Atlas V and Delta IV rockets in final assembly at ULA’s 1.6-million-square-foot plant in Decatur, Ala.
Photograph by Spencer Lowell for Fortune

THE GREAT ROCKET RACE

Elon Musk’s SpaceX is jousting with a Boeing-Lockheed joint venture for the lucrative business of sending satellites to space. The incumbent faces a daunting two-pronged challenge: Can it slash prices while overhauling a key part of its technology?

In recent years Elon Musk’s SpaceX and Jeff Bezos’s Blue Origin have rekindled the popular excitement that once surrounded the space program. The billionaires’ rocket startups have audacious aspirations—Musk talks of transporting passengers to Mars—and they’ve already chalked up some impressive accomplishments, including a series of missions to space. SpaceX even managed to land a spent rocket booster on a floating stage in the Atlantic Ocean. Faster, cheaper, reusable—theirs is the epitome of the disruptive vision.

Of course, if you’re the disruptee, you might take a more jaundiced view of all this. Tory Bruno, though, betrays no hint of irritation. He is the CEO of an entity with a bland, instantly forgettable name: United Launch Alliance. You likely haven’t heard of ULA, but you know its corporate parents, Boeing ba and Lockheed Martin lmt . A decade ago the defense and aerospace titans spun their noncombat rocket operations into a fifty-fifty joint venture.

ULA builds rockets that propel all manner of commercial and military satellites into orbit. Since 2006 the company has lofted 111 such payloads into space without losing a single one to accident or mishap. That’s a striking feat, one that Bruno repeatedly cites. He doesn’t need to mention that his upstart rivals have not come close to achieving this sort of record.

SpaceX and Blue Origin pose a long-term threat to ULA. But Bruno has a much more pressing short-term problem. His company is running out of the engines it depends on to power its workhorse Atlas V rocket, which could strand the rockets—and ULA’s revenues—on the ground.

ULA CEO Tory Bruno (left) in front of one of the company’s Delta IV rockets; An Atlas V (right) launches with a NASA space probe. ULA CEO Tory Bruno (left) in front of one of the company’s Delta IV rockets; an Atlas V (right) launches with a NASA space probe.Photos, Bruno: Courtesy of United Launch Alliance; Launch: Aubrey Gemignani-NASA via Getty Images

Indeed, finding a new engine is crucial to an even more daunting second challenge for ULA: designing a new, drastically lower-cost reusable rocket that will suit an era of fiscal restraint in Washington, one that is being facilitated by the presence of SpaceX, in particular, as a competitor. “If I boil it down to its simplest terms,” Bruno says, “we need to cut our prices in half while not losing the magic of our reliability. For an aerospace company, something like that usually takes about 10 years. We have to get it done in three years flat.”

ULA’s current troubles would have been difficult to envision just a few years ago. Until April of this year the company maintained a lucrative monopoly on the U.S. national security establishment’s satellite launch business, contracts that generated hundreds of millions of dollars per launch. How that monopoly crumbled is an unexpected tale, one that involves not only the likes of Elon Musk and SpaceX—but also Russian President Vladimir Putin, U.S. Sen. John McCain, and events ranging from Washington, D.C., to Crimea.

Perhaps most surprising to the layperson is this twist: For more than two decades the rockets that carry countless U.S. government satellites—including those that spy on Russia—have contained engines made in … Russia. That practice is coming to an end, leaving ULA in the (slightly less surprising) position of relying on upstart Blue Origin for a new—as yet undeveloped—engine.

A fuel tank for an Atlas V rocket at ULA’s factory.Photograph by Spencer Lowell for Fortune

Bruno, a 54-year-old mechanical engineer who has spent his entire career at Lockheed and ULA, is soft-spoken and analytical—so analytical, in fact, that he doesn’t even blink when asked whether his company is facing extinction. He has publicly acknowledged that ULA “cannot survive” if it doesn’t find a way to compete in the new environment.

His office at ULA’s headquarters outside Denver is standard-issue CEO, apart, perhaps, from the scale models of ULA rockets and a statuette of the legendary St. George slaying the dragon. (Bruno has a fascination with the Crusades, as we’ll see.) Even now, it’s hard to view mighty Boeing and Lockheed as underdogs in this contest.

But that’s the nature of the disruptee’s plight, and it leaves the existential question, Can Bruno slay the dragon? Developing a radically cheaper rocket built with a new engine in three years, as noted, is a very tall order. “Based on what I’ve seen in the last quarter-century watching companies develop rockets, it’s an extremely ambitious schedule,” says Marco Caceres, director of space studies at aerospace consultancy Teal Group. “If I were to bet, I’d say they can’t do it. But—and it’s a big but—I think they realize that this is an extraordinary situation for them. If they don’t do it, they may not be in the launch business much longer.”

It would be hard to overstate how deeply Boeing, Lockheed, and ULA have been tied to the U.S. space and missile programs. Boeing, for example, was one of three companies that built the Saturn V rocket, which took Apollo 11 into outer space on its way to the moon. Boeing eventually acquired the other two companies involved in the project. From the Mercury to the Minuteman, Boeing and Lockheed (and the countless companies that disappeared into their maw) encompass much of the history of U.S. space rocketry.

The role of private companies has intensified over the decades. In the old days Boeing or Lockheed would build a rocket, and the U.S. Air Force or NASA would handle most launch operations. In a series of evolutions that was capped, in 1995, by an Air Force program with the cumbersome name Evolved Expendable Launch Vehicle (EELV), the government handed even more responsibility to private operators. Now companies not only construct the rockets but also manage the launches and run the missions.

Atlas V tanks in the tank center at ULA's plant in Decatur, Ala. A finished Atlas V 500 series vehicle stands 205 feet tall – almost 19 stories. When finished, Atlas V tanks can stand almost 19 stories tall.Photograph by Spencer Lowell for Fortune

ULA (and Lockheed and Boeing before it) have long outsourced their large-rocket engines, only one part of the complex synthesis of engineering, materials, and design that allows a long tube filled most of all with huge quantities of fuel to shoot satellites into space. It costs roughly $1 billion to develop a large liquid-fuel rocket engine from scratch, and in the wake of decades of consolidation, very few companies can do it.

Cost is always an issue in aerospace and defense contracting. Indeed, another element of the 1995 EELV program was an initiative to cut rocket prices in half. That was meant to occur through competition, with the additional benefit that the Department of Defense would be guaranteed a minimum of two rocket lines to choose from, providing redundancy in case one line suffered a technical problem.

ROC.11.01.16.Sidebar

The move to save money partly explains how U.S. rockets came to be using Russian engines. In the early and mid-1990s, the Cold War had just ended; the Soviet government was gone. An era of good feeling, combined with the cost-savings plan and the Russians’ need to find new customers now that its old one had ceased to exist, led to what previously would have been unthinkable: The U.S. Department of Defense authorized the purchase of RD-180 first-stage rocket engines, manufactured by Russia’s NPO Energomash. Russian rocket-propulsion technology was proven, reliable, available, and priced to move. The RD-180 made its way first onto Lockheed’s Atlas III and later into the Atlas V.

The 50% cost reduction never materialized (surprise, surprise), but the issue periodically popped up. A decade later, in 2005, Boeing and Lockheed proposed merging the businesses into one entity (which would maintain the separate Delta dal and Atlas rockets). A single builder would operate more efficiently, the reasoning went. ULA was officially born the following year.

As that was happening, two ultra-rich entrepreneurs had begun preparing rocket startups. Elon Musk, best known now for electric-car company Tesla tsla , founded SpaceX in 2002. Jeff Bezos, of Amazon amzn , stealthily incorporated Blue Origin two years prior, in 2000, but the company’s existence wasn’t made public until 2003.

The two projects were initially dismissed by many as the indulgences of billionaires, expensive hobbies or wealth-destroying follies, depending on one’s point of view. But the two men weren’t seeking the space equivalent of the mine-is-bigger-than-yours yacht. Musk has spoken about his ambition to ultimately make humankind a “multi-planet species” through the establishment of a human presence on Mars. Bezos, though less vocal, has a similarly lofty vision. “I want millions of people living and working in space,” he told the Space Symposium in Colorado Springs in April. “I want us to be a space-faring civilization.” (A SpaceX representative was interviewed for this article but declined to be quoted; a spokesperson for Blue Origin said the company couldn’t provide anybody to speak in time for Fortune’s deadline.)

A SpaceX reusable rocket-stage approaches landing on an ocean platform (left); a Blue Origin reusable rocket booster.Photos, SpaceX: NASA via Getty Images; Blue Origin: Courtesy of Blue Origin

Both companies want to make space more accessible and profitable, and that involves getting people and hardware into space at a lower cost. For Blue Origin, that means building a space tourism business around its suborbital New Shepard spacecraft, which in theory will generate revenue while making human spaceflight far more routine. SpaceX aims to further its interplanetary ambitions by providing companies with low-cost access to orbital space and beyond aboard its Falcon rockets. Lowering the cost of launch is a core aspect of SpaceX’s mission, Musk says, as doing so not only is good for SpaceX’s business but also drives innovation in space-based technology and enables new business models. “We must bring in more money than we spend,” he told reporters after a Falcon 9 flight last year, “but maximizing profitability is not really what it is about.”

SpaceX’s offering is dramatically more economical than ULA’s. It advertises its Falcon 9 rocket at a launch price of $62 million, compared with ULA’s “list prices” of anywhere from $164 million to $350 million for an Atlas V launch (the terms of any particular launch are typically not disclosed). Undercutting the existing players wasn’t so much rocket science as smart business—SpaceX’s offerings were designed and built using the most up-to-date technologies, and the company is completely vertically integrated. Unlike ULA, SpaceX makes all its critical components in-house, including its engines. Blue Origin also makes its own engines.

In 2014, ULA’s fortunes began to turn. That March, amid a civil war in Ukraine, Russia seized and then annexed the Ukrainian territory of Crimea, causing a rift in Russo-American relations. In April, SpaceX sued the Air Force for the right to compete for military launch contracts.

In May 2014, deputy prime minister Dmitry Rogozin—the political overseer of Russia’s defense and space industries and one of the first individuals targeted by U.S. sanctions following Crimea’s annexation—threatened to cut off American imports of RD-180s for military satellite launches. That jeopardized the Pentagon’s access to space and illustrated precisely why one wouldn’t want crucial U.S. security technology to rely on Russian equipment.

Rogozin didn’t follow through on the threat—but the Russian engine was now a political hot button. Congress got involved. By summer the U.S. government began seeking ways to wean itself from the Russian engine. Congress decided to limit ULA’s import of RD-180s to five it had already ordered (though that number would later rise).

In January 2015 came a second blow to ULA: In response to SpaceX’s lawsuit, the Air Force agreed to open some launches to competitive bidding. Meanwhile, SpaceX was seizing the public imagination, making headlines with its attempts to land a rocket booster on a floating platform—part of the company’s long-term goal to reuse its costly first-stage engines, which would drive its prices lower still.

By this point, ULA had begun to respond. In August 2014, Boeing and Lockheed tapped Bruno—then a vice president overseeing Lockheed’s strategic and missile defense systems unit—to take over as ULA’s chief executive. The company needed someone to help it adapt to its new reality.

Bruno began cutting costs, thinning ULA’s executive ranks by 30%. He also slashed 36% from supply chain costs (suggesting that fat abounded) and cut the time to build and deliver a rocket in half. Bruno set a new mission: to develop a new, less costly, partly reusable rocket to replace both the Atlas V and Delta IV. Dubbed Vulcan, it marks the first time ULA is developing a rocket from the ground up. Its decision to make it partly reusable suggests just how much SpaceX has set the pace within the launch industry.

Despite these changes, ULA’s value proposition remains reliability. “I don’t want to be the best price provider,” Bruno says. “I want to be the best value provider.” He doesn’t mention—he doesn’t have to—the SpaceX rocket that exploded in September with a $200 million satellite onboard (which Facebook planned to use to provide Internet access in Africa) or the SpaceX Falcon 9 that blew up en route to the International Space Station in June of last year.

RD-180 engine mount for Atlas V rocket, left, and Delta IV in final assembly, right at ULA’s plant in Decatur, Ala. The rockets will travel by a large cargo ship, "The Delta Mariner" from Decatur, along the Tennessee River to the open ocean, the trip takes 8 days to get to Cape Canaveral Air Force Station, Fla., and 21 days to get to Vandenberg Air Force Base, Calif. At left, a mount for a Russian-made RD-180 engine. At right, a Delta IV rocket in final assembly. The rockets travel by cargo ship via the Tennessee River to the ocean to launch sites. It’s an 8-day journey to Cape Canaveral in Florida and 21 days to Vandenberg Air Force Base in California.Photographs by Spencer Lowell for Fortune

Bruno acknowledges that ULA is still adjusting to a new environment. When the company formed a decade ago, he notes, the Iraq and Afghanistan conflicts were in full swing. The Air Force needed a lot of hardware put into orbit quickly, and the focus was performance. Now, though, cost is paramount, which puts ULA at a distinct disadvantage. “Everything we’d done up to that point didn’t really apply anymore,” Bruno contends.

To the outside world, ULA didn’t look all that interested in entering a new world of competition. In 2015, when SpaceX was eligible to bid for the first time, ULA declined to vie for a Global Positioning System III satellite launch. ULA cited criteria that favored price over reliability and past performance as reasons for opting out, along with a lack of available RD-180 engines. SpaceX won the contract by default, ending ULA’s defense sector monopoly.

For more on the space race, watch this Fortune video:

ULA’s decision to sit out the bidding alienated some key members of Congress. McCain—a vocal advocate for an RD-180 ban who has considerable sway, given that he chairs the Senate Armed Services Committee—wrote a letter to Defense Secretary Ashton Carter calling ULA’s excuses “dubious” and “manufactured.” McCain wrote, “ULA’s use of these tactics is unacceptable. It artificially created a need for relief from legislative restrictions on its ability to continue using RD-180.”

But ULA also wields power in Washington, and this year it scored its own victory: It persuaded Congress to allow it to use up to 18 more RD-180 engines for military launches—enough to keep ULA in business into the early 2020s but not much longer. (McCain won one concession: RD-180s will be prohibited from national security space launches starting in 2022.)

“I don’t want to be burned at the stake,” says CEO Bruno, comparing ULA to the Knights Templar after the crusades.

This episode supports a classic narrative about an incumbent government contractor. ULA, in this view, is a cash machine for two aerospace and defense behemoths that for decades have used their pull on Capitol Hill to maintain a monopoly on federal launch contracts while charging astronomical prices to American taxpayers. Ensconced in its monopoly, ULA never had an impetus to lower costs or develop a breakthrough product (such as a reusable rocket), and now the market has come along to sweep it into the dustbin of history.

Bruno doesn’t flinch at the suggestion of this second narrative, though he certainly disagrees with it. Whether you view ULA as a lazy monopolist or simply a company facing new competition, it needs to transform itself or be left behind—a classic adapt-or-perish scenario that he illustrates through one of his favorite subjects. Bruno is something of a scholar on the Knights Templar, the medieval order of European knights most famous for their forays into the Holy Land during the Crusades. He has written two books about the group, (one of them called Templar Incorporated) that extrapolate modern management lessons from the 12th-century organization, whose vast banking, shipping, and security businesses constituted the West’s first multinational conglomerate. “For 187 years they were at the top,” he says. “They were the biggest company, if you will, for all that time. Then, within about seven years, they were extinct—the CEO was literally burned at the stake.”

What changed? Not the Templars, Bruno says, but external forces around them. With European Christendom’s last strongholds in the Holy Land lost toward the end of the 13th century, the order found itself without a real mission. It was a military enterprise built for holy wars with no holy wars to fight. “We’re doing what the Templars should’ve done,” he says. “We’re changing.” In three years, he says, ULA is going to have a new rocket, a new engine, and a significantly more competitive position. It’s an immense job. As he puts it, deadpan before breaking into a grin, “I don’t want to be burned at the stake.”

On a hazy Friday in September, an executive-level status meeting on ULA’s planned Vulcan rocket convenes in Bruno’s office. The topic is materials, specifically whether to construct its first stage from machined aluminum or a more exotic, lightweight stainless steel skin. Ultimately, the same kind of machined aluminum already used on Atlas Vs and Delta IVs gets the nod. Stainless steel would carry greater design risks and require changes to the engine, and right now the engine is just about everything to ULA.

The engine is the novel, liquid-natural-gas-burning BE-4 being developed by Blue Origin. ULA chose to partner with Blue Origin more or less by default; it’s the only U.S.-made rocket engine that will likely be ready to fly by 2019, when ULA needs to start testing Vulcan. “Time to market is important to us,” says Mark Peller, ULA’s vice president of engineering and the lead on Vulcan’s development. “The competition continues to move forward.”

Right now, he adds, “our entire focus is trying to improve our competitiveness.” ULA will continue trimming its work force by a quarter—up to 875 people—through the end of next year. ULA will also scale back its portfolio, phasing out its Delta II rockets next year and its larger Delta IV in 2018. Three of five launchpads will close, leaving a single launch site on each coast.

These efforts are aimed at making Vulcan as cost-competitive as possible. “We’re already bidding lower prices today, and by the end of 2017 we’ll be offering sub-$100 million launch costs,” Bruno says. “Vulcan is the last piece.”

Both rocket and engine remain untested, raising huge question marks for a company whose strongest selling point is reliability. The BE-4 marks Blue Origin’s first foray into a rocket engine of this size and type. A lot could go wrong—a full-scale static fire test of the engine, slated for later this year, has already reportedly slipped into 2017—and ULA’s schedule leaves very little room for error. Bruno has saddled up, lance in hand—but the dragon still looms large.

A version of this article appears in the November 1, 2016 issue of Fortune.

Sign In

Get

Thank you for your interest in licensing Fortune content. Please find information on various licensing contacts below and choose the one that best suits your needs:

  • 1. To license Fortune articles, excerpts, or headlines for republication in various media (including books, eBooks, film, web, newsletters, newspapers, magazines and others), please email syndication@timeinc.com.
  • 2. To license a Fortune cover, order reprint or e-print copies of an article or cover, or license an accolade, please contact PARS International at www.timeincreprints.com.
  • 3. To license text only photocopies of Fortunearticles as print or digital handouts in academic settings, or in academic coursepacks, please contact the Copyright Clearance Center at www.copyright.com