Rubin
  • University of Hawaii, Manoa

Not all astronomers are fans of SpaceX, which has already launched nearly 3000 of its Starlink internet satellites into low orbits, where sunlight glinting off them leaves streaks on the cameras of ground-based telescopes. The problems could multiply with Starship, which could launch hundreds of Starlinks at a time, enabling the company to build its planned constellation of up to 42,000 satellites even faster. SpaceX now equips the satellites with “sunshades” to reduce the reflective glare, but astronomers are still worried. “Making access to space and Earth orbit easier has a lot of benefits,” says Meredith Rawls, an astronomer at the University of Washington, Seattle, and a member of an International Astronomical Union center set up to mitigate satellite interference. “But we need to make sure that we’re doing it in a mindful way and not just having it be a Wild West disaster.”

As another reality check for the dreamers, Lionnet points out that discounted rides will only reduce the cost of missions by so much. For major scientific projects, Lionnet says, launch costs are usually between just 5% and 10% of the total price tag. For JWST, the fraction was even smaller. The typical cost for a ride on an Ariane 5 rocket, JWST’s launcher, is about $175 million, just 2% of the mission’s total price tag. “A complex telescope will still be a complex telescope,” Lionnet says.

But cheaper launches could allow the probes themselves to be cheaper, with less need for space-rated parts that save on weight or bulk. With Starship, planetary scientists wanting to outfit a rover with a spectrometer could just buy one on the internet. Astronomers could use a glass mirror instead of a featherweight beryllium one, like JWST’s. And, Rubin says, “You should just be able to shield your way into radiation hardness,” rather than soldering circuits out of specialized materials.

A close-up view of the side of the rocket, partially covered with hexagonal black tiles.
Thousands of hexagonal heat tiles will keep Starship from burning up when it reenters Earth’s atmosphere.Reginald Mathalone/NurPhoto/AP

Earth-observation researchers have already been down that road. For many years, remote sensing satellites were big and pricey—little different from JWST, says Aravind Ravichandran, founder of the spaceindustry consultancy TerraWatch Space. But about a decade ago came the “small satellite” revolution: Researchers shrank and standardized equipment and took advantage of rideshares on relatively cheap Falcon 9 launches and other small rockets.

Suddenly, university students were sending shoebox-size CubeSats to space. Using cheap cameras and consumer electronics, the company Planet built up a fleet of about 200 small satellites that gather daily images of all of Earth’s land. Ravichandran sees Starship making it easier to assemble ever bigger fleets—enough eyes on the planet to revisit a given spot multiple times an hour, rather than every few hours, days, or weeks. “Why can’t you do every 5 minutes? Every 10 minutes?” Ravichandran asks. Imagine, he says, what that kind of revisit rate might do for, say, tracking wildfires or floods.

Handmer, who now works as a clean-energy entrepreneur, wants astronomers and planetary scientists to adopt this sort of bold thinking. Instead of a 30-meter telescope, why not a 1000-meter one? Why not mass-produce probes that could survey dozens of asteroids? Why not fly by all the outer planets in the next decade? Or land on most planets annually?

In Handmer’s view, the problem is partly cultural: NASA engineers try to get everything right on the first try, at all costs—the vastly expensive, long-delayed JWST being a prime example. “It’s kind of like a medieval cathedral,” he says, of such flagship missions. To exploit Starship’s immense capacity, Handmer estimates NASA will need to make 100 times as much stuff for a fraction of the usual cost. It will need to be a fast-fashion factory, not a boutique. But having worked at JPL, Handmer isn’t necessarily hopeful that will happen. “It was just not set up to mass-produce anything,” he says.

Rubin says NASA centers could get left behind by nimbler companies, or privately funded scientists. “If thousands and thousands of tons are going to orbit, someone’s going to figure out how to put a telescope up there without NASA,” he says. Astrobotic, which calls itself a “lunar logistics company,” has sprung up with this sort of business model. It plans to send landers and rovers to the Moon, carrying instruments for paying customers.

Robert Manning, JPL’s chief engineer, doesn’t think the facility is quite so resistant to change. But he also questions Handmer’s vision of cheap, mass-produced probes. The equipment used at the scientific frontier is rarely standard. Every mission, with its fresh targets and questions, requires innovation. And it’s hard to take risks as a public agency, Manning says. “We have an obligation to make sure that we are not wasteful of taxpayers’ dollars,” he says. “We can’t throw things to space frivolously and say, ‘Well, if it doesn’t work, let’s build another one.’”

In 1992, then–NASA Administrator Daniel Goldin pushed the agency to pursue a “faster, better, cheaper” approach—a mantra that was discarded later in the decade after high-profile losses of the Mars Climate Orbiter and Mars Polar Lander. If a faster, better, cheaper culture is to return, Manning says both NASA and Congress will have to bless the risk-taking—and stand up for it when things go wrong. “It’s going to be difficult to politically communicate that that’s OK for us to try it out,” Manning says.

But assuming all those issues worked themselves out, he acknowledges that frequent, low-cost, high-mass-and-volume launches could provide “an incredible opportunity for us to change how we get things done and be willing to take more risks.” JPL has already been thinking about it, he says—for example, considering how to incorporate standardized, lower cost components in NASA’s custom deep-space missions.

There’s a major asterisk on the rocket revolution that Starship heralds. “We haven’t been able to act on it yet,” Manning says, “because it’s not true yet.” When Starship makes its first orbital launch attempt, many researchers will be watching, waiting to see whether that giant silvery rocket is a vision of space science’s future, or a mirage.

Support nonprofit science journalism

Help News from Science publish trustworthy, high-impact stories about research and the people who shape it. Please make a tax-deductible gift today.


Not Now