They found a garage Abe liked, attached to a house on a hill in Hacienda Heights, a suburb east of Los Angeles. The garage had 600 square feet of floor space and an equally spacious attic. By 1981, both spaces were crammed full of tools, computers, and handmade molds to fabricate aircraft parts from lightweight composites such as fiberglass and carbon epoxy. Working with Abe in the garage were two other believers in UAVs: Jack Hertenstein, a brainy, bashful engineer and radio control modeler Karem had met at Developmental Sciences, and Jim Machin, a pre-med student who’d impressed Abe at a free-flight modeling meet.
The trio produced a UAV demonstrator that was feather-light—it weighed only 200 pounds—and would carry a television camera in its nose. Hertenstein contributed avionics and ground control, and, Karem says, “tremendous expertise in flying his automated model aircraft without crashing.” According to DARPA calculations, it would stay aloft a stunning 56 hours. Karem named it Albatross.
During Karem’s brief stay at Developmental Sciences, he had met Ira Kuhn, a technology entrepreneur who visited the company to evaluate its UAV on DARPA’s behalf. In the course of conversation, Kuhn described an engineering problem he had been working on. A few days later, Karem called him with a solution. “It was extremely clever and much better than mine,” Kuhn recalls. Kuhn kept in touch, and was so impressed by the Albatross that he told DARPA director Bob Fossum, “This guy is a national asset.”
DARPA ended up funding the Albatross flight tests. The drone’s exceptional performance led the agency in 1985 to contract with Karem’s new company, Leading Systems Inc., to develop a larger endurance UAV the agency named Amber. Navy Secretary John Lehman was pushing the development of UAVs as spotters for the guns on Navy ships. Largely financed by the Navy, Amber also had champions in U.S. Southern Command, who wanted “persistent surveillance” of drug traffickers in Latin America, recalls former DARPA official Bob Williams, who initiated the project.
Thus were the seeds of the drone revolution sown. The Predator, though nearly twice as large, is a direct descendant of Amber, as their near-identical configurations suggest. Amber’s thrust came from a two-blade wooden propeller at the rear of the fuselage. Its tricycle landing gear retracted into its body. The V-shaped stabilizers at its tail pointed downward. Karem chose that distinctive feature mainly for aerodynamic stability but also because, in a rough landing, the stabilizers could hit the ground and skid, keeping the propeller blades from shattering on the runway.
Like the Albatross, Amber was radio controlled and could be configured to take off and land like conventional aircraft or fold its wings and stabilizers for launch by rocket from a canister. Most ingenious of all, Amber could be recovered by putting it into a “deep stall”—a feature used in free flight glider models to escape thermal updrafts and stay within flight time limits. The stall would bring Amber to “a near-vertical landing so it could be used from small ships, submarines or trucks/trailers,” Williams explains.
By 1988, Karem was expecting to build large numbers of the UAV for the Navy. With GPS becoming available and other key technologies gaining momentum, interest in UAVs was growing. Leading Systems now occupied 200,000 square feet of industrial space in Irvine, California, and had a contract on an airfield near El Mirage Dry Lake. In June of that year, an Amber flew over El Mirage at 5,500 feet above sea level for 38 hours. Even so, the Navy dropped a plan to buy 200, and the Army rejected the aircraft when Karem offered it as a short-range UAV, a mission in which endurance was no advantage. Around the same time, piqued by the money spent on programs like Aquila and Condor, which failed to develop useful and affordable UAVs, Congress froze funding for such programs and created a joint program office to consolidate drone development. DARPA transferred Amber to the joint program office, which promptly cancelled it.
Developing a useful and affordable UAV, however, is precisely what Karem had done. His achievement, says Bob Williams, “was not a high-aspect-ratio wing, good aerodynamics, or a lightweight structure, although good designers could do those things, and Abe could do them better than most. The issue really was reliability. Manned aircraft fly for tens of thousands of hours without crashing. So how do you even get in that ballgame? The crowning success of Amber was that by the end of the program, we were able to fly 650 hours without a loss. That was a huge order of magnitude increase in reliability [over other UAVs at the time].”
When Amber’s future looked unsure, Karem shifted his focus to a lower-tech export version his team had been working on with private funding: the Gnat-750. The number refers to the chord of the wing in millimeters at its root, a dimension of the airfoil from leading to trailing edge. The Gnat was bigger, heavier, and less capable than Amber, with a two-stroke Rotax 532 engine and no canister launch capability. While Karem tried to make a foreign sale, a $5 million bank loan came due.
Alerted to the opportunity by a mutual friend, the owners of General Atomics, brothers and reputed billionaires Neal and Linden Blue, bought the assets of Leading Systems out of bankruptcy in 1990. Their company, which had been working on its own UAV with minimal success, got everything Karem had achieved with Amber and Gnat, Karem says. The Blue brothers, lifelong aviators and entrepreneurs who also saw unmanned aircraft as a beckoning frontier, hired Karem and eight of his people, getting “a team that was flying advanced UAVs like nobody else,” Karem boasts.