The Imagineers of War

by Sharon Weinberger

In the early 1980s, Allen Atkins, now a DARPA official, was in Israel to give a talk at a conference and to brief the Israelis on some of the agency’s unclassified projects, including its unmanned aerial vehicles. As Atkins recalled it, he was sitting in the hotel’s lobby bar when a “little portly guy” came up to him and stuck out his hand. “I’m Abe Karem,” the man announced. And without much more of an introduction, Karem plopped down next to him and began to describe his concept for an unmanned aircraft that could stay aloft for days at a time.

  In the United States, drones on the battlefield had not progressed much since the Vietnam War, when DARPA sent the QH-50 to hunt Vietcong. DARPA had continued to sponsor work on small-scale drone projects, but without much interest from the military. The air force was run by pilots, who did not want to be put out of a job, and the army and the navy, though slightly more interested in drones, could not quite imagine how they should be used. For example, DARPA had handed the army a tactical drone called Prairie, which was powered by a lawn mower engine. The army eventually called its version the MQM-105 Aquila, which was supposed to be launched by a catapult and recovered in a net. Rather than keeping it simple, the army came up with more things the drone should be able to do. Cost estimates for Aquila grew to $2 billion before the army canceled it. The drone was never used in combat.

  Compared with the United States, the Israelis had embraced unmanned aircraft, particularly during the 1973 Yom Kippur War, when the Israel Defense Forces used drones to draw out anti-aircraft artillery over the Golan Heights and to perform reconnaissance. Though the Israelis might have employed drones more readily than their American counterparts, the Israeli military did not have the money to fund ambitious new aviation projects, at least compared with the Pentagon. Frustrated, Karem, an Iraqi-born Jew who had worked as an aircraft designer in Israel, moved to the United States. He set up shop in California, eventually working from a garage attached to his home. There, Karem started work on unmanned aerial vehicles that could stay aloft for long periods of time, and that was the idea that Karem presented to Atkins that night in the hotel bar. “Karem wasn’t really pushing any particular application,” Atkins said, he was just interested in building drones.

  Karem’s proposal intrigued Atkins. Up to that point, one of the biggest problems with military drones was simply that they crashed too often. The QH-50, the unmanned helicopter that DARPA had sent to Vietnam, had a reputation for falling out of the sky; the Aquila did as well. Karem was pitching reliability, so Atkins decided it was worth at least having someone at DARPA check out the Israeli inventor’s idea. Atkins assigned Robert Williams, a program manager who had a reputation as a brilliant aerodynamicist and a keen eye for spotting talent. “You are going to have to talk to him to find out if he has anything,” Atkins told Williams.

  There was another reason why Atkins thought DARPA might be interested, though he could not tell Karem, because it was classified. In 1980, DARPA had launched the top secret program Teal Rain, a code name for a series of unmanned aerial vehicles that would replace spy planes like the U-2 and the SR-71. Some of the projects under Teal Rain were classified, while others were conducted in the open. Even thirty years later, DARPA officials decline to speak about many aspects of Teal Rain, citing its secrecy.

  As it turns out, Karem and Williams, both known for single-minded pursuits, meshed well. Karem was a rogue aircraft designer whose vision left little room for tact. “Gentlemen, everything I see in this room is nonsensical,” he said at a meeting with a large defense firm. Williams was an engineer turned government bureaucrat with money to spend and whose love of visionaries could sometimes cloud his better judgment. As an initial step, Williams funded flight tests of the Albatross, a two-hundred-pound drone that Karem built in his garage. It flew an astounding fifty-six hours. When that design proved successful, DARPA then paid Karem to build Amber, an unmanned aircraft funded under the Teal Rain program. While Amber, which eventually flew 650 hours without a single crash, was ostensibly run in partnership with the navy, the real interest, according to Atkins, was from the CIA. “The navy was in it primarily for cover,” he said.

  By 1990, DARPA finished its work on Amber. As a research agency, it could only develop prototypes; it was up to the military to buy production aircraft. Lacking new orders, Karem was pushed into bankruptcy, forced to sell what was left of his company to General Atomics Aeronautical Systems. Then the CIA bought the Gnat, a derivative of the Amber that Karem built to sell abroad, and used it for surveillance during the war in Bosnia. Less than ten years later, the agency bought another derivative of Karem’s work from General Atomics Aeronautical Systems, called the Predator. By that time Karem was not even involved in the company, though his work had led to a weapon that was about to change how the United States conducted war. The Predator was sent to Afghanistan in the days after 9/11, armed with Hellfire missiles and used to kill “high value targets,” much as DARPA had wanted to do with the QH-50 back in the Vietnam War. This time, the plan worked, and the Predator ushered in an era of remote-control killing. For better or worse, the Predator, as Richard Whittle wrote in his history of the unmanned aircraft, “changed the world.” So, too, had DARPA.

  —

  In the 1980s, DARPA was expanding rapidly into secret aviation projects. The success of Have Blue, the first stealth prototype, prompted DARPA to start another radar-evading aircraft program, called Tacit Blue, a prototype plane based on Northrop Grumman’s design, which had originally lost out to Lockheed’s Skunk Works. Tacit Blue was a bizarre-looking aircraft that DARPA pursued, in part, to ensure that at least two companies would have the ability to build stealth aircraft. “Looking at it from the side, it looks just like a whale, with the fins,” said Atkins, who was recruited to DARPA to manage the program because of his experience on Have Blue.

  Tacit Blue’s odd shape earned it the nickname the Whale, and many of those who worked on it wore gold tie tacks that had tiny whales on them—a way of acknowledging that they belonged to an elite club. Some of the tie tacks even included tiny diamonds to represent the aircraft’s side-looking radar. As a spy plane, Tacit Blue was being used to test whether a narrow-band radar would be invisible outside its transmission range. “Most people didn’t know about the radar,” Atkins said. “They thought it was just building an airplane.” Like Have Blue before it, Tacit Blue was kept completely in the “black” and flown at Area 51.

  As DARPA’s black aircraft programs boomed in the 1980s, the agency would often announce a research program in aeronautics as a cover for building a secret military prototype aircraft. This enabled DARPA to award contracts and buy needed equipment without raising suspicion. Atkins described one example of what he called a “White World” program, meaning unclassified, that was run jointly with NASA. Yet the technology was also being looked at for secret military applications. “We built some full-scale models to see how we would militarize it,” said Atkins. The “black” project was a stealth rotorcraft.

  The cover project was called the Rotor Systems Research Aircraft/ X-Wing, or RSRA, a joint DARPA-NASA program that funded Sikorsky to design a hybrid of a helicopter and a fixed-wing aircraft. RSRA was a real program, but it was also a cover for one of DARPA’s more significant attempts to develop stealth helicopters. The “black” program involved taking “the rotor head off of the RSRA and putting it onto a stealth vehicle,” Atkins said. A helicopter’s rotor blades typically produce a Doppler shift. Those shifts are difficult to mask from radar, but it could be done, as DARPA learned with the X-Wing.

  One after another, “X-planes,” or prototype aircraft, seemed to spill out of Atkins’s office. One, called a CycloCrane, looked a bit like La Minerve, the nineteenth-century fantasy blimp dreamed up by the Belgian physicist and magician Étienne-Gaspard Robert. The CycloCrane was a hybrid lighter-than-air vehicle that incorporated elements of helicopter control. It looked as if someone had taken propellers off a beanie hat and placed them strategically around the body of the airship
. Highly maneuverable and with the ability to haul large amounts of cargo, the CycloCrane was proposed by DARPA to the navy as a possible way of off-loading ships in areas that did not have developed ports. The navy was not interested in the strange-looking airship. “If you could ever get beyond the giggle factor, this would be a great aircraft,” Atkins recalled one admiral telling him.

  Another “doomed by looks” project was the X-29 forward-swept-wing aircraft; the wings look as if they were put on backward. The aircraft, which in theory would be highly maneuverable, never went beyond DARPA’s investment, in part because the air force had no interest in it. “No, that’s too ugly an aircraft,” one four-star general told Atkins.

  For Atkins, those were exciting days at DARPA as money flowed into the aviation programs, initially run out of the Tactical Technology Office. The programs grew so quickly that the office was “swallowing the agency’s budget,” said James Tegnelia, the deputy director of DARPA at the time. So, DARPA created a separate Aerospace Technology Office, headed by Atkins. Splitting aviation off into a new division did not necessarily solve the problem, because the new office eventually swelled to $1.5 billion, or more than half of DARPA’s budget. “I had one program that lost”—Atkins then corrected himself—“spent $600 million, and it failed, but we knew where it failed at,” he said. That program, like many others, remains a secret. Those aircraft might not have been successes, because they never led to something used by the military, but they were the types of high-risk concepts that were encouraged at the time. “You don’t go after a minimum change or an incremental change,” Atkins said. “You go after something that’s going to force people to think outside of the box.”

  Secrecy and unbridled ambition were the hallmarks of aviation programs in the 1980s. DARPA had long managed classified programs, but secrecy seemed to be enveloping much of the agency as the black aerospace research burgeoned. As the head of the Aerospace Technology Office, Atkins enjoyed the cloak-and-dagger nature of those programs, even when they sometimes brought DARPA into conflict with the military services. DARPA was encouraged to charge forward, even if military officials were opposed.

  One time, Atkins needed to get James Ambrose, the undersecretary of the army, to sign a classified agreement with DARPA on a major black aircraft program. Ambrose, a master of bureaucratic wrangling, was trying to avoid signing the agreement, even though the army had agreed to participate. Unperturbed, Atkins got Ambrose’s travel schedule from his aides and hatched a plan to track him down at LaGuardia Airport. The only hitch was that the agreement was for a top secret code word program, so the documents had to travel with two individuals with appropriate security clearances. Atkins ended up taking his wife, Natalie, a DARPA secretary who was cleared onto the program. At LaGuardia, the husband-and-wife tag team ambushed Ambrose as he got off his flight.

  Cornered, Ambrose agreed to sit down at a dimly lit airport restaurant. “It was Ambrose facing me, and [my wife] and I were sitting so we could look out to see what was going on, and we had his aides fix up a perimeter around it,” Atkins said. “He’s sitting there with a candle, holding it over the document, trying to read it. Finally he turns and signs it, and I get back up, put it in an envelope, seal it again.”

  Atkins recalled another run-in with Ambrose at a meeting of the Defense Resources Board. Ambrose was known as a fierce protector of the army budget, and DARPA programs were a direct threat to the army’s own weapons development. Cooper, the DARPA director, believed Ambrose was blocking a top secret DARPA program, which the army was supposed to be supporting (Atkins declined to name the program).

  Secretary of Defense Caspar Weinberger had a habit in meetings of closing his eyes, leading some to believe he was asleep or not listening. Yet at the end of protracted discussions, and even heated arguments, he would open his eyes and make a simple pronouncement. As Cooper argued his point at the meeting, he stood up, all six and a half feet of him, towering over everyone in the room, including the diminutive Weinberger, who sat back silently.

  “There are over half a dozen ways you can undermine a program, and this wall-eyed son of a bitch has done all of them!” Cooper shouted, bending over Ambrose and pointing a finger in the army official’s face.

  “Bob, why don’t you tell us what you really think?” Weinberger said.

  Weinberger then turned to Ambrose and asked, “Is any of that true, Jim?”

  Ambrose started to defend himself, explaining what he had done to protect the army’s budget, which only bolstered Cooper’s claims.

  “Well, Jim, we’re going to do the program, and the Army’s going to do it with DARPA,” Weinberger said.

  To this day, Atkins will not say precisely which project was the subject of the contentious meeting, but three decades later an army stealth helicopter flew navy SEALs into Pakistan on a mission to kill Osama bin Laden. By that point, the descendant of another DARPA aircraft from Atkins’s office—Karem’s drone—had been hunting and killing suspected terrorists for nearly a decade.

  —

  As the 1980s progressed, Reagan’s techno-optimism permeated the Pentagon and DARPA. Projects that might once have been discarded as pipe dreams suddenly seemed plausible, even to Cooper, who had been a missile defense skeptic. In April 1983, just a few weeks after Reagan’s Star Wars announcement, a member of the duPont family came to talk to Cooper about a plan for a hypersonic space plane. It was fortuitous timing. A space-based missile shield would require putting satellites, weapons, and other technologies up in orbit quickly and cheaply, something that an aerospace plane could do, at least in theory. Rather than the months it takes to plan a rocket launch, the aerospace plane would just zip up to orbit and then return to earth, landing on a runway. “This is something DARPA ought to do,” Cooper enthused.

  Tony duPont was not involved in the family’s eponymous business, E. I. du Pont de Nemours and Company, better known simply as DuPont, but he shared an entrepreneurial spirit. He was a former Pan Am pilot and then later an aerospace engineer who had worked for more than a decade at Douglas Aircraft. There, he specialized in missile and space systems, focusing on how to get those vehicles to reenter earth’s atmosphere without burning up. In the 1970s, he struck out on his own, founding duPont Aerospace. He had moderate success working with NASA on a hypersonic engine but harbored grand ambitions of building an entirely new type of aircraft.

  The route duPont took to DARPA went through Tony Tether, the head of the Strategic Technology Office and a future director of the agency. Tether had been attending government meetings on trans-atmospheric vehicles—vehicles that combine attributes of aircraft and spacecraft—when he met duPont. Some in the aerospace community regarded duPont as a snake oil salesman who oversold his ideas, but the soft-spoken, earnest-sounding engineer had a way of winning over even ardent skeptics. Tether, a science fiction fan, took an immediate liking to duPont and his ambitious ideas and sent him to Cooper, who in turn assigned the space plane to Robert Williams, the same DARPA program manager who had championed Abe Karem, the maverick drone designer. Cooper described Williams as an “imaginative fellow” and thus the “right guy,” even if hypersonics was not his area of expertise. Cooper would later regret that judgment.

  Space planes were an ambition that dated back to the very origins of DARPA, but there was reason to be skeptical of duPont’s claim that he could design such a vehicle, which would need to travel at hypersonic speeds to reach orbit. Hypersonic vehicles—missiles or aircraft that can travel many times the speed of sound—have long been a dream of aerospace engineers. A hypersonic aircraft would make travel from the United States to Europe or Asia more like a brief train ride than a daylong journey. A hypersonic missile could strike an enemy halfway across the planet in just over an hour. And a hypersonic space plane, like what duPont was proposing, could lift things—people, satellites, or weapons—into orbit quickly and cheaply.

  One of the keys to building the space plane was a supersonic combustion ramjet, kno
wn as a scramjet. A conventional rocket carries its own oxidizer, which for space vehicles, like the U.S. Space Shuttle, means taking off with a massive tank filled with liquid oxygen and liquid hydrogen. A scramjet, on the other hand, takes in air from the atmosphere. The difficulty is that a scramjet only works at high speeds, around Mach 6. Even then, keeping the engines going is “not unlike keeping a candle lit in a hurricane,” as one writer described it.

  But duPont believed he had a design that could do it. His idea was to build a space plane powered almost entirely by a hybrid scramjet surrounded by ejectors channeling rocket exhaust. Rather than a big external booster, the duPont engine would incorporate a small rocket to power the aircraft until it reached high enough speeds that the scramjet could kick in, boosting the plane into orbit. It was a novel concept but also incredibly complex.

  One of the reasons that a space plane had not been pursued for several decades—along with the high price and complexity—was that it was not necessarily clear that such an exotic technology was needed. With Star Wars ascendant, DARPA now had a motivation to support duPont’s idea. The Pentagon was geared toward space weapons for the foreseeable future, and duPont’s aerospace plane could help put them in orbit. The Defense Department also had another highly classified mission for the space plane.

  “Can you do this mission?” Williams asked duPont.

  “I’ll study it,” duPont told Williams.

  “It was exciting; someone was finally interested,” duPont recounted. He stayed up all night, running the numbers, based on the hypersonic work he had done for NASA, and extrapolating out from the models he had built to determine whether ramjets could run to Mach 25, taking the hypersonic plane to orbit. A few days later, he called Williams back. “You can do it,” he told him.

  At that point, duPont was awarded a study contract for just $30,000 to draw up a theoretical design of an aerospace plane that could take off from a runway and then accelerate up to Mach 25. The study was to design the smallest plane possible that could reach polar orbit and then return to earth. The magic number was a fifty-thousand-pound plane with a twenty-five-hundred-pound payload, and that was the design that duPont submitted. DuPont delivered the results of his study at 6:00 p.m. on September 30, 1983, just as the fiscal year ended.