Predator (9 page)

By mid-1985, Williams had persuaded the Navy, Marine Corps, and Army all to help DARPA develop the Amber, with the Navy taking the lead. The Navy Department agreed in December 1984 to a “remote control vehicle joint program” aimed at creating a drone able to fly twenty-four hours or more at altitudes of up to thirty thousand feet. DARPA would award Leading Systems a five-million-dollar contract to get under way and manage the project until the company built and demonstrated an Amber prototype. The Navy would invest twenty-five million dollars and take charge once the prototype proved the basic design was sound. Under Navy supervision, Leading Systems would develop different Amber versions able to carry various payloads, from daylight and infrared video cameras to special radars and sensors able to intercept electronic communications. Eventually, the company might also develop an Amber with a warhead in its nose.

In July 1985, the Army joined the program, raising the total Amber budget to forty million dollars. Two months later, Karem met with Melvyn R. Paisley, an assistant secretary of the Navy, and came away believing that the Pentagon would hire Leading Systems to produce up to two hundred Ambers a year once he proved the technology. The CIA was also interested in the drone, though in just what way and to what extent still remained a secret that participants in the program were forbidden by law to reveal three decades later.

The Amber project commenced during a memorable period in Karem's life. On March 16, 1984, seven years after arriving in America from Israel, he and his wife, Dina, at last became U.S. citizens. Hearts swelling, they took the oath of allegiance along with hundreds of others in a naturalization ceremony held under the glittering chandeliers of the Dorothy Chandler Pavilion, a Los Angeles venue that hosted the Academy Awards that year as well. Karem's new country also seemed to be recognizing that its new citizen was an aviation pioneer and an invaluable resource: a couple of weeks after his naturalization, Karem got a Top Secret clearance from the Defense Department. Following what a friend later described as his “wilderness years” in the garage, Karem was sure his vision was at last about to be realized.

*   *   *

From the moment Williams gave him the green light, Karem was a man in motion, doing a thousand things at once to get Leading Systems out of his garage and Amber into the air. He was designing, calculating, negotiating, researching, budgeting, buying, leasing, hiring—seemingly every day, seemingly all at once. Some nights he got by on only two hours of sleep. Weekends were workdays. He was going full throttle because at last he was living his dream. He was determined to build a company—a team, to be precise—that would defy the norms of the military-industrial complex, whose bloated bureaucracy, corporate mentality, and corrupting politics had made cost overruns, schedule delays, and outright failure in defense programs commonplace since World War II. Above all, Karem dreamed of leading the way into a new era of unmanned aviation the way pioneers of the early twentieth century led the way into the air—people such as Glenn Curtiss, Donald Douglas, Jack Northrop, Igor Sikorsky. These aviation greats—“people who should make us humble,” in Karem's words—worked aeronautical wonders with small but talented teams of collaborators. They produced aircraft that pilots and passengers could rely on and nations could use to deliver the mail, expand commerce, and win wars. That was how Karem wanted to make drones.

Knowing government money was on its way, he tapped his savings and borrowed cash from his mother, brothers, and other family members. He leased an industrial building in Irvine, thirty miles south of his Los Angeles home, figuring a town in coastal Orange County would be more attractive to the kind of people he wanted to hire. With eighteen thousand square feet of floor space, the structure had more than enough room for offices and the various shops needed to create prototypes. Karem also planned to use the facility to refine and produce a less capable drone called the Gnat. He and his first employee, Jim Machin, had created the Gnat in Abe's garage, intending to sell it as a target drone and a trainer for RPV pilots. The Gnat was smaller than the Amber, and its wing attached at the bottom of its fuselage, rather than at the top as with both the Amber and the Albatross. The family resemblance among the three drones, though, was clear. All three had a high-aspect-ratio wing, a pusher propeller, and inverted-V tail fins.

By the time Air/Space America 88 came along, Karem was in his fifth year of developing the Amber. A company brochure boasted that Leading Systems had “a team of 95 exclusively devoted to development, production and flight operations” of UAVs, or unmanned aerial vehicles. The term UAV was becoming more fashionable than RPV because of the increasing autonomy in air vehicles made possible by the rapid evolution of computer technology. Small as it was, Leading Systems now had six engineering departments, five manufacturing departments, and two test groups. Among the company's key players was Frank Pace, a software engineer with degrees in math and computer science whom Karem had hired to help run the company. Pace had come from Brunswick Company's defense division, and his talent for writing software would be crucial to making a digital flight control system for the Amber. Only four such systems had been built at the time, including one Brunswick had done for a decoy drone.

Karem remained his company's chief designer, and the Amber was his creation. The prototype of the new drone weighed nearly five times as much as the Albatross, held three times as much fuel, and could carry video cameras, radar, or other payloads weighing up to a hundred pounds. Karem designed and constructed one version that would take off and land from a runway conventionally; DARPA wanted this version built first to prove the drone's endurance and test its flight characteristics. He also designed a model with features he and Williams assumed the Navy would need. This “Amber I” lacked landing gear, and its wings, fins, and propeller were meant to fold against the fuselage so it could be launched at sea from a torpedo tube or a canister. The drone's appendages would unfold after launch, blossoming into their functional positions after the UAV was blasted into the air by a booster rocket, which would fall away after doing its job. Upon returning to a ship or sub, this version would go into a deep stall—much like the free-flight models Karem had flown years before—then fire a retro-rocket downward to ease its descent into a shipboard net. Ever inventive, Karem had even designed this version so it could be launched straight out of its transport box by ground troops or put into flight from under the wing of a manned aircraft.

Karem's creativity didn't stop there. The Cold War had been heating up in recent years, as President Reagan challenged the Soviets rhetorically and beefed up the military to deter and, if necessary, fight the Warsaw Pact in Europe or other Moscow allies elsewhere. Karem imagined a range of possible missions for the Amber, so for added versatility he designed three different fuselages. One had a nose shaped like a snake's head to house both a TV camera and a small radar. The nose of the second bulged downward to accommodate a “moving target indicator” radar, which could distinguish objects in motion from ground clutter. A third Amber design had a missile-shaped nose to carry a warhead. Karem wanted his potential military customers to see all the possibilities.

In mid-1986, Karem moved his company to a two-story, 180,000-square-foot building in Irvine that was far too large for his team but provided enough space for the sizeable manufacturing orders he anticipated from the Navy and Army. He also leased, with an option to buy, a small airfield near El Mirage, California, a dry lake bed in the desert about ninety miles north of Irvine. He wanted his own airfield where flight testing could be done, pilots could train, and the Gnat could be used to evaluate software and technologies developed for the Amber. Besides doing all its own testing, Leading Systems built nearly every component of its drones in house, from composite structures to the “actuators” that move aircraft control surfaces; from flight control and ground station electronics and computers to gearboxes, alternators, and even propellers. The company's machine shop fabricated mechanical parts of all kinds, right down to the landing gear. After hiring more than a dozen engineers and technicians with Formula One and other race car experience, Karem even set out to build the first engine expressly designed for UAVs.

*   *   *

On the surface, Leading Systems certainly seemed to be a success. Two weeks after the Air/Space America show ended in May 1988, visitors to Karem's booth at yet another trade fair in San Diego, the Fifteenth Annual Association of Unmanned Vehicle Systems Technical Symposium and Exhibit, could see just how extraordinary his new drone was. At 7:48 a.m. on June 6, minutes after the show opened, an Amber took off from Karem's airfield at El Mirage. Simultaneously, a digital electronic display at his booth in the trade fair began tallying the Amber's flight time. The display added minutes all that Monday. It added minutes all that night and throughout the next day. It added them into Tuesday night. At 10:10 p.m. on June 7, Karem's team finally landed the Amber—a phenomenal thirty-eight hours and twenty-two minutes after takeoff, an unofficial record for a UAV. The Amber had stayed aloft several hours longer than any other drone ever demonstrated.

That stunning performance wasn't the only reason outsiders might assume Leading Systems was a success. Nearly a year before the Amber set its endurance record, the officer running the Navy's unmanned air vehicle office, Captain Penn E. “Pete” Mullowney, had told
Aviation Week
that his service was “working to procure a total of 96 Amber vehicles capable of ship and ground launch,” and that they “could be fielded with the Marine Corps and Navy units in early 1990.” A year later, Mullowney gave
Aviation Week
another rosy prognosis for the Amber. To all outward appearances, then, the Navy was happy with the Amber, and in May 1988 the drone was apparently on a steady course toward putting its unprecedented capabilities to work helping defend America. But, as Karem learned the hard way, appearances can be deceptive in the byzantine world of military procurement, where politics and personalities can be more decisive than performance in determining what ultimately is purchased.

As early as August 1987, in fact, Karem had begged DARPA to hold off on transferring his project to the Navy. A seismic shift in the politics of military drone programs was taking place, and Karem feared that his project might get buried in the legislative avalanche that followed. The trigger was the boondoggle known as Aquila, which earlier that year had attained the status of public scandal. In tests at Fort Hood, Texas, between November 1986 and March 1987, the Army UAV performed correctly on a mere 7 of 105 flights—after the service had spent fourteen years and $1.2 billion developing the artillery-spotting drone. Outraged, defense experts in Congress took a closer look at all drone programs. Finding “excessive redundancy” in the few drones the armed services were developing, they decided radical surgery was needed. The fiscal 1988 defense appropriations bill directed the Pentagon to consolidate drone research and development under a new multiservice Unmanned Air Vehicle Joint Program Office—the JPO, as insiders called it, pronouncing it “JAY-poe.” The bill also scrapped all funding for the Aquila, and to “encourage” the JPO to eliminate duplication, Congress slashed the year's budget for drone research and development in half—from $103 million to $52.6 million. For good measure, the legislation froze all spending on drones until the Pentagon produced an “improved RPV master plan.”

A few months later, the Pentagon designated the Navy to run the JPO, and now Karem knew his project was in trouble. From the time the Navy created his RPV office in 1985, Captain Pete Mullowney's top priority had been to bring into the fleet a smaller short-range drone like the one Navy Secretary Lehman had seen in Israel, the Mastiff. Mullowney's office was developing a mini-RPV similar to the Mastiff, called Pioneer, and the Pioneer was being built by a joint U.S.-Israeli team of contractors with considerable political clout. With the Pentagon budget for drones cut in half, Karem felt all but certain that if push came to shove in a contest for dollars between the Pioneer and the Amber, Leading Systems would get pushed and the Amber would get shoved.

To make matters worse, Karem and his JPO overseer were constantly at odds. Mullowney had assigned the Amber program to Lieutenant Colonel Joseph G. Thomas, a Marine Corps aviator and a recent graduate of the Program Managers Course at the Defense Systems Management College, whose core curriculum was bureaucracy. Cultural and temperamental opposites, Thomas and Karem clashed over the marine's focus on process and the inventor's focus on performance, the bureaucrat's imperative to manage the Amber and freeze its design for production and the inventor's impulse to improve his creation as much as possible before production began. Karem was terrified that too much government regulation would turn his Amber into another Aquila.

Karem and his overseers also battled over the government's money, which Thomas thought Karem spent far too freely. “I could not control Abe Karem,” Thomas complained years later. “I couldn't control Abe's burn rate, and I was running out of money.” His time overseeing Karem's project, Thomas mused, was “my most notorious failure of any kind as a manager.”

Karem bridled at what he saw as inept or political interference by the government. His relationship with his Navy overseers wasn't helped by his tendency to display disdain for those who either failed to appreciate his insights or offered ideas he deemed wrong. Ronald Murphy, who oversaw the Amber for DARPA before the Navy took charge of it, later admitted that he had been as worried as Karem about what would happen when the JPO took charge of the project. “My concern was it would be cancelled fairly quickly, because nobody would put up with Abe, and he no longer had an intermediary,” Murphy said. “You can only tell people they're stupid for so long.”