The memo said nothing about how Rutherford was to navigate the bureaucratic maze so he could meet that deadline, but a little more than a week after Deutch signed the UAV memo yet another star came into alignment. Larry Lynn, an expert in radar and surveillance technology who had been deputy director of DARPA from 1981 to 1985, arrived as Deutch’s deputy on July 21—personally recruited by Deutch to help prune some of the bureaucratic bramble out of defense procurement. Within weeks, Lynn came up with an idea. He and Deutch would create a new category of projects: instead of developing equipment from scratch, the Pentagon would buy “mature” technology—a weapon or a piece of equipment that was ready or near ready to use—in small batches and let military users try it out. If the users liked the product, more could be bought under normal procurement rules. If not, little time or money would be lost. These Advanced Concept Technology Demonstrations, as Lynn called them, would essentially allow the military to take new products for a test drive. Lynn and Deutch decided the new endurance UAV would be the first such project.
On November 17, Lynn signed a memo much like Deutch’s of July 12, except that Lynn’s directed Rutherford’s office to award a contract for an endurance UAV “within 40 days after money is appropriated.” Money was appropriated a few days later, and not by coincidence. Rutherford and others working on the drone project had briefed key congressional aides even before Deutch officially created the project. They also arranged for Deutch, Joint Staff intelligence director Cramer, and other senior officers to meet with key members of Congress about the new UAV.
Perhaps no meeting proved more important than the one Jayjock arranged for Cramer with Representative Jerry Lewis, a conservative Republican from Southern California with big hair and a toothy smile who was serving his eighth two-year term. Lewis was a power in the House, especially on defense spending. Edwards Air Force Base and Fort Irwin, the Army’s largest training center, were located in his district, and as a senior member of the House Appropriations Committee and its powerful Defense Subcommittee Lewis had a lot to say about which Pentagon programs got funded and at what levels. His district also happened to include El Mirage, and Lewis was delighted to hear that General Atomics was providing a UAV to the CIA and might build one for the Pentagon, too. Soon one of his top aides, Letitia White, was hearing from Tom Cassidy almost as often as Jayjock was. That fall, as House and Senate committees wrote the next year’s defense bills, Lewis made sure twenty million dollars was added for an endurance UAV.
Five days before Congress finished its work on the defense bill, Rutherford’s office sent “draft solicitations” to contractors asking for bids on the new drone. General Atomics and three other contractor teams responded, and on November 29 a competition officially began. Exactly forty days later, on January 7, 1994, the UAV JPO awarded the contract to General Atomics. Within six months, the company would have to demonstrate a UAV that met the requirements in Deutch’s memo. Within thirty months, it would have to deliver ten drones and three ground control stations for use in military exercises or actual deployments. General Atomics agreed to do all that for the relatively modest sum of $31.7 million.
A month after the award, Rutherford’s commanding officer got a seething letter from an executive of TRW, a major defense contractor, which had teamed with Israel Aircraft Industries to bid on the contract. TRW Vice President Robert J. Kohler charged that the award to General Atomics was “of questionable legality.” The “40 day rush,” he complained, gave contractors too little time to form good teams and created “the perception that the government really wanted to go sole source to General Atomics.” Holding the competition “over the Christmas holiday made it even tougher” and “gave a huge advantage to General Atomics,” he wrote. Kohler conceded that TRW and IAI had bid sixty million dollars more than the winner—three times as much—but he confidently declared that “this job cannot be done for the contract value awarded to General Atomics.”
Rear Admiral George Wagner, Rutherford’s commander, replied by letter that every point in Kohler’s letter was “incorrect,” and he offered to “debrief” the TRW-IAI team to explain why. But after Wagner’s reply, Rutherford never heard another word about Kohler’s protest. Years later, Rutherford insisted that no one directed or induced him to award the contract to General Atomics. His office, he said, simply picked the best design for what was by far the cheapest price.
* * *
Not long after Neal Blue’s company won the Predator contract, a small team of CIA and General Atomics personnel began flying a Gnat 750 over Bosnia. They launched the drone from a military air base in western Albania called Gjader, located on the inland side of some foothills on the coast of the Adriatic Sea. The team’s footprint was small; members lived on a military-style patrol boat docked at a nearby port. Line-of-sight communication to the skies above Sarajevo, roughly 140 miles distant, was possible, but a nearby range of mountains required the CIA team to relay its ground station’s signals to the Gnat 750 through a manned aircraft, a small, quiet “motor glider” known as a Schweizer RG-8. Operational flights were limited to a mere two hours by the Schweizer, whose crew of two needed six of the craft’s eight hours of endurance to get to and from an orbit area over the Adriatic Sea.
Despite this limitation, the Gnats gave U.S. forces exactly the sort of intelligence President Clinton wanted. Flying at six thousand feet or lower, but apparently undetected, the Gnat 750 carried sensors that allowed its operators to distinguish real artillery from decoy, find surface-to-air missile sites, and spot tanks and gun movements. The experiment wasn’t entirely successful, however: the video relayed by the Gnats was often degraded by interference at Gjader from unshielded power cables and fluorescent lights, and bad weather and problems with data links—the signals sending video and other imagery back to the ground station—also caused problems. After a couple of months, the CIA halted the flights but, overall, the operation was considered worthwhile. Later in the year, the agency’s two Gnat 750s flew again, from an island off the coast of Croatia, and Aviation Week reported that the CIA might buy three of the six Gnat 750s that Turkey had ordered and outfit two of them to serve the relay role played by the manned Schweizer. To those involved in the drone project, though, it was clear that the relay system was no better than a stop-gap solution. Ultimately, the drone’s designers would have to come up with something better.
Meager though the results of the missions were, Woolsey would later recall with great satisfaction seeing Gnat 750 video of pedestrians crossing a bridge in Mostar, Bosnia. He was proud of the fact that the CIA’s modest program helped spur the Pentagon to get its own endurance UAV—one that would turn the history of drones upside down.
* * *
Contract in hand, the General Atomics team had just six months to build a prototype of the new drone, and redesigning the Gnat 750 to carry the satellite dish the Pentagon wanted, plus up to 500 pounds of payload instead of 130, would require a superhuman effort. To add the satellite dish, Karem gave the Gnat a serious nose job, adding a large upward bulge to house the round antenna, whose diameter was thirty inches. To carry the additional weight of the satellite antenna and a synthetic aperture radar, the drone had to grow in every dimension. The wing on this new UAV would be 1,100 millimeters at the root, not 750; its fuselage would be 9 feet longer than the Gnat’s 17.5 feet; and its wingspan would be more than a third longer than the Gnat’s wingtip-to-wingtip length of 35 feet. All this would require fabricating a lot of new parts, engineering and testing how those inside the fuselage fit together, and writing new flight control and other software. Because it would be so different, the plane clearly needed a new name.
Rutherford and others in government had been calling the aircraft Tier II because Deutch and Lynn had decided the Pentagon should build UAVs for different altitudes, starting with a low-altitude Tier I and going up to a high-altitude Tier III. That, at least, was the official story: in fact, Tier I was a cover for the CIA’s Gnat 750 purchase, which was s
ecret until Aviation Week revealed it in September 1993. Two months later, the magazine reported that one of the CIA’s Gnat 750s had crashed near El Mirage, brought down by an apparent software malfunction, which forced the agency to lease a third Gnat 750 from General Atomics. As the engineers at General Atomics were putting together the first prototype Tier II for the Pentagon, Cassidy told Rutherford he had a name for the new UAV. They would call it, Cassidy declared, the Predator.
Rutherford cringed. Tier II was just a surveillance and reconnaissance drone, an eye in the sky. “Predator” sounded like a weapon. Nobody had suggested arming the new drone, though some in Congress were talking about equipping it with a laser designator so the UAV could guide bombs and missiles dropped by manned aircraft to targets. If Tier II began looking like a weapon, Rutherford feared the Air Force might try to kill the project. The fighter pilots who ran that service seemed instinctively hostile to the idea of unmanned combat planes. But when Rutherford tried to resist, Cassidy kept repeating “Predator” over and over, and finally it stuck.
No one ever mentioned that General Atomics had used the name before.
* * *
As aircraft go, the new Predator’s name was the sexiest thing about it. Constructed of graphite epoxy composites and lighter than an economy car, it was powered by a four-cylinder Rotax 912 piston engine, an Austrian motor used in ultralight sport aircraft that buzzed like a big mosquito. The drone’s top speed was just over a hundred miles an hour, and its cruising speed in a windless sky would be somewhere between eighty and ninety—about the velocity of a professional baseball pitcher’s changeup. Like its Karem-designed forebears—the Albatross, the Amber, and the Gnat—the new Predator’s appearance resembled a glider more than a powered aircraft. Its thin wings stretched almost forty-nine feet from tip to tip—nearly twice the length of its fuselage. The fuselage, while much longer than the Gnat 750’s, was eight feet shorter than a Piper Cub’s. Those wispy wings would help the Predator stay airborne more than a day and a night without guzzling fuel.
Another Karem feature familiar from the Albatross, the Amber, and the Gnat was the Predator’s tail. It sported two rectangular stabilizers jutting down in an inverted V—a configuration that added aerodynamic stability but also, in a rough landing, allowed the stabilizers to serve as skids, preventing the drone’s propeller blades from shattering on the runway and slinging dangerous debris across an airfield. Perhaps the most striking aspect of the reborn Predator, though, was the bulge Karem had given it forward of the wings. That aerodynamically sculpted hump, situated about where a manned aircraft’s cockpit might be, would hold the satellite dish. Directly beneath that hump and protruding from the bottom of the fuselage was another prominent feature: a “chin turret,” holding daylight and infrared video cameras.
The ground control station, where operators would sit to fly the new Predator and aim its sensors, would be housed in the same boxy brand of three-axle trailer NASCAR teams used to haul stock cars from race to race. The GCS, as insiders called it, was thirty-six feet long, eight feet wide, and eight feet tall. At one end of the interior was a console holding an array of computer screens and controls. There, one pilot and one “sensor operator” would sit in faux leather chairs, manipulating control sticks like those of a conventional airplane to fly the drone and aim its sensors. Their computer screens would display flight information such as altitude, airspeed, and other vital data. Two other screens would show video from the Predator’s cameras, and the console would hold keyboards and communications gear, including a radio the pilot could use to talk to air traffic control towers.
On July 3, 1994, four days before the contract’s six-month deadline for a demonstration, program manager Rutherford traveled to El Mirage to witness the Predator’s first flight, an event planned for just after sunrise, when the desert winds would be calmest. The dress code that summer Sunday morning was casual—Frank Pace was in tennis shorts—and the mood optimistic but nervous. To meet their extraordinarily tight deadline, the members of the General Atomics team had put in twelve to fourteen hours a day, six days a week, since Christmas.
Tim Just, a champion model aircraft pilot Karem had hired at Leading Systems to fly the Amber, would operate the Predator on its maiden flight. At about 6:30 a.m., Rutherford, Cassidy, Pace, and other observers watched the demonstration from a spot near where the CIA’s Twetten had posed with Karem fifteen months earlier, roughly two hundred feet from the El Mirage runway and a third of the way down its three-thousand-foot length. A technician turned a switch on the drone’s fuselage, then darted to safety as the Rotax engine grumbled to life and the Predator’s pusher propeller began buzzing. Pace, a born-again Christian, whispered a hurried prayer.
Soon pilot Just added power, and the Predator began rolling down the runway. A moment later, the drone eased into the air, nose tilted upward about five degrees. The takeoff was going beautifully—until the Predator’s engine choked and sputtered just as it passed the audience. Pace nearly choked himself, panic replacing the tingle of nervous excitement he’d enjoyed earlier as he waited for the flight to start.
Inside the GCS, Just saw the engine’s revolutions per minute plunge on his control panel and made a snap decision. He turned the ignition off and then, as gently as he could, maneuvered the Predator’s broad nose downward. The sleek, long-winged little bird wanted to keep flying, but Just set the drone back down on its spindly tricycle landing gear as quickly as he could. Then he desperately applied the brakes, nearly standing on his rudder pedals as the Predator rapidly rolled down the last thousand feet of runway. It stopped with barely two feet left.
Wheels up to wheels down, the Predator’s maiden flight had reached an altitude of about fifteen feet and lasted, by Pace’s count, fourteen seconds.
“First flight!” Rutherford cried gleefully, declaring the program’s six-months-to-fly deadline met.
As those on the tarmac watched, a technician ran out and shut off the drone’s battery. When he did, the loss of voltage triggered a small explosive charge used to deploy an emergency parachute General Atomics had insisted the plane carry, and which the technician should have disarmed before killing the battery. Startled by the loud “pow!” the technician ducked and covered his head as the orange-and-white parachute puffed out, then settled gently over him. Rutherford was annoyed to learn that the company’s only parachute packer was away that day, so there could be no second flight. Others were relieved.
As in most flight tests, the prototype was rigged with telemetry that beamed data about its mechanical and other systems to recorders nearby. A study of the results showed there had been a sharp drop in fuel flow as the aircraft took off, and the engineers quickly saw why. To make the Predator as light as possible, and avoid the risk of a big explosion if it crashed, the General Atomics team had put just enough fuel in the tank to orbit the airfield a time or two and land. As the drone’s nose had risen during takeoff, almost all its fuel sloshed out of the engine feeds in the bottom of the Predator’s two fuel tanks, starving its motor of gas.
* * *
Abe Karem wasn’t present that day at El Mirage. Had he been, he might have caught such a simple mistake, for he was meticulous about testing. In the days leading up to test flights, he would always ask his teams one question over and over: “How can we fail?”
But a month earlier, just before the Predator was finally ready to fly, Karem had quit General Atomics. His job had never been a good fit. Tom Cassidy was one reason; the retired admiral couldn’t seem to construct a sentence without four-letter words, and though Karem had heard those words long before he knew Cassidy, he didn’t enjoy hearing them again, or so often. Cassidy’s temper bothered him as well: he would never forget how Cassidy had once grabbed Karem’s briefing papers after a presentation, tore them in half, and threw them in a trash can to show how violently he disagreed with him. Karem also had little financial incentive to stay at General Atomics. Neal Blue was poised to reap the rewards of an inventi
on Karem had devoted more than two decades of his life to developing, yet he was paying Karem nothing beyond his salary.
Those were probably reasons enough to leave, yet they weren’t the main impetus. In truth, Karem decided to part ways with General Atomics for the same underlying reason he had quit IAI in 1974 and started designing drones in the first place. At heart, Karem was an inventor, pure and simple. To be happy, he simply had to be working on something new—and he had to be doing it his way. As he put it himself, “I am not a guy who can stay in standing water with high nitrogen and frogs jumping around the water lilies. I am a salmon, and a salmon needs oxygen—fast streams with a lot of oxygen.”
After leaving General Atomics, Karem set out on his own to compete for the next big UAV contract the Pentagon was going to award. This one was “Tier II Plus,” a surveillance drone able to fly at altitudes of sixty thousand feet and ranges up to three thousand miles. Karem already had a design, an unmanned flying wing he called the W570. He also already had a new company, Frontier Systems Inc., a legal entity he had created in 1991, right after Leading Systems went out of existence.
Frontier Systems didn’t win the Tier II Plus contract, which ultimately produced a drone by Northrop Grumman Corporation called the Global Hawk. But not long after Karem lost that bid, he called his former DARPA ally Bob Williams, who had given him his contract for the Albatross UAV demonstrator in 1984, and Williams, now a special adviser to the general in charge of the U.S. military’s Southern Command, offered some thoughts. Over the next few years, on contracts provided by DARPA, Frontier Systems developed the A160 Hummingbird, a small, lightweight, unmanned helicopter whose rotor changed its speed in flight to turn at the most efficient number of revolutions per minute. The revolutionary design brought Karem full circle: in 1974, Israeli Air Force commander Benny Peled had urged him to design a drone helicopter, and though Karem had done so, he’d never found a buyer for it. This time he did. In May 2004, Boeing acquired Frontier Systems and its A160 for a price never officially announced but rumored to be as much as sixty-five million dollars. Just over a month shy of his sixty-seventh birthday, Abe Karem had finally turned a profit on one of his dreams, and a handsome one.
Predator Page 10
