Deep inside a giant but little known NASA facility, crews have for years been staging elaborately faked space missions. This is not a conspiracy theory. It is the sad tale of NASA's Michoud Assembly Facility, the sprawling New Orleans complex where the space agency had for decades built its biggest rockets.

After the space shuttle's last flight in 2011, Michoud's massive hangarlike facilities were rented out to Hollywood studios, housing some of the production for Ender's Game and other science-fiction movies.

But lately a growing cadre of NASA engineers and other workers have been engaged on an important new production here—a sequel to the agency's greatest days of human spaceflight. Michoud is back in the rocket-making business, serving as a factory for the biggest, most ambitious space vehicle ever to undergo construction: the Space Launch System, often called by its acronym, SLS.

The SLS is the rocket in which NASA hopes to thunder a crew of astronauts skyward from Cape Canaveral, Fla., for roughly a year's journey to the surface of Mars while hauling the living quarters, vehicles and supplies they will need to spend at least a few weeks shuffling through the rusty dust there. That mission is still about 25 years away. But between now and then, the SLS could carry people to Earth's moon and an asteroid and send a probe to search for life on Europa, one of Jupiter's moons. It is an interplanetarily groundbreaking project, one of the most audacious NASA has ever undertaken.

Why, then, do so many people seem to hate it?

Replacing the Shuttle
After the giddy triumph of the Apollo moon exploration program in the 1960s and early 1970s, the space shuttle was supposed to make Earth-orbit access relatively cheap and routine. Instead the shuttle averaged more than $1 billion a trip, flew only a few times a year and was twice afflicted by catastrophe. In 2004, a year after the Columbia disintegrated on reentry, killing seven people, President George W. Bush charged NASA with replacing the shuttle with a more Apollo-like program that would bring us back to the moon and then to Mars. The resulting effort, called Constellation, led to the design of two new Ares rockets, a crew launch vehicle and a giant, Saturn V–like version intended to haul cargo. But by 2011, after having burned through some $9 billion, all Constellation had produced was an Orion crew capsule that was being constructed by Lockheed Martin and a rocket that had been launched once as a test. President Barack Obama canceled the program, directing NASA to refocus its energy on a mission to an asteroid. The agency was to turn to the private sector for an orbital ferry service to get cargo and crew to the International Space Station (ISS).

Still, many in Congress pushed hard to continue the quest for a new heavy-lift rocket capable of getting humans to the moon and Mars. The resulting compromise was the SLS, a single big rocket for both crew and cargo that would eschew much of the new technology planned for Ares and instead rely on space shuttle engines, boosters and tanks for most of its kick. The SLS was Ares on the cheap.

From the beginning, the SLS has been dogged by the perception that Congress cooked it up to protect jobs at NASA and its major contractors. “This vehicle has the distinction of being the first rocket designed by a committee of politicians rather than by scientists and engineers,” wrote the editors of the Economist last December. Some critics deride the SLS as the “Pork Rocket” or “Senate Launch System.” Southern senators whose states are home to large NASA or contractor facilities have indeed been the SLS's loudest proponents in Congress. Supporters include, for example, Senator Richard Shelby of Alabama—some 6,000 people are employed at the NASA Marshall Space Flight Center in Huntsville, Ala., where the SLS is managed—and Senator David Vitter of Louisiana, home to NASA's Michoud facility, where SLS core-stage prime contractor Boeing is deploying many of the 1,500 people it already has working on the program.

And a big program—and rocket—it is. The SLS will initially have a bottom core stage powered by four RS-25 space shuttle engines that use standard liquid hydrogen and oxygen fuel. Attached to each side of the core stage will be solid rocket boosters, which provide the extra push needed to get the heavy rocket airborne [see illustration on next page]. A second stage, atop the first, will take over at an altitude of about 50 kilometers to push the rocket into orbit, and the Orion crew capsule will sit on top of the entire structure. At 98 meters, the rocket will be slightly shorter but more powerful than a Saturn V, which powered every manned mission to the moon, and will carry three times the payload of the shuttle. None of the components are designed to be reusable. Over the next decade, SLS upgrades will include more powerful engines and boosters. The eventual Mars-capable SLS would get even more power in its upper stage, giving it twice the thrust of the first version.

Critics charge that by specifying that the SLS rely on shuttle components, Congress ensured that the shuttle's big aerospace contractors would profit. “Once again, Boeing is making out like a bandit,” says Peter Wilson, senior defense research analyst at RAND Corporation. Others contend that the shuttle-recycling approach will leave the SLS a troubled Franken-rocket with stitched-together parts from a dead program. The use of the shuttle boosters has already led to a problem with gaps in heat insulation, for example.

Estimates of the SLS's final cost vary wildly. NASA has publicly projected that it will take $18 billion to get the SLS to first launch—$10 billion for the rocket itself, $6 billion for the Orion crew capsule and $2 billion to get Cape Canaveral fitted to handle SLS launches. (Incidentally, Senator Bill Nelson of Florida is another big supporter of the SLS.) But a leaked internal study came up with a cost of more than $60 billion over the next 10 years. Others predict that delivering a crew to Mars will cost up to $1 trillion. NASA's stated target is $500 million per launch, but others have put it as high as $14 billion when all program costs are figured in.

Critics insist that the government and public will never back their enthusiasm for space exploration with the many hundreds of billions of dollars the SLS's grandest missions will require. Several analyses, including one internal study performed by NASA, have suggested that we can get to deep space and Mars without a heavy-lift rocket. It might be cheaper, some argue, to rely on smaller rockets akin to the Delta IV, used for about a decade to launch satellites, to heft into low-Earth orbit the fuel, components and materials needed to construct deep-space vehicles and then build the big craft there. And if it turns out we do need a giant rocket, many say, why not turn the job over to so-called new space? SpaceX, the company founded by Silicon Valley icon Elon Musk, has already won orbital ferry contracts with NASA using its well-regarded Falcon 9 rockets. The “SLS is only adding small incremental improvements to technology developed 40 years ago,” says James Pura, president of the Space Frontier Foundation, an advocacy group dedicated to advancing space exploration. “NASA ought to tell private industry what sort of payload it wants to get into deep space, offer a set amount of money for the job and let companies like SpaceX build it.” SpaceX is developing a 27-engine, SLS-class heavy-lift rocket and is working on new, more powerful engines that, if successful, would allow that rocket to outpull even the largest envisioned SLS. And SpaceX is designing all its major components to be reusable; the SLS, in contrast, is entirely disposable.

Despite these objections, SLS mission planning is under way. A 2018 first flight will send a crewless SLS and Orion out well past the moon, and a second, not yet formally scheduled flight will do much the same with a crew perhaps a few years later, taking humans farther from Earth than ever before. What happens after that will ultimately be up to Congress and a new president, but right now a crewed asteroid visit is tentatively planned for the mid-2020s, with a human mission to Mars to follow in the 2030s.

The Rocket Factory
NASA tests its biggest rockets at Stennis Space Center, which lies in a web of lakes, rivers, bayous and canals near the southernmost tip of Mississippi. As we gear up in hard hats and safety vests, Tom Byrd, who until his retirement in January was a NASA deputy manager here, tells me there are three reasons for the center's proximity to water: the activities at Stennis require access to large barges, to marine construction expertise and to a ready way to cool giant slabs of metal exposed to temperatures approaching those found on the surface of the sun.

Each test stand here is a huge metal-and-concrete structure that looks something like a cross-sectional slab taken from the middle of a mega ocean freighter. We climb up through one of the stands, and along the way I am shown a control room that would not look out of place in a circa 1950s Soviet power plant—mostly steam gauges and big, clunky dials. I ask why they have not been upgraded to digital panels. The answer is one that will prove to be a sort of mantra for the SLS program: it has taken decades to get this stuff to work well despite unfathomable forces and innumerable glitches, so why mess with it?

From the top of the stand, however, I can see that Stennis is actually awash in upgrades. Canals and roads are being reworked to handle larger loads, and the test stands themselves are getting renovations and reinforcements because the SLS is going to subject them to greater stresses than any previous rocket. “The forces generated here are bigger than during actual launches because a rocket in the test stand can't escape its own plume,” explains Byrd. Throughout an approximately nine-minute test-firing, thousands of nozzles will shoot high-pressure jets of water at the stand's walls—not for cooling but to tamp down ferocious vibrations that could otherwise rip the stand apart. Even before the SLS, no private structure was allowed within 13 kilometers of the stands because the sound waves alone from a test could shake it apart. And the SLS engines will generate the most powerful rocket thrust ever produced on Earth.

Just across the Mississippi-Louisiana border, a few hours away via canal (or, in my case, 45 minutes by car), sits Michoud, which I visit the next day. In contrast to the isolation of Stennis, Michoud is in the middle of an industrial area on the outskirts of New Orleans. In some ways, Michoud is a factory like any other, with welding stations, forklifts, cranes and parts bins. It is just all done on a much larger scale.

Inside, Michoud is gleaming. To tour the complex is to watch it fill up, minute by minute, with new gear—towering robot arms that can move at blinding speed, wheeled platforms and cranelike handlers that whisk components weighing tens of metric tons from one station to another, parts-organizing systems that ensure that an engine consisting of hundreds of thousands of parts does not end up with one too many or few. When you build a machine as powerful as an SLS rocket engine, you must have a very low tolerance for assembly deviation. “If our parts-tracking system told us that one of these tiny washers here is left over, all work would stop until we found it,” says Patrick Whipps, one of NASA's managers at Michoud.

Many of the components that will go into the rockets built here originated in other vehicles. “We're not going to have many one-of-a-kind components on the SLS,” says William Gerstenmaier, the NASA associate administrator who heads up the agency's human space exploration efforts. Yet new manufacturing equipment and methods should make the those components much less expensive to build than they have been in the past, Whipps adds. Upgrades include a friction-stir welding machine the size of a municipal water tower tank. Massive aluminum-alloy rocket sections can be dropped whole into this leviathan, where drills will meld the two sections together. It is the largest machine of its type in the world.

The SLS goes beyond shuttle technology in many other ways as well. To analyze the stresses on the SLS from buffeting and other aerodynamic instabilities during its climb through the atmosphere, NASA turned to state-of-the-art fluid dynamics software. Without it, the engineers would have had to redesign the rocket to provide more stress resistance to cover a much bigger margin for error. In addition, new avionics and digital controllers relying on computer chips that are several generations ahead of those used in the space shuttle will enable automated flight and engine controls to react many times faster to sudden changes and dangerous conditions.

Leftover shuttle engines will get the SLS airborne for the first four flights, but new versions will be needed starting in the 2020s. For those, NASA is using machines that will produce the thousands of required coin-sized turbine blades by laser-welding powdered metal into the right shapes instead of individually machining them, cutting production time for an engine's worth of blades from a year to a single month. “We're using computer control everywhere to minimize labor costs and improve precision,” Gerstenmaier says.

The Case for the SLS
When the SLS program is in full swing, the aim will be to turn out at least two rockets a year—possibly as many as four. In the rocket world, that is mass production. But it will grind to a halt if NASA cannot convince the American public that the SLS is worth building.

The two broadest objections—that $18 billion is too much to spend on a rocket and that we should focus on sending probes and robots, not humans, into space to do science—can be addressed as matters of perspective. Eighteen billion dollars is not all that much for the capability of sending humans to another planet and back; it cost a third more than that to improve traffic flow in Boston via the “Big Dig.” It is easy to claim there are cheaper ways of doing it, but NASA's success and safety records have set the bar high, and it is unlikely that the American public would put up with higher chances of a catastrophic failure in order to shave off what amounts to a few thousandths of the federal budget.

As for sticking with probes and robots, the case is often made that the science haul from a human-crewed mission is likely to be bigger than what a probe or rover can deliver. But the real justification for human spaceflight is to take steps toward expanding the human race's stomping grounds.

The SLS does have many fans. These supporters include NASA's current leadership and rank and file, a number of space experts and a growing chunk of the American public, much of which was thrilled last December by the flawless orbital flight of the Orion crew capsule that will be sitting atop the SLS when it heads into deep space. The experts among them can easily argue, point by point, with the critics.

Use smaller rockets to heft components and fuel into space for orbital assembly? Some 500 metric tons of materiel will be needed for a crewed Mars mission, Gerstenmaier calculates. That is a feat that the SLS could manage in four launches but that would take at least two dozen launches of a maxed-out Delta IV. Gerstenmaier contends that every one of those launches raises program risk a bit because the worst things are most likely to happen in the first minute of a mission. The approach is also more vulnerable to delays, with the effect of stretching out individual launches cumulatively across all the launches. “We used the many-launches approach with the space shuttle to build the space station, and it ended up taking decades,” he says.

But the most significant potential drawback to a lift-it-in-small-chunks approach, Gerstenmaier says, is the massive amount of in-orbit construction that would be required, including of habitats, interplanetary vehicles and fuel depots. That is a daunting task, given our limited experience with the very tricky craft of in-space assembly. “You'd have a huge number of dockings; you'd be fabricating in space,” he says. “Inevitably some of the pieces wouldn't work right and would be difficult to fix there. It adds an enormous amount of complexity and risk.” The SLS's sheer girth will also allow packing in bulkier, ungainly payload shapes up to 10 meters across, such as those with solar panel and antenna arrays, that otherwise would have to be complexly folded and thus more vulnerable to damage or malfunction.

Another big advantage to the heavy-lift route: some of an outsize rocket's extra thrust can be converted into higher speeds that get spacecraft to their destinations more quickly. That is a critical consideration for crewed flight to Mars, where radiation exposure and supply requirements set tight upper limits on mission duration. Distant robotic missions benefit, too, because planning for follow-up missions has to wait for data to come in from predecessors to maximize the scientific returns. Because of its sheer power, the SLS can send missions into deep space using its own fuel, as opposed to gravitationally slingshotting around planets as the Voyager and Galileo missions did.

The “SLS will cut the time for a Europa visit from six-plus years to 2.5 years,” says Scott Hubbard, a consulting professor of aeronautics and astronautics at Stanford University. “It would be an enabler for a very compelling scientific mission.” Add these shorter transit times to the higher payload masses and packaging flexibility, and you have a powerful case for a heavy-lift rocket. That helps to explain why both China and Russia are working on SLS-class designs.

The same goes for SpaceX. Yet new space is not as natural a source for deep-space rockets as it is for transport to the ISS and back. There is no existing market, and none envisioned, for deep-space exploration beyond the handful of missions NASA has tentatively planned for the SLS. That eliminates the opportunity for SpaceX to leverage development costs for a heavy-lift rocket over various commercial customers, as it has with its smaller rockets. Stripped of that advantage, SpaceX is no better positioned than Boeing, Lockheed Martin and other conventional aerospace contractors, says former NASA astronaut Scott Parazynski, a veteran of five shuttle missions who is now at Arizona State University. “Those are very capable contractors, and I don't see SpaceX in a dramatically different light,” he explains.

Hewing to the tried and tested instead of innovating might be a recipe for failure in the automobile, cell phone or software industries, but when it comes to zipping a crew of heroes into deep space on the wings of a barely controlled explosion, a certain level of conservatism is not necessarily a bad thing. SpaceX suffered several explosions and losses of control in its earlier rockets—par for the course in the development of new designs. Last October a crew member was killed when a craft that Virgin Galactic built to bring tourists into suborbital space crashed during a test flight—just three days after the explosion of a crewless rocket built by private company Orbital Sciences, one that was headed to the ISS.*

These accidents serve as reminders that in spite of decades of experience, rocketry is hard. It carries a high risk of pure catastrophe. That is one reason leaders at the Inspiration Mars Foundation, a privately funded organization that has been trying to facilitate a mission to Mars, are among those who have, after initial skepticism, been lining up behind the SLS. Other Mars experts agree. The “SLS has been criticized from day one as a rocket to nowhere,” Hubbard says. “But it now has clear-cut, defensible missions, and it's time for everyone to get behind thinking about how we can make sure it all comes together.”

Escape Velocity
For 500 seconds on a cool night this past January, one of the Stennis Space Center's hulking engine tests turned into a fireball. It was the first test of an R-25 shuttle engine since 2009, and it went perfectly. If the successful tests keep coming, time may be on the SLS's side. The longer the program lasts—if it remains on budget and on time—the more it will stand as its own proof of concept. In its first three years, the program has achieved smooth and rapid progress, gliding through design reviews and entering into early manufacturing steps. That is blindingly fast for a major new human-rated rocket. Only a few glitches have cropped up; those insulation gaps were just about the worst of them, and the problem was quickly fixed with a layer of adhesive.

Anything could happen in the years ahead, under new presidents and congresses, contends Joan Johnson-Freese, a professor at the U.S. Naval War College who specializes in space. Maybe the consensus in government will become that we should abandon Mars for now and focus on setting up a base a little closer to home. “Some in Washington have an almost criminal nostalgia for the moon,” she says. Others think NASA should forget both the moon and Mars for now and concentrate on asteroids, not only because they may contain answers to important questions about the origins of the solar system but also because we might learn how to divert or destroy any that end up heading toward Earth.

But the allure of Mars remains widespread. Lately that allure has been building, as it dawns on more people that we could reach the Red Planet within their lifetime. “We'd all like to see us go there,” Parazynski says. “Other missions would be a distraction.” He has concerns about the SLS, but not because he thinks it is a lousy way to get to Mars. He worries that because it will not be cheap or immediate, we will abandon the SLS before we get there.

At the moment, there are no showstoppers in sight for the SLS. That claim alone, which cannot be made for any alternative Mars rocket proposal, may ensure that the project stays the course. Sure, it was cobbled together from congressional mandates. Yes, it lacks the innovative verve of rival schemes. But there is every indication it will work as planned, and it is funded for the foreseeable future. That should be good enough to make the SLS the rocket that takes us to Mars. And if it does, the criticisms will be quickly forgotten.

*Editor's Note (7/23/15): The original sentence from the magazine article was corrected online after posting. It erroneously stated that Virgin Galactic's spacecraft was a prototype and that it exploded.