When the pilotless, wing-shaped warplane lifted off a runway at California’s Edwards Air Force Base for the first time on the morning of April 27, it was like the resurrection of the dead. The Boeing Phantom Ray — one of the most advanced drones ever built — came close to never flying at all.
In late 2007, according to company insiders, U.S. military officials ordered Boeing to destroy an earlier version of the Phantom Ray, the X-45C. Exactly why the feds wanted the robotic aircraft dismantled has never been fully explained.
Boeing had just lost out to rival aerospace firm Northrop Grumman in a contest to develop a so-called “Unmanned Combat Air System” for the Navy, capable of taking off from, and landing on, aircraft carriers. That contest, known by its acronym N-UCAS — “N” for “Navy” — was actually the third time in five years Boeing had gone toe-to-toe with Northrop over a government contract to build killer drones, and the second time it had lost.
With each round of competition, Boeing had made enemies.
Even so, the kill order came as a shock to the Chicago-based company. Rare if not unprecedented in the world of military contracting, the command represented the climax of a nearly decade-long drama pitting a rotating field of corporations and government agencies against each other and, bizarrely, even against themselves — all in an effort to develop a controversial, but potentially revolutionary, pilotless jet fighter.
The UCAS development story has all the trappings of a paperback technothriller: secret technology, a brilliant military scientist, scheming businessmen, and the unseen-but-decisive hand of the military’s top brass.
And the story’s not over. The X-45C barely survived the government’s alleged assassination attempt. And after three years of clandestine development, a modified version of the flying-wing ‘bot leaped into the air that day in late April, an event depicted in the video above. The Boeing drone’s first flight opened a new chapter in the ongoing struggle to build a combat-ready, jet-powered robot warplane — and to convince the military to give the new unmanned aircraft a place on the front lines of aerial warfare.
What follows is the Phantom Ray’s secret history, reconstructed from news reports, interviews with government and corporate officials, leaked documents, and a treasure trove of information from Boeing insiders who spoke to Danger Room on condition of anonymity. Officials at Northrop largely declined to answer in-depth questions about their unmanned aircraft’s development.
This isn’t a complete retelling of the competition to build the combat drone. By virtue of its subject and sources, this portrays largely Boeing’s point of view over those of its rivals and customers. And Boeing played just one role, however prominent, in the continuing drama.
With traditional manned fighters growing more expensive — and consequently rarer — by the day, unmanned warplanes are rising to take their place. Boeing isn’t alone in testing pilotless jet fighters. Northrop Grumman, Lockheed, General Atomics, European firm EADS, British BAE Systems and Swedish plane-maker Saab are also working on killer drones. Each company’s UCAS surely has its own secret history.
The future of aerial warfare is more robotic than ever. Boeing’s decade-long struggle to launch the Phantom Ray, and the drone’s ultimate takeoff, is one reason why.
The X-45 and other UCAS can trace their roots to the first Gulf War. In January and February 1991, a U.S.-led air armada hammered Iraqi positions in occupied Kuwait. In the course of around 100,000 sorties, 42 coalition airplanes were lost to Iraqi air defenses, and 38 aviators died.
An Air Force officer named Mike Leahy was determined to make future aerial assaults safer for pilots — by removing the pilots from the most dangerous missions. Leahy’s ambition was bound to face opposition from the Air Force establishment, symbolized by the white linen scarf worn by World War II aviators, that was determined to keep men behind the yokes of America’s warplanes.
Leahy was an unlikely pioneer. In an Air Force dominated by fighter pilots with perfect eyesight, he was a glasses-wearing, ground-bound engineer — the opposite of a white-scarfer. Leahy started his Air Force career in 1980 in a laser laboratory. He eventually published 50 academic papers and earned four degrees, including a doctorate in engineering. He was, in short, a nerd.
And a revolutionary. In the middle of his career, Leahy’s concentration shifted toward robotics, and in the late 1990s he was temporarily assigned to Defense Advanced Research Projects Agency, the Pentagon’s fringe-science wing, to continue his efforts. There, Leahy led an alphabet’s soup of programs that guided the gradual evolution of combat drones from neat idea to deadly weapon. “The father of the X-45,” is how one Boeing insider described Leahy.
On April 16, 1998, the Air Force and Darpa, under Leahy’s guidance, awarded $4 million contracts to four companies: Boeing, Northrop, Lockheed and Raytheon. Each company had 10 months to come up with a preliminary UCAS design for the “post-2010″ time frame.
Boeing produced the best studies, and in April 1999 the plane-maker was awarded a contract to continue its killer-drone work. It was a so-called “cost-share” contract, with the government ponying up $131 million. Any additional cost, Boeing would have to cover itself, to the tune of at least $300 million over the first six years.
Northrop, meanwhile, beat out Boeing in a parallel contest launched by Darpa and the Navy to produce a killer drone that could take off of, and land on, aircraft carriers. In 2001, Northrop snagged government cash to build several of its X-47A drone prototypes; Boeing said it was looking at ways of “navalizing” the X-45, likely by strengthening the landing gear for hard carrier landings.
The first of two X-45As took off on its inaugural flight on May 22, 2002, reaching an altitude of 7,500 feet and a top speed of around 200 mph. It was a modest flight for an airplane, but “a significant jump” for a combat drone, to borrow one Air Force official’s description.
At a Darpa conference in Anaheim, California, in 2002, Leahy described his strategy for developing pilotless fighters in an Air Force still proudly wearing its figurative white scarf. He directed the drone designers to optimize their robots for destroying enemy air defenses — easily the most dangerous job in all of aerial warfare. “It is a mission that doesn’t directly threaten the white scarf crowd,” Leahy said, “but enables them to better perform their primary mission of air supremacy” — that is, dogfighting.
At that point in the UCAS’ development, Leahy aimed for Boeing to build a dozen or so test drones by 2007, wring them out in a series of tough exercises, then begin manufacturing combat-ready bots around the year 2010, at a unit price lower than the roughly $100 million a typical manned fighter would cost. It was an plan: It’s rare for American warplanes to go from blueprint to flight-line in fewer than 20 years, and even rarer for per-plane price to decrease from one generation of technology to the next.
Building the robot planes themselves was relatively easy. Much tougher was writing the software needed to fly the drones. “The operating system is the part that’s hardest to deal with,” Michael Francis, Leahy’s successor, said later. Ideally, killer drones would fly in a choreographed “swarm,” swooping down to overwhelm an enemy’s defenses. But swarm behavior required a fast-reacting blend of navigation, communication, targeting and formation-flying that had never been demonstrated before.
Leahy was aware of the difficulty of pulling off what he called “multi-vehicle, coordinated control,” even using the latest data-links, GPS, sensors and algorithms. But without it, the X-45 would never match human pilots, and would go nowhere. “Demonstration of that capability will culminate in a graduation exercise” for the Boeing drone, Leahy said. He hoped that would occur sometime in 2003.
But the Pentagon had other ideas. In April 2003, before Boeing and Darpa could complete the X-45′s final graduation, the military decided what was good enough for the Air Force should work for the Navy, too. Even in the flush years following 9/11, the idea of two combat drone programs seemed a little excessive for the Pentagon. The two UCAS programs were ordered to combine into one, competitive effort, known as “Joint-UCAS.”
Blending the two initiatives essentially overturned Boeing’s and Darpa’s carefully-laid plans for the X-45. Now Boeing would have to compete again with Northrop. And there was another catch — one that planted a ticking time-bomb inside the Boeing drone team, the J-UCAS program and, arguably, the Pentagon’s entire warplane plan stretching for decades. The military required that Boeing and Northrop jointly develop common drone-control software that would be compatible with the X-45 and the X-47, pictured above.
That seemingly innocuous requirement put Boeing in an awkward position. With unmanned aircraft like the Hunter and the high-flying Global Hawk, Northrop had a proven track record as a drone-maker. Boeing, in contrast, hadn’t built many robotic planes. Their advantage lay in the software, company insiders felt.
With at least a year’s head-start on Northrop, in 2003 Boeing was in possession of a mostly complete control software, while Northrop was not. Working together basically meant Boeing handing over to its biggest rival, for free, what Leahy had described as the most important part of the drone architecture — and, by, extension the foundation of the future’s unmanned air force.
The way the Chicago company handled that awkward edict made a huge splash in the U.S. aerospace industry. The ripples are still spreading.
At the time of the merger, Boeing believed it was on the way to achieving Leahy’s goal of debuting swarming, combat-capable drones around 2010. The key to this progress was the company’s Distributed Information-Centralized Decision mission-control software. “Dice,” as it’s known inside Boeing, is a software suite that allows human operators on the ground to feed, via radio, mission parameters to drones in the air: Go here, do this, attack that.
Dice’s first big test was already in the works when the Navy and Air Force killer-drone programs merged in 2003. On Aug. 1 the following year, the two X-45As rolled down the runway at Edwards Air Force Base in California. The two drones took to the air and performed a series of preplanned moves, “autonomous[ly] maneuvering to hold their relative positions,” according to the company press release. A full swarm, it wasn’t — but it was the “first ever multiple air vehicle control flight demonstration,” Boeing trumpeted.
Over the next year, Boeing steadily expanded the X-45A’s autonomy and formation-flying skill. An X-45 flying solo had already dropped bombs, back in March 2004. By 2005, Boeing was flying the two X-45s simultaneously with two simulated drones that existed only in Dice’s computer brain, and doing it “over the horizon” — that is, with the drones in California and the ground-based operator sitting at a console in Seattle.
On the drones’ 50th test flight in February 2005, they orbited a simulated battlefield, scanning for “enemy” activity below. Simulated surface-to-air radars flickered on and pretend missiles arced into the sky, all merely impulses inside Dice, following a digital script prepared by Boeing engineers. The drones executed pre-programmed tactics to swoop in and drop mock satellite-guided bombs. It was the long-delayed graduation exercise that Leahy — now promoted out of the UCAS program — had hoped for years earlier.
With growing confidence in its ‘bot design — and, more importantly, in Dice — Boeing began building two larger, more powerful X-45C versions of its killer drone. They would be faster, longer-ranged, fully radar-evading like an F-117 stealth fighter and fitted with probes for in-air refueling.
As the X-45, pictured above with program officials, moved from strength to strength, the X-47 appeared to lag behind. Northrop’s diamond-shaped drone flew for the first time in January 2004, two years after the X-45′s aerial debut. Northrop’s second-generation killer drone, the X-47B, wouldn’t appear until 2007.
But because of the government’s edict that the two drones share a common operating system, Boeing was expected to help Northrop catch up. “Darpa wanted us to give Northrop all our key products,” the Boeing source said. “We felt it was criminal, but the company knew the backlash [from refusing] would have killed us.”
The U.S. military was funding a big chunk of Boeing’s killer drone work. So the sharing edict may seem perfectly reasonable. But since the Chicago company had paid for most of Dice using company funds, it could argue that all the software was proprietary until the J-UCAS program identified a clear, specific need for Boeing to share. “This led to an unusual working relationship,” the source said. “We answered questions,” but if Boeing employees saw Northrop doing something wrong with regards to its own drone, they “couldn’t say anything.”
Northrop declined to comment on the company’s work on the common operating system.
With every bit of knowledge Boeing handed over, Northrop caught up. More and more, the only major differences between the two killer drones were in the airframes themselves, as their control software — based mostly on Boeing’s Dice — converged.
Though competing for the same contract according to the same requirements and with increasingly similar control systems, the X-45 and X-47 airframes could not have been more different. The X-47 originated with a Navy program; the X-45 was a response to an Air Force need. Each was optimized for its original customer.
So the X-45 was smaller, ostensibly more nimble and stealthier thanks to its thin wing and body. For long-range missions, the X-45 would rely on aerial refueling, rather than carrying lots of gas on its own. The X-47, by contrast, was built tougher to survive the brutal carrier landings. Since the Navy doesn’t have large aerial tankers of its own, to reach distant targets the X-47 had to have big fuel tanks. That increased the thickness of the Northrop drone’s wing and body, compromising its stealth.
In 2011, Navy Capt. Jaime Engdahl, the officer overseeing the X-47B, carefully described the drone as “LO-relevant.” “LO” stands for “low-observable,” or stealthy. Pressed for an explanation of the unwieldy term, Engdahl admitted that the X-47B was not actually radar-evading, per se. Rather, its design could accommodate stealthy enhancements in the future.
The X-45, by contrast, is an inherently stealthy design, Boeing officials insist — especially in its C model, pictured above. “I expect it will beat the others in that department, both heading into and away from a threat radar,” a company source said.
As long as both the Navy and Air Force were in the killer-drone business, jointly sponsoring the UCAS program, each company had reason to hope its design would win out when the two drones went head to head in a planned 2007 fly-off. As long as the two military branches were equal partners, neither bot had a clear advantage based on its origins. In principle, either could eventually be modified to satisfy — however imperfectly — the needs of the Navy and Air Force.
With the common software slowly coming together and no fewer than four war-bots buzzing around hitting test points, in 2005 Darpa decided to hand over the J-UCAS program to full Air Force and Navy control, in order to speed along the process of bringing the robots into front-line service. The transfer had unintended consequences, however, that nearly killed off the program’s original drone.
Just a few months after Darpa bowed out of J-UCAS, the Air Force did, too. After investing a decade and several billion dollars of government money, the flying branch had changed its mind about killer drones — and just as the X-45 was proving itself ready for combat and a second generation of the drones was taking shape. J-UCAS would survive in a different form, as a Navy-only program renamed N-UCAS.
J-UCAS’ abrupt ending came as a shock to Boeing, in particular. Northrop clearly had a leg up in a Navy-only competition. Boeing had reason to fear J-UCAS’ collapse would start a domino effect that could lock the firm out of any major killer drone business for the foreseeable future.
So in March 2006, Dave Koopersmith, then Boeing’s X-45 program manager, and his boss Darryl Davis met with military officials to discuss J-UCAS’ collapse — and figure out if the company still had a future in killer drones.
The men made a powerful team. Koopersmith is tall and lean. Easygoing but inscrutable, he’s earned a reputation for technical savvy, and for being an excellent manager of engineers. Koopersmith knew his killer drones, and their makers, inside and out.
Davis is, in many ways, Koopersmith’s opposite. Small in stature, Davis is a politician and salesman more than an engineer — the kind of guy you can find forging strategic partnerships through a well-played game of golf.
The two were ready to give a pitch for the X-45, covering all possible bases, from the technical to the political. Instead, they just listened as the Air Force explained its rationale for abandoning the killer drone. To hear the Boeing employees tell it, the Air Force killed off J-UCAS to protect its new, ultra-pricey manned fighters, the F-22 and the F-35 Joint Strike Fighter, or JSF.
“The reason that was given was that we were expected to be too good in key areas and that we would have caused disruption to the efforts to ‘keep F-22 but moreover JSF sold,’” the Boeing employee said. “If we had flown and things like survivability had been assessed and Congress had gotten a hold of the data, JSF would have been in trouble.”
In other words, Leahy’s strategy had backfired. The former combat drone champion had hoped that the unmanned aircraft’s improving performance would overcome any opposition by the “white scarf crowd,” determined to preserve a human being’s place inside U.S. warplanes. Instead, the Boeing drone spooked the old guard with its advanced capabilities, provoking what seemed like an emotional, irrational backlash — one that shattered the Air Force-Navy alliance and doomed the Air Force-optimized drone.
Boeing made an effort to keep the X-45 viable in the Navy-run N-UCAS program, but the company knew the X-47 was the assumed winner. To beat the Northrop drone, Boeing would need to demonstrate superior technical performance and offer a lower price. “The challenger in a title fight rarely wins by decision, they must win via TKO or knock-out,” Koopersmith explained in a letter to his UCAS team.
The Navy required the winning company to launch and land its drone on an aircraft carrier no later than 2013. It’s a harder task than it sounds. Carrier decks are small and crowded by airfield standards, and constantly moving. And the airspace around a flattop teems with helicopters, fighters and resupply planes. Threading a pilot-less aircraft through this aerial tangle represents “a big challenge,” Engdahl said, as does maneuvering the bot around the carrier deck without running into anyone or anything. “Unmanned operations on the carrier: That is the big shift.”
The company prepared what it viewed as a thorough and realistic bid based on what it knew about the difficulties of perfecting drone software. The cost, according to Boeing: $1.2 billion over five years.
The answer came back from the Pentagon on August 3, 2007. It was a gut punch. Northrop had won the UCAS-N contract with a $650-million bid — just over half the price Boeing believed was realistic.
The Boeing engineers weren’t shocked that they lost, but they were shocked how they lost. How could Northrop, with what they strongly believed was inferior software, possibly pull off a robotic carrier landing cheaper than Boeing? The X-45 team was already hurt and suspicious when the Navy allegedly made their final, shocking demand. According to a company insider, the Navy ordered the company to destroy the two X-45Cs then under construction in St. Louis.
In late 2007, Koopersmith and Davis, along with corporate lawyers, went to the Pentagon, looking for an explanation.
“It got very heated,” according to a company source. When asked why the Navy had ordered the destruction of the two Boeing X-45C systems, the answer was that they didn’t “meet the mission requirements or otherwise have usefulness.” Boeing then asked: If that was true, was [Northrop's] X-47 system developed in the same period going to be destroyed, too? A military lawyer told the Navy official not to answer.
It got worse. Just hours after the meeting, Northrop practically admitted that it had under-bid the contract. Rick Ludwig, Northrop Grumman’s director of business development, told Aviation Week that the company was still negotiating the “funding profiles.” After adding aerial refueling and other capabilities the X-45 already possessed, the cost of the X-47 carrier demonstration could, Ludwig said, rise to $1.2 billion. Exactly the price Boeing had proposed.
At Boeing, there were all kinds of threats about lawsuits in the days afterwards. But the threats never materialized. Northrop, for its part, declined to comment further on the bidding controversy.
After that, Boeing didn’t try to fight the N-UCAS award, despite the huge ramifications for an aerospace company struggling to stay in the warplane design game. To many industry insiders, it appeared Boeing had given up on killer bots, essentially surrendering the future combat drone market.
The X-45As wound up in museums. Ground equipment was placed in storage. The X-45 team disbanded and its members moved to other Boeing programs. For two years following the N-UCAS drama, not a word was heard from Boeing regarding its once record-setting killer drone.
Then in mid-2008, Boeing quietly rebuilt the X-45 team and, in May 2009, surprised everyone by announcing the UCAS’ resurrection, in the form of the bigger, smarter, more powerful X-45C, now called Phantom Ray.
The Navy never followed through on the alleged order to destroy the X-45Cs. In St. Louis, engineers were putting finishing touches on two of the enlarged killer drones. A special Boeing 747, usually used to transport the Space Shuttle, carried the first Phantom Ray on its back from St. Louis to California. First flight was slated for 2010, but some last-minute modifications delayed that to April 27 of this year.
The Dice control system was mostly unchanged. It was revolutionary in 2005, and despite Northrop’s recent advancements, remains some of the best drone-control software in the world.
In an echo of Boeing’s very first UCAS effort in the late 1990s, the revitalized killer drone was entirely company-funded, and not exclusively associated with a single government requirement. Instead of tying itself to the Air Force, Navy or Darpa for development and risk getting burned again, Boeing would refine the Phantom Ray on its own terms and at its own pace.
The approach carried a bit of a stigma; in the Pentagon’s weapons-development community, anything that’s not a military-funded “program of record” runs the risk of being seen as an ugly stepsister. But there were advantages, too. “Since we’re not a government program of record, we’re able to do some things in a rapid fashion,” Davis said. He added that the Phantom Ray would probably compete for the Navy’s follow-on program to N-UCAS and maybe the robot component of the Air Force’s next-generation bomber program.
Koopersmith had unknowingly predicted the Phantom Ray’s resurrection. “You have laid the foundation for the future of Boeing with all of the technology you developed and the aviation firsts you accomplished,” Koopersmith wrote to the drone team in 2007.
The X-45 drama has also laid the foundation of a new approach to warplane development — and to aerial warfare. Stung by the Boeing’s and Northrop’s UCAS spats and other weapons-buying disasters, the Pentagon wants more companies to pay for their own prototypes, rather than relying on the military bureaucracy to lead and fund every effort. That could have the effect of producing better weapons, faster.
With Boeing back in the UCAS game on its own terms – and with Northrop and General Atomics testing their next-generation, jet-powered drones – unmanned aircraft could be in a position to gradually gain the support that Leahy envisioned all those years ago. Drones may finally win a position in the ranks of front-line U.S. warplanes. Air combat may never be the same.
Photos, videos: Boeing, Northrop Grumman, Air Force, Darpa