Aerospace startup making 3D-printed rockets now has a launch site at America’s busiest spaceport

A rendering of what Relativity Space’s rocket will look like, launching from LC-16.
Image: Relativity Space

America’s busiest spaceport in Cape Canaveral, Florida, is about to get a new tenant: a startup that shares SpaceX’s ambitious plans of turning humans into a multiplanetary species. The new occupant is LA-based launch provider Relativity Space, a company that wants to revolutionize how rockets are manufactured through the use of fully automated 3D printing. The company will soon have its very own launch site at the Cape for its future 3D-printed vehicles.

Thanks to a new deal with the US Air Force, Relativity will be taking over a site at Cape Canaveral Air Force Station called LC-16. It’s a pad that was once used by the US military to launch Titan and Pershing ballistic missiles. But since the late 1980s, LC-16 has been dormant. The Air Force picked Relativity to move into the area after a very competitive bidding process, and the company will modify the pad to suit its rocket technology. “Getting the launch site agreement was a huge checkmark,” Tim Ellis, co-founder and CEO of Relativity Space, tells The Verge. “That was the final infrastructure piece we need to have a clear path toward launching.”

Over the last year, Relativity has quickly established itself as a serious player in the commercial space industry. The company, which was founded in 2016, has raised more than $45 million in funding. It also has multiple workspaces in Los Angeles, and it’s currently using facilities at NASA’s Stennis Space Center in Mississippi to test the Aeon engine it’s been working on. As of now, Relativity has done 124 test fires of its rocket engine, in pursuit of launching the company’s first rocket by 2020.

A rendering of what Relativity Space’s pad at LC-16 should look like.
Image: Relativity Space

That vehicle will be called the Terran 1, a nod to the human explorers in the computer game Starcraft. It won’t be terribly large, either: at about 10 stories tall, the rocket will be a small- to medium-sized launcher that’s capable of putting about 2,755 pounds (1,250 kilograms) into low Earth orbit, which is about the same weight as a Honda Civic. That’s a small load compared to the 50,000 pounds (22,800 kilograms) that SpaceX’s Falcon 9 can carry to orbit. But Ellis says the real groundbreaking aspect of the vehicle will be in how it’s made.

Relativity’s goal is to disrupt the entire process of manufacturing rockets. “For the last 60 years, the way rockets have been built hasn’t really changed,” says Ellis. Instead of relying on the traditional, complicated assembly line of machines and people sculpting and piecing together parts of a vehicle, Relativity wants to make building a rocket almost entirely automated. The trick? Using giant 3D printers that can create all of the parts needed to build a rocket — from the engines to the propellant tanks and structure.

At the company’s Los Angeles headquarters, Relativity has the largest metal 3D printer by volume, a machine that’s capable of creating parts that are up to 20 feet tall and 10 feet wide. It’s called Stargate, another nod to Starcraft, and the team designed this printer from scratch, which means they can scale it up if needed. Ellis says that by relying on printers like this for manufacturing, the team will be able to produce about 95 percent of the rocket through 3D-printed automation. The last 5 percent still requires human labor. Most of that human interaction will be centered on testing, shipping, and very small amounts of manual assembly.

Relativity’s Stargate 3D printer at the company’s LA headquarters.
Photo: Relativity Space

Building rockets this way is meant to serve two purposes for the company. First, it’s meant to save money by consolidating the parts needed for each vehicle. Ellis says that the 3D printer they’ve developed can make incredibly complicated parts in just one piece, and he argues Relativity will be able to produce rockets with 100 times fewer parts than normal. For instance, Relativity’s engine injector and chamber are made of just three 3D-printed parts; traditionally, such sections would require nearly 3,000 parts, says Ellis. “All the complexity is really in the software,” he says. “It’s really what the file and CAD model looks like. The 3D printer doesn’t really care how complex it is. It’s able to make shapes of almost any complexity.”

The team can also quickly adjust the design if needed, simply by changing the software. And 3D printing will allow the company to simplify the manufacturing process, shortening the time it takes to build each rocket. Ellis says the goal is to get to a point where it only takes 60 days to manufacture one vehicle. “We’ll be able to achieve that because of the robotic automation and 3D-printing technologies,” he says.

If all of this works out, Relativity will try to realize its more ambitious long-term goal: taking this manufacturing process to the Red Planet and building a rocket using 3D printers on Mars. Ultimately, Relativity wants to establish a construction business on our planetary neighbor, along with launching rockets here on Earth. “I really view us as having two products,” says Ellis. “One is the rockets, the other is the factory.”

Ellis, a former propulsion engineer at Blue Origin, says he was inspired by SpaceX’s core mission to start a settlement on Mars. But he thought there needed to be at least dozens or hundreds of companies working on Mars technologies to make that vision for the future a reality. To illustrate his point, he refers to one of SpaceX’s animations, which shows the company’s future Mars rocket, the Starship, landing on the Red Planet’s surface. “The door opens and the astronauts get out, and there’s just nothing there,” he says. “There really needs to be a company that’s actually going to work on the build technology.”

Relativity’s Aeon engine undergoing a test fire at Stennis Space Center.
Photo: Relativity Space

That’s where Relativity hopes to come in. Once it masters its automation process here on Earth, the company hopes to shrink its printers and ship them to Mars via rockets to see if they can create vehicles capable of launching from the Red Planet using raw metallic materials. If successful, Relativity could provide a service that both scientists and engineers have dreamed about for decades: a way to leave Mars once you get there. So far, we’ve only ever been able to land hardware on Mars, but not bring it back. Being able to launch from Mars would be useful for getting humans off the planet or even collecting samples of Martian rocks in order to return them to Earth for study. Ellis says that the company has already piqued the curiosity of NASA, which hopes to bring samples of Mars to Earth someday.

For now, Relativity is focused on proving that it can build and launch rockets from Earth in this newly automated way, and it has a fairly credible team working toward that goal. The company has grown to a team of 60 people, 12 of which are already well-respected leaders in the private spaceflight industry. That includes Tim Buzza, one of SpaceX’s earliest engineers. The company has also raised all of its funding purely through venture capital. So with this new deal announced today, Relativity is the first venture-backed launch company to set up shop at Cape Canaveral Air Force Station, which also hosts aerospace giants like SpaceX, Blue Origin, and the United Launch Alliance.

Customers seem to be excited about what Relativity has to offer. Ellis says the company has garnered a little more than $1 billion in potential contracts from customers, and with the new launch site, it’s building out a manifest of future launches. The team is also collaborating with NASA on new propulsion capabilities.

In the end, though, Ellis hopes the success of Relativity serves as an inspiration for other companies to work on technologies needed for Mars. Perhaps other organizations might want to work on new remote energy generation or mining technologies that could be used on both our planet and the one next door. “I hope we inspire 12 or 100 companies to want to go to Mars and do the same mission,” says Ellis. “And then we all work on different parts of this. That’s really the vision to me.”


Really excited for this new Space Race!
We need a good new name for this time!

One tiny question: Where is the range capacity coming from? If SpaceX is serious about Starlink (and their FCC licenses say they are), then they’re going to need pretty much every slot that the Eastern Range can cough up to have a prayer of making the FCC deadlines in 2024 and 2027. They can’t be happy having another company competing for the scarcest resource they have.

I can see that SpaceX might also launch from a boat in the future, they already land on one sometimes..

Maybe that’s why they aren’t building a SuperHeavy pad at Canaveral. But that’s a lot of new tech in a very short timeframe. They have to have 2205 birds up and running by 4/2024, another 3759 by 11/2024, another 2205 by 4/2027, and the final 3759 by 11/2024.

They are going to want Starship launching from Canaveral by 2023. Although an offshore launch facility is looking more and more likely.

That seems to be the obvious solution, but there’s been zero indication that SpaceX plans to spin up an SH/SS pad there any time soon.

Offshore would solve a lot of problems, especially if it were a mobile platform.

The range can do back to back launches from 40 and 39A in one slot. 25 slots per year, 2 launches per slot, 24 birds per launch, and 5 years gets it done.

Starship could probably do 100+ per launch starting around 2023.

Back-to-backs would solve a lot of problems, but I’d think that you’d have to put something like that out for public comment, and get the USAF’s buy-in at least a couple of years in advance. AFTS is a necessary but not sufficient condition.

I doubt they can turn early SH/SS launches around faster than a month, so a 100+ cadence would require a huge fleet—which they won’t have. Beyond that, BC isn’t magical when it comes to closing air and sea space. Unlike the Eastern Range, the BC range will have large volumes of air traffic to and from South America, and hundreds of oil platform servicing runs per day. Carving out a slot every third day ain’t gonna happen, no matter how reliable SH/SS’s launch cadence becomes.

I meant 100+ birds per launch, not 100 + launches.

I think the USAF has already indicated that they can handle 2 SpaceX launches within 24 hours back to back.

Sorry—read that wrong.

I don’t get any volume restrictions at all for SH/SS, which would mean that, if you assumed 500 kg wet mass and 75 kg of dispenser per bird, you’d get about 175 birds per launch. However, if you have to do the plane change, you sometimes need multiple refueling launches to go with it. When I average the Starlink launches with the fueling launches, I get between 30 and 86/launch for the mid-latitude launches.

Do you have a link to the USAF statement? I’m sure that they could turn the launches in an emergency, but 24 hours isn’t exactly an operational back-to-back launch. And the aviation folks are pretty annoyed already. If they could launch within a couple of hours back-to-back, that would be tolerable.

The 45th says they can in some cases support 2 launches in 24 hours if one of them uses AFTS:

Plane changes are unnecessary for Starlink, and so is injection by the LV to the final operational altitude. The birds will have to launch to a transfer orbit and can use nodal precession to change planes while they raise themselves to operational latitude. The injection will be no higher than ~500 km to avoid DOA birds clogging up operational orbits.

SpaceX also has launch facilities in California and plans to have them in Texas. Florida is not their only option.

California is only good for the near-polar launches, which is a fairly small fraction of the Starlink constellation. You can’t reach the main Starlink inclinations (42, 48, 53, and 53.8 degrees) from Texas without overflying either the southeastern US or the Yucatan peninsula before Starship has reached orbit.

SuperHeavy/Starship can launch to about 33.2 degrees, through the Yucatan Channel, but then it needs to execute a very expensive plane-change to get to the target inclinations. That will require refueling on-orbit, which effectively halves the payload capacity of Starship (since every other launch won’t contain any birds, just propellant).

And I’m not sure how much range capacity there will be at BC. The main limitation is clearing airspace and sea lanes, both of which are very busy off the Texas coast.

Overflying before reaching orbit is not an issue. The problem is the IIP trace and probability of a failure causing injury. SpaceX can launch on any azimuth they want if they can show a low enough probability of injury to the FAA.

I agree that probability of injury to the public is the metric—I was using the overflight as a simplified proxy for that.

I don’t see how they can hit the 1E-4 probability without a lot of empirical data about SH/SS reliability. They might have that data after a couple hundred launches, but until then, my guess is that the FAA is just going to use a 1500 km rule-of-thumb and leave it at that.

The Louisiana and Campeche coasts are both about 600 km downrange.

There is no rule of thumb, the calculation is extremely involved and includes many specifics about the vehicle, trajectory, and the population of the overflown area.

The Eastern Range won’t let an Atlas V go less than the 35 degree azimuth restriction, which prevents an impact in the Outer Banks, about 900 km downrange. An A5 is many things, but unreliable isn’t one of them.

Given that, I have a hard time believing that the FAA is going to let a completely new launcher, with a design that they can’t assess very well, overfly populated areas 600 km downrange, the first 100-200 times out of the chute. After that, we’ll see.

Atlas V sheds lots of parts as it flies downrange; Starship sheds none. Intentionally dropping up to 5 SRBs, a booster, and two fairings on every launch is going to have a much larger effect on PoC than the small chance of an upper stage designed for human rating blowing up.

View All Comments
Back to top ↑