How NASA Built Its Mars Rovers

Pete Theisinger stands at the back of the mission control room, his round, mustachioed face frozen in a nervous grin. Hunkered down at long rows of computer consoles, his engineers sit on the edges of their chairs. NASA’s Jet Propulsion Laboratory is hanging on the brink of a jubilant victory—or a devastating failure.

Then the black-and-white images appear on a big projection screen, and the room explodes in cheers. Some 200 million kilometers from Earth, a little robotic rover called Spirit, built here in Pasadena, Calif., has awakened and called home, sending images of what it is seeing. And what it is seeing is the rocky plain of Gusev Crater, in the southern highlands of Mars.

Theisinger, the project manager of the Mars Exploration Rover mission, has only one word: “Wow.”

Wow, indeed. Since that night in 2004, Spirit and its twin rover, Opportunity, which landed three weeks later, have embarked on an extraordinary journey of discovery. Designing, constructing, launching, and landing those rovers on Mars has become NASA’s most thrilling and successful planetary mission ever.

Why bother to study rocks and dirt from a cold, desolate, remote world? Because the geology of Mars embodies a history that should help unravel our own, and because those Martian rocks may also hold the answer to a question we’ve been asking ourselves for a very long time: Are we alone?

“Finding evidence that life arose independently on another planet would be one of the most profound discoveries that humans could ever make,” Steve Squyres, astronomy professor at Cornell University and the mission’s chief scientist, writes in Roving Mars (Hyperion, 2005), his candid account of the project. Mars, he adds, “is a world that can help us learn our place in the cosmos.”

In the past decade, planetary rovers have emerged as one of the most amazing exploration tools humanity has ever seen. They have also fostered scientific and technological innovations that should find applications on Earth, in areas such as autonomous robotics, remote sensing, and materials engineering.

Above all, these robotic explorers have demonstrated that unmanned missions offer formidable rewards, with immensely smaller costs and risks than manned ones. Manned missions will surely remain on NASA’s agenda; human boot prints on extraterrestrial soil are too powerful a draw to relinquish. But the success of the Mars rovers has proved that before we send humans, we ought to send robots.

NASA is not alone in advancing rover technology. The European Space Agency, in a joint mission with NASA, is building a next-generation Mars rover, called ExoMars, for launch in 2018. And other countries and even private companies have dreamed up rover plans of their own.

Right now, however, JPL is commanding all the attention. A new rover, the Mars Science Laboratory—named Curiosity in a contest—is scheduled to launch later this year. Compared to its golf-cart-size predecessors, it’s a monster of a machine, the size of a Mini Cooper, weighing in at 900 kilograms, equipped with a nuclear power supply, and carrying 10 scientific instruments of unprecedented sophistication, including an advanced analytical system for detecting organic molecules. The mission: Determine whether conditions for life existed on Mars and were preserved—and if they were, find a sample.

NASA is like a planetary system where personalities, politics, budgets, and schedules revolve around each other in erratic orbits. Every now and then, these celestial bodies align and a promising mission becomes possible. That’s how the idea of sending rovers to explore the geology of Mars emerged in the mid-1990s.

Every space mission builds on its predecessors. Before Spirit and Opportunity, a Soviet program called Lunokhod put an eight-wheeled, solar-powered rover on the moon in 1970 and another one in 1973. Controllers on Earth steered the rovers and operated their cameras and instruments in near real time.

But as interest shifted from the moon to more distant parts of the solar system, flyby probes, orbiters, and landers came to dominate the scene. Rover missions seemed too daunting—or merely unnecessary, in the case of a ball of gas like Jupiter.

In the mid-1970s, NASA’s Viking program put sophisticated landers on Mars to search for signs of life. They didn’t find any, but then again, they were pretty much looking at their feet. That’s the main drawback of landers: They’re stuck in one place. You could have the most intriguing rock sitting in front of you, but if you can’t get to it, you’ll never know what secrets it might hold.

Rovers made a comeback in 1996, as part of NASA’s Pathfinder program. Pathfinder was essentially a lander mission, but it carried a small rover named Sojourner in its belly. The rover successfully roamed around, snapping photos and analyzing the chemical elements of rocks.

The results led NASA to start planning a more ambitious rover mission. In 1999 alone the agency lost two spacecraft, the Mars Climate Orbiter and the Mars Polar Lander, and it desperately needed a successful mission.

It was good timing. JPL had been developing a number of navigation and robotic technologies that it could put to use. After a series of solicitation bids, review panels, and approved, canceled, and newly revived proposals, one idea stood out in particular. Could NASA use Pathfinder’s successful landing approach to place not a lander but a rover on Mars? This approach would rely on slowing the spacecraft as it entered the Martian atmosphere, using rockets and a parachute, and then deploying air bags to cushion the touchdown.

JPL engineers believed it could work. They would strip the Pathfinder lander down to its basic structure, cram a bigger and better-equipped rover inside, and keep the rest essentially the same. The mission got the green light in mid-2000, and the clock started ticking. The launch window—when the orbits of Earth and Mars would be best aligned—was just three years away. That wasn’t a whole lot of time to build a complex planetary rover. Yet NASA told JPL to build two, to double the chances of success.

Cornell’s Squyres led the science team building the instruments that Spirit and Opportunity would take, while Theisinger headed the engineering team working on the design of the rovers. It didn’t take long for them to come to a terrifying realization.

Though things looked good in preliminary studies, when it came time to refine their designs the teams discovered they had underestimated the size the rovers would eventually become. The rovers were too big for them to reuse the Pathfinder landing system. The EDL team—responsible for entry, descent, and landing—scrambled to redesign the parachute and air bag systems. The first tests failed badly, with prototype chutes and air bags blowing to pieces.

Other problems cropped up. Instruments that had been working misbehaved inside the thermal vacuum chamber that simulated the Martian atmosphere. An error in a telecommunications module deep inside Spirit forced engineers to reopen its electronic guts. In Roving Mars, Squyres describes one of the commands the software team sent to the rovers quite often during development: SHUTDOWN_DAMMIT.

At one point, when Spirit was already inside a Delta II rocket awaiting launch, engineers discovered they had accidentally blown the rover’s main fuse during the final assembly. The problem nearly made NASA administrators kill the entire mission.

But in the end, the engineers did what engineers do best: They solved problems, one after another, with solutions that were sometimes ingenious and other times just good enough.

The machines they created—each cost some US $400 million, including launch and operation—are beautiful pieces of technology. Their solar panels unfold like origami. Their so-called rocker-bogie suspension systems allow each wheel to move up or down independently so the vehicles won’t tilt excessively. And their software lets them receive navigation instructions and then drive autonomously, avoiding big boulders and stopping before cliffs.

With panoramic cameras, a microscopic imager, a rock abrasion tool, and three different spectrometers, the rovers made many discoveries about the geology and mineralogy of Mars. Among the most important was convincing evidence that Mars once had lots of water, an ingredient essential to life as we know it.

Expected to last 90 days, the rovers have worked for seven years. Spirit is stuck in sand and probably hasn’t survived the last Martian winter, but Opportunity is still roving around. These robots answered many questions and also raised fresh ones. And that’s why NASA is going back—this time with a bigger rover.

Few engineering projects compare to building a spacecraft in terms of cost, complexity, and risk. These are one-of-a-kind machines that will face the harshest conditions—crushing accelerations, extreme temperatures, radiation storms—while stuffed with sensitive instruments and moving parts. And of course, once they leave the launchpad, there’s no recalling them. The number of organizations that can build jet fighters and nuclear reactors is small. Fewer still can build spacecraft. JPL is one of these.

Tucked in a small campus in the San Gabriel foothills, JPL is NASA’s lead facility for the exploration of the solar system. Hardware built here has flown to the moon, Venus, Mars, Mercury, Jupiter, Saturn, Uranus, and Neptune.

Late last year, when I visited Theisinger, he proudly showed me photos and decals on the walls of his cramped office—trophies of his past victories. The 40-year JPL veteran is again leading an engineering team, the one building the Curiosity rover.

Compared to Spirit and Opportunity, the new rover is “an order of magnitude” harder to design and build, Theisinger says.

With a massive 2.3-meter-long arm, the rover will be able to push a percussion drill against rocks to extract samples. Wheels half a meter in diameter will let it traverse difficult terrain, off-road-style. And thanks to its plutonium-238 thermoelectric generator, it will be able to operate in the winter and at latitudes farther from the equator.

These capabilities are key to the success of Curiosity’s mission. To find out whether environments habitable for microorganisms ever existed, the rover will have to go to places and perform scientific experiments beyond what its predecessors could do. The goal is to better understand how the availability of water, energy, and elements like carbon evolved on the surface of Mars.

"With [Spirit and Opportunity], what we can do is imagine what might have happened," says Caltech geologist John Grotzinger, the mission’s chief scientist. "With Curiosity, we’re going to determine what actually happened."

Curiosity will carry 10 instruments from five countries. The most important is known as SAM, or sample analysis at Mars. This set of instruments takes in a rock or soil sample and uses a mass spectrometer, a tunable laser spectrometer, and a gas chromatograph to characterize its molecular structure and isotopic composition, and also to test for the presence of organic carbon.

Getting the 80 kilograms of science hardware—Spirit and Opportunity each carried 5 kg—to fit on the rover was one of the project’s biggest challenges. But perhaps even harder was the design of the landing system. The new rover is so big and heavy that air bags won’t work. Like Spirit and Opportunity, Curiosity will travel aboard a capsule known as an aeroshell, and after it enters the Martian atmosphere a parachute will unfurl. But then, still plunging at supersonic speeds, the craft will fire retro rockets, decelerating to a gentle descent until it’s just 20 meters from the ground. That’s when the rover will detach from a supporting structure and lower itself on cables—much like a commando rappelling from a helicopter.

When the rover touches the ground, explosive bolt cutters will release the cables, allowing the aeroshell to fly away and crash at a safe distance. Steve Lee, a member of the EDL team, describes all this—the "sky crane," they call it—with a satisfied look on his face. When I suggest that this would make for a good movie, he smiles even more broadly: A downward camera on the rover would capture not only a top view of the landing site but also the entire touchdown action.

NASA knows the power that images of extraterrestrial worlds have to capture people’s imaginations. So it’s no surprise the rover is also equipped with cameras capable of obtaining high-definition stills and movies, which could be converted into 3-D. (James Cameron is a member of the camera team. Really.)

Knowing what the JPL crew went through with their previous rovers, you might think they’ve grown too bold. Can they pull it off?

The project has already hit some major snags. A novel lubricant-free motor that NASA wanted to use failed during tests, forcing the engineers to go back to traditional designs.

Another setback involved a turbomolecular vacuum pump in the rover’s SAM instrument. It had to be redesigned and retested, delaying its delivery—and leaving project managers wringing their hands.

But as before, the engineers march forward. At JPL’s vast clean room, construction and testing of the rover’s final components proceed at a frantic pace. Launch is scheduled for late November or early December.

Why go to Mars? There are many reasons, JPL engineers will tell you. But some on the team have personal motivations. Pete Theisinger says he once received a letter from a man in Ohio. The man wrote that one day he was watching TV with his young son, and they saw Spirit and Opportunity and the images their cameras had captured of the Martian landscape. The letter included a photo of the man’s son. He was building a rover out of Lego blocks.

This article originally appeared in print as “Are We Alone?”

For all of IEEE Spectrum’s Top 11 Technologies of the Decade, visit the special report.