The once pristine white floors featured in Relativity Space’s PR photos are now scuffed and coated with the residue of a typical machine shop. Inside its warehouse on the outskirts of Los Angeles, three robot arms hang imposingly next to a container filled with a coil of metal wire. The container’s lid has a jagged hole as if someone punched through it on a bad day; duct tape has been slapped on to cover the sharp edges. This is a machine that’s been pushed to its limits, in service of a lofty goal. Led by its founders, Tim Ellis and Jordan Noone, Relativity is attempting to create 95% of its rocket, Terran 1, using 3D printing, in just 60 days.
You read that right: the plan is to go from raw material to a launch-ready rocket in two months. If it sounds audacious, that’s because it is. Hugely. 3D printing is having a moment in the spaceflight industry—everyone from SpaceX to Blue Origin to lesser-known startups and old-guard rocket shops are tinkering with the technology, and some have gone so far as to print their own engines from scratch. But even engineers on the cutting edge of 3D-printed rocketry don’t know what to make of Ellis and Noone’s upstart firm. And more than one think they’re just crazy.
Traditionally the aerospace industry hasn’t been quick to change, and for good reason: rockets are controlled explosions that put huge sums of money and, sometimes, human lives on the line. Relativity is aiming to win over skeptics and holdouts with a test launch in 2020. Thing is, they haven’t even printed a whole rocket yet.
At their core, rockets consist of four main systems: payloads, guidance, propulsion, and structures. The payload is whatever the rocket is carrying. The guidance consists of sensors that keep the craft on target, and propulsion is made up of the fuel and engine that make it go. The structures are the rest of the frame, cone, and fins of the rocket—parts that are typically fabricated using ultra-precise CNC milling machines and hand welding.
That’s all a way of saying that behind every successful launch is a tremendous amount of labor and a vast network of suppliers working in concert to assemble each vehicle. By streamlining the supply chain, Relativity hopes to sharply cut production time.
But this goal of printing Terran 1’s more than 100-foot-tall (30-meter) exterior and fuel tank comes with an additional challenge: creating printers that can accomplish the task. “Building a rocket company is hard, building a 3D-printing company is hard, and building both together at the same time is borderline nuts,” says Ellis, Relativity’s CEO. “But while it’s the hardest part of the job, it is also the secret sauce that will make Relativity a world-changing company.”
There’s still a way to go before doing any world changing, though. “We’re not going to fly a rocket unless we get these metal 3D-printing technologies developed,” Ellis admits. “So that provides quite a bit of existential kick in the butt to figure it out, because this is the only way we’re going to actually make it to our goal.”
Children of the Stargate
Relativity’s solitary 20-foot-tall printer, Stargate, has been serving the company since it exited stealth mode in 2017, but it’s finally about to get a break. In a nearby building are four updated, fresh-out-of-the-box models. Each one is shielded by long black flaps that run from the warehouse ceiling to the floor and betray their newness with a pungent plastic smell. One has a small toy basketball hoop hanging on it—as if, so far, it’s more often served as a backboard than a rocket printer.
A giant image splashed across the wall depicts a hoped-for vision of the company’s future: a warehouse filled with nothing but Stargates, smaller printers, and robot arms. An engineer’s paradise, and a machinist’s nightmare. It’s the “robots are taking our jobs” headlines in mural form.
The hulking machines seem to smirk at decades of rocket assembly. During the Apollo program, engineers faced extreme difficulty achieving perfect welds on the Saturn series of rockets. Even experienced welders had to be given specialized training to complete the long, precise welding passes required. Now a robot is welding the entire thing.
Stargate and its offspring use a variant of what’s known as directed energy deposition. Traditional manufacturing methods involve carving a finished product from a block of material. 3D printing builds up an object layer by layer instead, enabling the creation of lightweight objects with intricate internal structures that are impossible to make any other way. The most prevalent form of 3D printing is called fused deposition modeling—a material, often plastic, is melted and squeezed out of a nozzle in precise patterns to build an object. Combine that with welding and you have directed energy deposition.
The basics of welding involve supplying a steady stream of metal wire with one hand and heat with the other. Stargate does this automatically, feeding wire out of an extruder on the end of a tall robotic arm. The metal is heated using electric plasma (and sometimes a laser) and then laid down according to a computer’s instructions. A combination of electronic controls, thermal imaging cameras, and sensors mounted near where the material is deposited adapt the print as it’s created. “Our vision of 3D printing is software-defined automation for aerospace,” says Ellis. “That’s getting toward the long-term vision of 3D-printing rockets on Mars. These are exactly the tools we’re going to need to actually build stuff on other planets.”
The way Ellis talks about his company brings to mind Elon Musk’s exultations about SpaceX and Tesla, only Ellis says he is completing a piece of the Mars puzzle Musk isn’t yet tackling. “The thought is having two products. One is the rocket launch vehicle. The other is the factory,” he says. “Over time, the factory we see being able to shrink down smaller and smaller and smaller until it’s eventually something that we can actually just launch on a big rocket.” You build the machine that makes the machine. And then launch it to Mars. Simple.
Relativity took a small step toward that vision last month, announcing it had signed a nine-year lease with NASA on a 220,000 square foot facility in Mississippi that will become its first autonomous rocket factory.
Even fellow rocket companies aggressively pursuing 3D printing (a.k.a. additive manufacturing) aren’t entirely convinced this is how the future looks. Rocket Lab, one of only a few small satellite launchers that fly commercial flights, has relied on additive manufacturing to create engines, valves, manifolds, and a number of other complex components; its CEO, Peter Beck, says, “There’s no way that we can produce the volume and the performance of the engines that we’re producing now without 3D-printing technology.” But an entire rocket? “To go and print an avionics box or tank or something like that doesn’t make any sense, because there’s much more efficient processes for doing that,” says Beck. “I don’t want to rain on Tim’s parade. I wish him the absolute best, but from an engineering perspective, it makes absolutely no sense to us.”
In the end, customers are the ones who will need proof of the wisdom of Relativity’s method. Like most rocket companies before their first launch, Relativity is selling its customers on test data and the team that’s been assembled. “Ultimately, it’s a belief and a leap of faith that we’re going to go execute,” says Ellis. “But yeah, it’s a pretty big one. And definitely, it’s a process to get to it.”
Evidently some customers are willing to take that leap. Relativity has already publicly announced three clients with launches booked for 2021 and 2022: the Canadian communications company Telesat, Washington-based Spaceflight (which helps coordinate satellite ride shares on larger launches), and Thailand’s mu Space. Noone says that once Relativity shows it can launch successfully in 2020, it plans to increase the number of flights it launches each year to 12 to 24.
These kinds of aggressive time lines are baked into the company’s lore. Three years ago, shortly after Ellis and Noone each left their first jobs out of college at Blue Origin and SpaceX, respectively, they pitched investor Mark Cuban via email to ask for seed funding. The message had the subject line “Space is sexy: 3D printing an entire rocket.” Cuban, who conducts the majority of his business through email, replied five minutes later saying he wanted to invest $500,000. Two months later he did. According to Cuban, it wasn’t just the additive-manufacturing element that caught his eye. “The idea was unique. I wish I had thought of it,” he says. “They were qualified, and they were local.” (Ellis is from Texas, where Cuban lives.)
Since the infusion, Relativity has put its foot on the gas. In the past year it’s grown from 14 people to more than 80. The team now includes Tim Buzza, one of the first SpaceX employees and former VP of launch for both SpaceX and Virgin Orbit, and David Giger, a 12-year SpaceX employee who served as the senior director of engineering for the company’s Dragon capsule.
Ellis, the front man for hiring and raising capital, doesn’t seem to have trouble winning people over at all levels. He’s got a spot on the White House’s National Space Council Users Advisory Group, and contracts and cash are flowing into the company. Relativity has closed a $35 million series B funding round, scored a deal with NASA to test its engines at the Stennis Space Center in Mississippi (the same facility where its autonomous factory will eventually go), and received permission to launch at one of the most competitive launch sites in the world: Florida’s Cape Canaveral.
This last coup, announced in January, lines the Terran 1 up to launch from the hallowed Launch Complex 16, which once played host to Titan missile launches, the Apollo program, and the Gemini program. High-profile moves like that have forced Relativity’s name into conversations about companies like SpaceX, Blue Origin, and United Launch Alliance, previously the only three outfits with permits to lift off from Cape Canaveral.
Printing takes off
Relativity is far from alone in hoping that 3D printing will propel it into the elite of spaceflight. Startups including Virgin Orbit, Firefly, and Electron are all vying to prove that they, like Rocket Lab, have what it takes to launch small satellites to space. Even established companies like Aerojet Rocketdyne are trying to prove 3D printing is on par with—or even more reliable than—traditional manufacturing techniques.
But no one is going for it as hard and fast as Relativity. Aerojet builds engines for government contracts and human-rated rockets like NASA’s Space Launch System, which have to be extra consistent and reliable. The company says that more than 60% of its research and development for 3D printing has been nothing more than establishing a database of the chemical and structural properties of different materials. “Others may kind of skip over that, and that’s their right to do that as a risk-accepting posture,” says Jeff Haynes, Aerojet’s senior manager of advanced programs.
By contrast, at Relativity, “if we put a fully printed engine on the test stand, successfully fire it, and then fly it, that for us is success,” says Noone. “You could write hundreds of pages of specifications telling you how to get there, and how to manufacture it, but we have our ways that we do it. I wouldn’t want to be hung up on creating the specification rather than just trying something and demonstrating that it works.”
That “move fast and break things” mentality would lead to a lot of sleepless nights for most rocket designers. Virgin Orbit, a competitor of Relativity’s, has additive-manufactured parts on its first LauncherOne rocket, but the company is happy to go easy on trendy tech. “The LauncherOne vehicle engine right now uses very reliable manufacturing methods that NASA has proved out since the ’50s and ’60s, because [priority] number one for first-launch vehicles is reliability,” says Virgin Orbit’s advanced-manufacturing manager, Kevin Zagorski.
The other companies giving additive manufacturing a chance run the gamut from Jeff Bezos’s Blue Origin—where Ellis had a hand in purchasing the company’s first metal 3D printer during one of his three internships there—to Launcher, a small startup that claimed to have made the world’s largest 3D-printed rocket engine. Heavy hitters like SpaceX, NASA, Rocket Lab, United Launch Alliance, and ArianeGroup have entered the 3D-printing ring as well.
The reasons most of these organizations give for using the technique are twofold: you can build something with fewer parts and tweak designs more quickly. Initially, Rocket Lab’s Beck saw additive manufacturing getting a bad reputation because it wasn’t being used effectively. “Someone would take a subtractively manufactured [i.e., machined] component and attempt to 3D-print it. It would turn out more expensive and more time consuming,” he says. “But like any new technology, it’s all about designing for the process. Where 3D-printed parts really excel is where you have really high complexity and you merge a lot of parts into one.”
For its part, Relativity boasts that Terran 1 will have just a hundredth as many parts as a standard rocket. Its engine, Aeon 1, is made from only three parts pieced together.
How much of this is a PR stunt, though, is hard to sort out. Announcing you’ve made the first whatever is tempting, especially for small startups. Relativity, for example, claims to have built the largest metal 3D printer—as do Sciaky and Titomic, two industrial hardware companies that aren’t in the space business. “Everybody’s looking to try and have a point of differentiation and trying to grab some headlines,” says Beck. “If someone wants to talk about 3D-printing something, then fine, but it’s somewhat amusing.”
Even if 3D-printing an entire rocket isn’t practical, “I’m really confident that in any case it will result in useful spin-offs,” says Dan Erwin, head of astronautical engineering at the University of Southern California. Erwin ran USC’s rocket lab when Ellis and Noone studied there but hasn’t worked with them since. “I have the intuition that this is one of those ‘If you build it, they will come’ kind of things,” he says. Regardless of whether Relativity launches a rocket by next year, it is forcing a slow-moving industry to take a closer look at, and perhaps advance, a technology that has uses outside spaceflight. The end result might be nothing more than a new breed of printer. Or it might be the Mars-bound rocket we’ve all been promised. “Life is too short to just wait for the future to happen faster,” says Ellis. “We should create it.”
Meta has built a massive new language AI—and it’s giving it away for free
Facebook’s parent company is inviting researchers to pore over and pick apart the flaws in its version of GPT-3
The gene-edited pig heart given to a dying patient was infected with a pig virus
The first transplant of a genetically-modified pig heart into a human may have ended prematurely because of a well-known—and avoidable—risk.
Saudi Arabia plans to spend $1 billion a year discovering treatments to slow aging
The oil kingdom fears that its population is aging at an accelerated rate and hopes to test drugs to reverse the problem. First up might be the diabetes drug metformin.
Yann LeCun has a bold new vision for the future of AI
One of the godfathers of deep learning pulls together old ideas to sketch out a fresh path for AI, but raises as many questions as he answers.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.