by late 1944, the tide of World War II had turned against Nazi Germany. Under the leadership of General Dwight D. Eisenhower, Allied forces were preparing for an invasion of Europe to liberate occupied countries and destroy the government of Adolph Hitler. Nazi military forces were stalemated or losing on both fronts, while their leaders were getting glimpses of the ruin that awaited the German people once the invasion began. This dark future left the Nazi leadership desperate for ways that they could forestall a disaster and, if not, punish their enemies with a last, desperate strike. German scientists were put to work by Hitler to create what he referred to as “vengeance weapons” which could strike at civilians in the United Kingdom and other Allied countries.1 Among these were the V-1 rocket, otherwise known as a “buzz bomb” or the “doodlebug” for its buzzing sound as it slowly flew over British cities, and the terrifying V-2 rocket. Fueled by liquid propellant and equipped with a sophisticated gyroscope for direction finding, V-2 rockets were lightning fast—they could reach their target in under five minutes from launch without warning—and extremely lethal.2 One estimate suggests that V-1 and V-2 rockets killed more than 8,000 people and injured another 22,000 in Britain alone.3 In some ways, the psychological effect was almost as bad as the physical impact. In language that echoes the contemporary criticisms of drones, the writer Evelyn Waugh described the V-1 rocket as “impersonal as a plague, as though the city were infested with enormous venomous insects.”4 Beyond their impact on their victims, the V-1 and especially the V-2 rockets were an alarming symbol of the technological sophistication of the Nazi regime. Under the leadership of Werner Von Braun, German scientists had created nearly space-age technology with a V-2 rocket that could fly hundreds of miles to a preprogrammed destination according to an automatic guidance system. Aware that Germany had plans for even more lethal rockets and nuclear weapons on the drawing board, Allied leaders began to consider ways to destroy V-1 and V-2 launch sites to slow their progress. The problem was that V-1 and V-2 launching pads located on the coast of northern France were difficult to hit. Most of the contemporary Allied missiles were not precise enough to strike them directly or powerful enough to shatter their steel and concrete frames.5
To tackle this problem, US Air Force officials began to consider ways to turn nearly out-of-service aircraft into missiles that could strike hardened German bunkers along the coast. These explosive-laden planes would fly low, at around 2,000 feet, toward the target but be followed by a larger plane, dubbed a mothership, which would adjust its auto-pilot functions and help steer it to its final outcome.6 To hide their approach in daytime, they were painted yellow or white. At the time, these “aircraft turned cruise missiles” were called drones because they were controlled by radio from another aircraft.7 According to US Air Force Chief General Henry “Hap” Arnold, these drones were desirable because “if you can get mechanical machines to do this, you are saving lives at the outset.”8 Working with General Carl A. “Tooey” Spaatz, Arnold ordered the Air Force to retrofit B-17 planes with the newest military autopilots and radio systems and to pack them full of explosives. Their secret plan, called Operation Aphrodite, began test flights in Florida and Britain to see if strikes against German missile bunkers would even be possible.
As the plans moved forward, Air Force officials gradually hollowed out both B-17 and B-24 planes and transformed them into remotely controllable cruise missiles. They packed the planes, newly dubbed BQ-7s, with 30,000 pounds of RDX explosives by stripping out every piece of payload not necessary for converting the plane into a flying missile.9 These planes were heavy, weighing over 64,000 pounds, but still airworthy. The biggest technological problem that they faced was the remote control of the aircraft. While the mothership flying behind the explosive-laden aircraft could make adjustments to the trajectory of a BQ-7 in flight and even convey television pictures of the cockpit and control panel, Operation Aphrodite required pilots to be in the aircraft to get it airborne and to point it in the right direction of the target. The control of the plane from the mothership was not sophisticated enough to allow for the plane to take off on its own as a truly unmanned vehicle. It was also not good enough to ensure that it was going toward the right target. The remote control system, originally designed for directing weapons to narrow railroad bridges, did not allow for the precise targeting of its destructive payload across significant distances.10 As a result, two airmen—a pilot and an autopilot technician—would need to be on board for the initial phase of the flight. The pilots would get the plane into the air and make sure it was headed toward the intended target. If all went to plan, the pilots would bail out with parachutes over the English Channel and later be collected by boat.
The problem was that these missions rarely went to plan. The dry-run tests were beset by technical problems and difficulties reaching their targets. Some planes—nicknamed “baby ships” or “topless roadsters” by the pilots—strayed far off course. Others were shot down by friendly fire.11 Although the English antiaircraft artillery crews were warned not to fire on the flights, they often did, causing the pilots to frantically bail out.12 Some pilots found it difficult to get out of the plane in time, with one pilot winding up in the English countryside trying to explain to a puzzled crowd how he got there.13 On August 4, 1944, during a particularly disastrous test run of four BQ-7s, one aircraft exploded and the other three missed their target.14 Two airmen—a pilot and an engineer—were killed in the disaster, and two acres of English countryside were scorched in the fireball that followed one of the crashes.15 A few days later, additional tests failed, resulting in the death of another pilot when his parachute did not open.16 By one estimate, none of the eleven test flights of Operation Aphrodite were successful.17 Yet the test flights continued, and the US Navy even joined in the exercise, retrofitting PB4Y-1 bombers (equivalent to B-24s, also called Liberators) to be flown at the German submarine pens in Heligoland.18
In September 1944, the US Navy set its sights on a massive underground complex in France, called the Fortress of Mimoyecques, where, it was rumored, Germans were developing next-generation V-3 rockets. The target was even more difficult than usual: the navy would need to guide the plane into a doorway of the fortress, no more than 15 by 16 feet, which was the most vulnerable part of the structure.19 To hit such a precise target, they needed skilled pilots to fly a B-24 Liberator (renamed by the navy a B-8) packed with Torpex explosives to the target. Among the volunteers was Joseph P. Kennedy Jr., the older brother of future president John F. Kennedy. The oldest Kennedy brother was already being groomed for high political office by his ambitious father and had served for two tours of duty with the US Navy flying B-24 Liberator planes. He volunteered for the mission despite being eligible for stateside duty. He was publicly optimistic about its outcome, but privately he was aware of the risks involved. Before the mission commenced, Kennedy passed a message to a family friend: “if I don’t come back, tell my dad—despite our differences—that I love him very much.”20 Paired with co-pilot Wilford “Bud” Willy from Texas, Kennedy took off on August 12 from an airfield near Norwich.21 Shortly after the handoff of remote control to the mothership, less than a half an hour into the flight, the plane exploded in an enormous fireball, raining down debris on the English countryside.22 No trace of Kennedy or Willy was ever found. The Kennedy family lost one of their favorite sons, one tipped for high political office, in what the navy only described as “an exceptionally hazardous and special operational mission.”23
The disaster was ultimately investigated by an Air Force top secret panel which identified a number of technical reasons for the disaster, but could not agree on a single cause. Some believed that it was due to a technical fault; others speculated that signal jamming may have detonated the charge and produced the explosion.24 The navy concluded that the Torpex explosive was too unstable to be reliably used and soon replaced it with TNT. As a program, the conversion of these old aircraft into flying bombs was deemed an abject failure. Almost no Nazi targets were ever hit, and the program was canceled as retreating Nazi forces abandoned the missile launch sites along the French coast.25 For the Kennedy family, the disaster was compounded by the navy’s decision to keep the details of the failed mission a secret.26 This disaster also subtly altered the course of history. It transferred the ambitions of the Kennedy family to the younger son, Jack, who was also serving in the navy at the time. The persistence of the US military with this dangerous project showed how far it would go to achieve something it had long dreamed of: the ability to fly and to strike long-distance enemy targets precisely and by remote control. This dream, long in the making, would ultimately not be realized for nearly another fifty years.
The early history of drones is not a linear or clean one. It is a story of mistakes, false starts, disasters, and failed prototypes.27 There were dozens of prototypes and failed experiments that contributed in varying ways toward the drone models that we see today. These prototypes were also not often seen, or described, as “drones” in the modern sense, but they did help to clarify and resolve some of the technical issues that afflicted early drone development. The evolution of the drone was also inextricably tied to the development of satellites and cruise missiles; indeed, in many respects, the technology behind Kennedy’s failed mission was closer to modern cruise missiles than to today’s drones. The early history of drone technology is also not a story that could have been told without the financial support of the military. Like the Internet itself, drone technology was incubated over decades by the military-industrial establishment and was used for secret military operations for years before appearing in the public domain and broadened their capabilities. Drones also might have also remained an obscure, military-only technology in the absence of significant technological changes—for example, the rise of Global Positioning Systems (GPS) and satellite communications—that spilled from the military into the commercial domain.
The earliest evidence of the thinking behind drones lies in efforts to create remote-controlled cruise missiles at the beginning of the aviation age. In some respects, the effort to create unmanned aircraft predates the revolution that the Wright brothers kicked off with manned aircraft. The eccentric but brilliant inventor Nikola Tesla was among the first to propose a remote-controlled cruise missile in 1898, creating a prototype called the “teleautomaton” which he later tried to mass-produce.28 One of his colleagues, Elmer Sperry, was the first to solve several of the technical problems that Tesla encountered—for example, stabilization of the plane in mid-air and a crude version of remote control by radio to allow the plane to be steered in mid-air. The steering was limited, however, as these aircraft were generally able to fly only toward a pre-determined location. Funded by the US Navy, Sperry’s experiments were designed to develop an “aerial torpedo” to be launched into German U-boats and other targets, much in the same way that Kennedy’s craft would later aim to do.29 He had to experiment with different ways to get these planes into the air, including most notably attaching the aircraft to the top of his car and driving it at 80 miles per hour down the Long Island parkway to create a wind tunnel. The airplane nearly lifted the car off the track once it hit flying speed.30 By March 1918, he had devised a way to launch the bomb-laden unmanned aircraft via a catapult system that would allow it to fly approximately 1,000 yards and still be reused.31
This issue—reusability—would later become central to the distinction between drones and cruise missiles. Today, the typical assumption is that drones are reusable but cruise missiles are not. But in the early twentieth century, the distinction was not as clear. Many of the forerunners of drones were explicitly designed to be flying bombs or cruise missiles. Some advocates of unmanned aircraft experimented with reusable prototypes; others worked with crafts that were deliberately expendable. Many of the technical issues—remote control, precision, and payload capacity—were the same. As usual, wartime necessity forced inventers to confront these problems and develop solutions. In World War I, British military officials commissioned an eccentric engineer, Archibald Low, to create a radio-controlled aircraft that could be used as a flying bomb against the Germans.32 The prototype worked and showed in-flight navigation was possible over short distances, but it was hardly a rousing success. After watching one unsuccessful test flight, one British military observer snapped, “I could throw my bloody umbrella farther than that!”33 Across the Atlantic, the US Army turned to Charles F. Kettering, founder of the Dayton Engineering Laboratories Company, and to Orville Wright.34 Together, these men—sometimes rivals but here joined in the war effort—produced the Kettering Bug, one of the first mass-produced, widely used unmanned aircraft. The Kettering Bug was relatively small—only 12 feet long and made of wood—but it could carry a 180-pound explosive.35 The underlying idea was that the Kettering Bug would be able to fly approximately 75 miles to a predetermined target so long as the engineers had preprogrammed the engine to do enough revolutions to reach the target.36 The Kettering Bug was the first airplane that could be both stabilized and navigated remotely through the use of pneumatic sensors and gyroscopes.37 The army considered developing between 10,000 and 100,000 Kettering Bugs, but in the end fewer than fifty were actually made.38 The British Royal Navy also experimented along these lines, creating aircraft that blurred the lines between drones and cruise missiles. One of their experiments was the Larynx, a fixed-wing, unmanned aircraft which could fly as fast as 200 miles per hour. The early trials of the Larynx showed promise as it was able to fly 112 miles toward its target, even if it crashed 5 miles short of ultimate destination. Just as the US military would later do, British military officials flew the Larynx over the deserts of Iraq in 1929.39 But only a few prototypes were made, and research on drones as cruise missiles halted for a while.
In the 1930s, government attention shifted from turning planes into bombs toward forms of remote control of a craft by radio. Both the United States and United Kingdom started looking for small unmanned aircraft that could be used for target practice for anti-artillery weapons. The idea was to use inexpensive drones to improve the accuracy of gunners in shooting down moving targets. In 1932, the Royal Navy introduced an early target drone prototype, called the Fairey Queen, which was then used for a series of test runs but was eventually destroyed.40 Royal Navy officials then took parts from two manned aircraft—the de Haviland Tiger Moth and the Moth Major—and created a radio-controlled aerial target aircraft. The Queen Bees, as they became known, were controlled remotely by a radio signal to circle boats in the water for anti-aircraft artillery practice.41 They were considered so successful that eventually 400 were constructed by the Royal Navy, a fact which drew attention and investment from the US Navy.42 By the mid-1930s, the Northrup company had a division wholly devoted to UAVs and planned to develop a thousand of their own US version of the Queen Bee, called the OQ-2A target drone.43 In the end, the Queen Bee was influential in establishing the contours of what a drone is supposed to be—small, mobile, and reusable—and it also arguably gave its descendants the name “drone.”44
By the mid-1930s, the growing sophistication of radio technology had led to the emergence of a hobbyist community devoted to radio-controlled tiny planes. Among these hobbyists was Reginald Denny, a British émigré to the United States. Denny began his career as B-list Hollywood actor in silent films during the 1920s and continued to act sporadically until the 1960s.45 After founding a company in Van Nuys, California, in 1934, Denny and his collaborators came up with a prototype radio-controlled plane—nicknamed the Dennymite—which attracted the interest of the US Air Corps (later the US Air Force).46 In 1940, he successfully pitched the aircraft to the US Army, which agreed to buy fifty for target practice. After the Japanese attack at Pearl Harbor, the air force and navy also expressed an interest and that number increased to a planned 15,000 drones.47 The Radioplane Models had a wingspan of nearly 12 feet, weighed 130 pounds, and later models could fly as much as 140 miles per hour.48 Because of the insatiable needs of the military service during World War II, thousands of Radioplane models were quickly constructed and put into service, but never anything close to the numbers the military originally planned.49 This effort drew the interest of Captain Ronald Reagan of the US Army’s First Motion Picture unit based in Culver City, California. Recognizing a good story that would aid the war effort, Reagan assigned Private David Conover to take pictures of workers on the shop floor.50 Conover spotted a beautiful, brown-haired woman named Norma Jean Dougherty whom he thought had potential as a model. Dougherty—who dyed her hair blond and later became known to the world as Marilyn Monroe—was first photographed as a young model smiling with a Radioplane propeller (fig. 2.1).51
Figure 2.1 Norma Jean Dougherty—later known to the world as Marilyn Monroe—photographed in June 1945 with one of Reginald Denny’s OQ-2 Radioplanes.
Among the most important technological innovations was the adaptation of televisions for use in unmanned aircraft in order to provide a view of target beyond the line of sight of the operator. The US Navy began experimenting with placing TV transmitters on aircraft to allow operators located in a nearby aircraft to see out its cockpit and steer it to the proper destination. By 1941, the navy expanded this to a full-fledged, 70-pound RCA television in the drone that could transmit a picture up to 30 miles away.52 That picture was sent to an accompanying aircraft that could make decisions on targets. Initial tests showed that these drones could hit targets with torpedoes with remarkable accuracy even when the drone was beyond the operator’s line of sight.53 Navy officials soon began to work on radar technology to allow transmissions of images to the accompanying manned aircraft at night and in poor weather.54
During World War II, US military production of drones involved an array of other target drones, including both biplanes and monoplanes, the efforts to convert B-17 and B-24 bombers into radio-controlled missiles, and the construction of Katydid drones which could compete in lethal accuracy with the German V-1 rockets.55 There was also a program for a US version of the V-2 rockets, but this was beset by logistical problems.56 Under the leadership of Lt. Commander (and later Rear Admiral) Delmar Fahrney, the US Navy developed aerial assault drones, and even contemplated using drones to ram enemy fighters.57 The navy converted a TG-2 biplane torpedo bomber into a remotely controlled drone and conducted a demonstration attack on a US destroyer. As Ann Rogers and John Hill have noted, “this can be seen as the first use of an unmanned combat air vehicle (UCAV).”58 By late 1944, the US Navy had developed drones that were capable of ramming targets and also dropping bombs and returning home. For the first time, drones were used in combat in the Pacific theater with several TDR-1 drones dropping bombs on and then later diving into stationary Japanese cargo vessels.59 The TDR-1 drones were built with a wood and steel frame using parts supplied by famous US companies such as Wurlitzer and the Schwinn bicycle company.60 It was controlled in flight by a manned aircraft flying alongside it and beamed its crude television images back to the pilots in that plane. Although hundreds of TDR-1 drones were ordered and constructed, the program was canceled in October 1944. Admiral Fahrney bitterly remarked that “the great broom of victory swept all new projects into the ashcan of forgotten dreams.”61 Yet the World War II investment in drones also deepened the commercial and industrial base for drone technology, with companies like Northrup, McDonnell, and Lockheed, the Glenn Martin Company, and others—all later big players in the modern military drone industry—getting their first experience with drones during this period.62
While the use of drones during World War II was more extensive than was widely known, the conduct of the war also led to a series of changes in the way that drones were conceived. Until World War I, drones were seen primarily as flying bombs. By World War II, military planners had concluded that they could also be used as target practice. Perhaps more importantly, they also saw the demand for intelligence on the activities of the enemy was increasingly essential, especially if the United States wanted to avoid the large-scale bloodshed that marked so many engagements during the world wars. The problem was that unmanned aircraft lacked the technological sophistication to be useful for this purpose.63 Once the war ended, drone production was significantly scaled back, and the successes of many prototypes were hidden from the public due to government secrecy. Yet privately the US military never wholly surrendered its desire to use drones, especially for surveillance and reconnaissance, and found new, often secret, uses for drones in the early days of the Cold War.
One of the first uses of drones for reconnaissance occurred at the birth of the atomic age. Shortly after the nuclear blasts at Hiroshima and Nagasaki, Japan, the US military realized that drones would be suitable for measuring the lethal radiation in the cloud that emerged after these blasts.64 In July 1946, the US Air Force flew two B-17s into the radioactive cloud that emerged after a nuclear test over the Bikini Atoll in the Pacific Ocean. These drones—called “babes”—were controlled by motherships 25 miles away, well out of the range of the radiation.65 Although the drones were damaged and buffered by the force of the explosion, they were still functional and managed to fly from Hawaii to California, a 2,174-mile journey, after capturing data on the radiation. In an August 1946 edition of the New York Times, the military expert Hanson W. Baldwin hailed the use of the drones as a rehearsal for a “push button war” and speculated “that what this can mean for the future of warfare—for that matter the future of air transport—requires little imagination to conceive.”66 He even speculated about the use of infrared that drones can see and how swarms of drones might be used in the future. Yet despite the advances in radio control, stability and the ability to deliver explosives, and the ability to “see” a target from a drone, drone technology remained too unstable to have the widespread impact that many assumed it would at the end of World War II.
Instead, drones were initially used for novel experiments to cope with some of the problems produced by the atomic age. For example, US military leaders began to consider how they could use drones to deliver nuclear weapons to targets. The first problem with thermonuclear bombs was that they were heavy (10,000 pounds) and would need to be hauled distances up to 10,000 nautical miles.67 The second, and far more serious, problem with thermonuclear bombs was that they “would produce a lethal area so great that, were it released in a normal manner, the carrier would not survive.”68 Thermonuclear weapons were so powerful that they would create an “inferno capable of charring wood at a distance of 20 miles” and a small-size hurricane with gusting winds that could destroy a manned aircraft.69 In 1949, in Operation Brass Ring, the US Air Force secretly examined delivering a thermonuclear bomb via a retrofitted B-47 bomber flown directly into the target. The idea was very similar to the mission that killed Joseph Kennedy Jr.: that B-47, equipped now with a nuclear payload, would be rammed into a target almost as a cruise missile. The air force experimented with deploying the aircraft from a mothership and using a ground control station, but its efforts were beset by technical problems. Ultimately, bureaucratic infighting, an increase in the size of the nuclear payloads, and technological innovations that would allow a manned fighter to survive this operation led the US Air Force to abandon plans for drone delivery of thermonuclear weapons.70
By the early 1950s, the US military began to experiment with drones that could conduct reconnaissance of Soviet battle positions. For example, one US Army experiment involved drones laying communication wire across a battle front; others were designed to be more flexible and to allow army units to cast the drones skyward for a picture of the battlefield, much in the same way that Raven and Black Hornet drones are used by the military today.71 One all-weather reconnaissance model, later known as the Falconer, was a precursor of the small drones like the RQ-11 Raven now seen on battlefields. A number of successors followed, and by 1964 the army had drones that were capable of real-time transmission of “over the hill” imagery.72 The US Marine Corps developed its own reconnaissance drone—known as the Bikini because, in the words of its Commandant, it was “a small item that covers large areas of interest”—which was able to produce photographs of the battlefield within minutes of landing.73 The US Navy also developed a number of prototypes, including anti-submarine drone helicopters. One series of models, called DASH drones, was designed to be launched off an aircraft carrier in order to spot enemy drones and even launch missiles at approaching submarines. All of these models were beset by technical difficulties and cost overruns, making it difficult to put them into service.
Within the US Air Force, research also continued on drones that could be used as targets to test pilot skills in aerial combat and the accuracy of air to air missiles. By 1947, President Truman became the first president to observe the operation of a drone from aboard a US aircraft carrier. In 1951, the air force introduced the jet-powered Firebee target drone, built by the Ryan Aeronautical Company.74 This fixed-wing drone resembled a missile and could be launched on the ground or in the air. It was eventually purchased by the other military services and became one of the most popular and widely used drones of all time. The US also sought to develop fast aerial target drones which could be used to simulate the movement of a missile or manned aircraft. One example was the MQM-74 subsonic drone, also known as the Chukar, which was developed by Northrup and was put into regular use in the United States, United Kingdom, and Italy in the 1960s.75 By 1975, an even faster stratojet aerial target drone, the MQM-107 or Streaker drone, emerged. It could simulate the heat and radar signals of real aircraft and maneuver accordingly, which was valuable for training operators of surface-to-air missile. Today, there are over 2,000 Streaker drones in a dozen countries.76 Many of the technical issues surrounding the use of target drones were largely resolved by the early 1960s, but conceptually they remained closer in function and form to cruise missiles than to the drones we know of today. The United States was also not alone in its pursuit of target drones. Russia and China had programs for target drones by the mid-1950s, with both converting manned aircraft into drones for naval gunnery practice.77
The real technological advances seen in modern drones came about through the use of drones for high-level Cold War surveillance from the early 1960s onward. Many of the novel developments with drones occurred in the shadows, as their development was ever more entwined with the priorities of the intelligence agencies during the Cold War. But as Lawrence Newcome has observed, the problem was that the Cold War often ran hot when it came to high-level surveillance. Between 1946 and 1990, 179 airmen were lost due to routine high-level surveillance of sites of interest.78 Crashes of military aircraft could result in international incidents or even crew being held as hostages; for this reason, unmanned aircraft were seen as preferable, assuming that the technology could be made to work. The necessity of developing unmanned aerial surveillance planes was made apparent in May 1960, when Francis Gary Powers’s U-2 spy plane was shot down by an SA-2 missile over the Soviet Union. This has been called the “genesis event” of UAV systems.79 Almost immediately, both the CIA and US Air Force announced bids for contracts for unmanned surveillance aircraft.80 The air force secretly awarded its contract to the Ryan Aeronautical Company to adapt the Q2 Firebee target drone.81 Yet the entire effort to build surveillance drones ran into bureaucratic headwinds. When the air force sought $1 million to invest in drones, they were flatly refused by Secretary of Defense Thomas S. Gates Jr., who responded, “I thought we weren’t going in this direction.”82 The Kennedy administration also toyed with drones but shied away from sustained investment. Instead, the immediate effect of the U-2 crisis was to accelerate efforts to invest in other surveillance platforms, such as satellites and manned aircraft. Thomas Ehrhard, who wrote one of the most detailed histories of US Air Force investment in drones, concluded that in the three-way competition between drones, satellites, and manned aircraft, “drones lost all the early battles.”83 Drones particularly lost the battle compared to the CIA’s manned spy plane, the SR-71, which was capable of flying at three times the speed of sound and could operate at altitudes that no Soviet missile could reach.84 Many of the problems usually associated with drones—control, stability, and the risk of accident—made them pale in comparison to manned aircraft like the SR-71 and even early Cold War satellites.
With the US military divided on the utility of drones, research and development into many drone prototypes only survived due to sustained investment by branches of the military intelligence agencies. The geopolitical competition with the Soviet Union meant that these agencies were rich with funds and could take risks by funding prototype drones, even if the models never paid off. The National Reconnaissance Office (NRO)—an agency so secretive that its name was once classified—poured millions into prototype research for drones that could perform the functions of U-2 aircraft.85 Created in 1961, this agency became a crucial vehicle for investment in drones by both the US Air Force and the CIA. Since much of the funds for drones were drawn from secret intelligence budgets, developers could experiment without drawing the Congressional attention and public scrutiny that they would with normal military procurement. One way that they did this was to work through the auspices of the US Air Force’s Big Safari program, which was designed to streamline the acquisition and development process for experimental aircraft and enable the building of riskier, cutting-edge prototypes.86 This secret infrastructure behind drone development meant that the US public remained largely unaware of the scale of drone development in the Cold War, even when drones became widely used in the military and began to have real consequences.
One moment when spy drones mattered was during the Cuban Missile Crisis. When the Soviet Union placed missiles in Cuba in October 1962, the Kennedy administration had to somehow get aerial reconnaissance photos without risking an accident that might spiral into nuclear war. The SR-71 was an obvious option, but it was high cost and had only limited availability.87 It also carried with it the risk that it would be shot down by a Soviet SA-2 missile, thus escalating the crisis. To address this, Undersecretary of the Air Force Joseph V. Charyk sought to use a new Firefly drone—essentially a Firebee drone modified for reconnaissance—and had these drones loaded on a C-130 mothership at a Florida air force base in the middle of the crisis. Air Force Chief of Staff General Curtis E. LeMay reacted badly to this plan. In the middle of the crisis, he killed the potential flight for fear that the overflights would reveal the capabilities of the US drone program to the Soviet Union and jeopardize their potential as a constant asset for strategic surveillance of Soviet military facilities. LeMay, known for his aggressive style and love of cigars, stormed into a briefing on the potential drone flights, threw out the participants, and told Charyk “hell no.”88 The need for unmanned aircraft became apparent soon enough. On October 27, 1962, only days before the end of the crisis, a U-2 plane was shot down over Cuba and its air force pilot, Major Rudolph Anderson Jr., was killed.89
This incident did not lead the air force to give up on drone surveillance. Rather, it began to institutionalize its drone operations. In July 1963, the US Air Force developed the first drone reconnaissance unit as part of the 4080th Strategic Reconnaissance Wing.90 Even more importantly, the NRO authorized the creation of a new and expanded Firefly model, dubbed the Lightening Bug (147B), to conduct surveillance at an altitude of up to 62,500 feet.91 The Lightening Bug could operate at high and low altitude, conduct surveillance of territory, and provide tactical benefits to US units on the ground.92 It was expensive, but it was so capable that it became among the most important of the Cold War’s drones. The Lightening Bug could pick up electronic intelligence on the Soviet Union’s SA-2 missile system;93 it could be deployed underwing from larger aircrafts like DC-130s and could travel a preprogramed range of up to 1,300 miles;94 it was capable of avoiding radar lock while in flight; and it could reach speeds of 600 mph.95 This small drone, with a wing span between 13 and 32 feet, could fly as low as 300 feet off the ground.96 Even though its initial performance left something to be desired, due to poor navigational accuracy and the risks of damage in flight, it nevertheless became the first drone to be put to widespread use by the US military for a range of surveillance purposes.
The first major use of the Lightening Bug was for surveillance over China’s key military and nuclear sites following its intervention in the Korean War. Under the code name “Blue Springs,” the United States began deploying Lightening Bug drones over Chinese military locations starting in 1964. The pace of drone surveillance picked up as the operational demands of the Vietnam War became more pressing. China’s decision to supply arms and other material to North Vietnam and its first nuclear test in October 1964 made US officials eager for greater surveillance of its activities. By 1965–1966, the United States had flown 160 Lightening Bug sorties over Chinese targets. At the same time, US officials were wary of another U-2 incident and wished to ensure that they kept plausible deniability in case one of the Lightening Bug drones were shot down. One way that they did this was to cover the drones with Nationalist Chinese markings to make it look as if it came from nationalist forces based in Taiwan.97 This ruse was unlikely to have convinced Beijing, but it did provide nominal political cover if drones were shot down.
The United States had good reason to be worried about drones being shot down. On November 15, 1964, only months after the drone overflights began, China shot down a US drone over its territory. This Lightening Bug was designed to trigger Chinese air defenses so the United States could see how they reacted. Although both sides knew the truth, the United States claimed to be baffled by the claims of intrusion and news of the incident gradually died down. This incident demonstrated to the United States that the Chinese and Soviets would implicitly accept that drone overflights were part of normal intelligence tradecraft, rather than a military incursion demanding a response.98 By 1965, the Chinese military had shot down eight Lightening Bugs. The United States was not deterred by losing the drones, and the intelligence collection via drone continued unabated.99 Unlike the Francis Gary Powers U-2 incident, no international crisis followed from any of the Lightening Bug interceptions. It did, however, spur the NRO to fund new experimental prototypes that could operate at high speed and in stealth in order to reach Chinese nuclear sites that were out of reach of existing drones and too risky for U-2 flights.100 Among these was the D-21 Tagboard drone, which had impressive speed and range but also high rates of accident and loss. Only the Lightening Bug remained in common use, and its models became adapted to a wide variety of tasks, including signals intelligence. By 1971, a modified version of the Lightening Bug, called the Buffalo Hunter, was even capable of flying amid cloud cover and poor weather conditions, two factors that had made flying with manned aircraft difficult if not impossible.
Even though they were commonly used, the Lightening Bugs had a capacity for accuracy that left something to be desired: they were not always able to direct themselves exactly to the target and could be as much as 3 or more miles off.101 One estimate even suggested that the Lightening Bug could drift as much as 6–9 miles off target during a 100-mile photo run.102 Course corrections could be done from the mothership, but this was difficult given the telemetry to the C-130 ship. Early forays with the Lightening Bugs produced little in the way of usable film: the first seven missions in 1964 produced only two reels.103 For tactical operations, Lightening Bugs needed to be extremely close to the target to be effective in producing photos. The film would also have to be flown to a location outside the flight path to be developed and used, which took time.104 Launch and recovery was difficult, even after the air force developed a sophisticated mid-air retrieval system to “catch” the drone in flight. Drone crashes were common, and recovery itself tended to damage the drone. This would not have been a problem if they were as cheap as typical target drones, but they were relatively expensive for the time.105 As a result, advocates of the Lightening Bug often had to make the case for a faulty technology against superior, manned alternatives, like the SR-71. These arguments became easier to make when the Soviet SA-2 missiles became much more effective and capable of knocking manned aircraft out of the sky, because pilots became more eager to hand off dangerous missions to their remotely operated cousins.106
One less-discussed obstacle to the adoption of the Lightening Bug was the organizational culture of the US Air Force during the 1960s. The air force often struggled to find commanders willing to support drones, in part because flying unmanned aircraft was seen as a poor career choice for pilots, as it was “drab and unexciting” and not professionally rewarded by the air force itself.107 It also ran against the organizational culture of the air force which promoted becoming a pilot, with all of the prestige and glamor that this implied, as the pinnacle of a service member’s career. This organizational culture produced resistance throughout the air force, in some cases because pilots feared being replaced and in others due to a lack of confidence in the technology.108 The command was not any more enthusiastic; in fact, some were so hostile that they did not even want to look at the drones. For example, in 1964, one former vice chief of staff for the air force, John Dale Ryan, was ordered by his boss to inspect the drones personally to reduce his skepticism. He flew to Eglin Air Force Base, got off his plane, touched one of the Lightening Bugs in its hanger, and said “There, I touched that little son of a bitch, now I can go home.”109 This strong organizational preference for manned aircraft could have led to drones being wholly abandoned during the Cold War, but the programs survived due to a combination of factors, including some motivated leaders within the air force, generous Cold War budgets that permitted experimental work through Big Safari and Lockheed’s famous Skunk Works lab, and the relentless demands of the intelligence agencies to know more about the battlefields that they might face if the Cold War ever turned hot.110
This last factor is what ultimately led to the expansion of the drone fleet and routine use of Lightening Bug drones during the Vietnam War for a wide variety of tasks, including gathering electronic signals, providing targeting data for missiles and ground units, and producing photographic and infrared imagery for pre- and post-battle assessments. Some uses were less straightforward. A few Lightening Bug drones, modified to drop propaganda leaflets on North Vietnam, were affectionately renamed “bullshit bombers” by their air force operators.111 Others were used as decoys to deliberately draw fire from Soviet SA-2 missiles stationed in North Vietnam, which allowed the CIA to find new ways to track and avoid those missiles.112 By the early 1970s, a few Lightening Bugs were equipped with their own Maverick missiles to directly fire at targets, making them capable of crude versions of the targeted killing missions that modern armed drones perform today.113 Initially used for reconnaissance of military sites in the Chinese mainland, Lightening Bug drones were flown in Vietnam in 1964 and were eventually relocated and flown directly out of Da Nang in South Vietnam by 1966. Although not widely reported at the time, their missions became routine. The 350th Strategic Reconnaissance Squadron reported flying 3,466 sorties in Vietnam with Lightening Bugs between 1964 and 1975. Perhaps more surprising was their survival rate: the air force managed to recover them in mid-air nearly 97% of the time, and only 578 drones were destroyed by Chinese or Vietnamese air defenses.114 One Lightening Bug drone, nicknamed the “Tom Cat,” survived sixty-eight missions before being shot down in 1974. Yet these drone operations, now routine and survivable, were not always effective: by one estimate, only 40% returned reconnaissance images, with a cost five times greater than that of manned aircraft.115
Once the Vietnam War ended in 1975, the use of reconnaissance drones was scaled back and US interest in developing the technology began to wane. The economic recession that afflicted the Carter administration made the Pentagon tighten its belt and cut back on programs that were not needed. In this environment, drones would naturally suffer: they remained more expensive and less reliable than satellites, which were showing tremendous gains in their capabilities as technology developed. They were also ill-suited to use within the European theater of war because of a congested flying environment with more manned military and civilian aircraft; moreover, their use was governed by some arms control regimes put in place by agreement with the Soviet Union.116 In this respect, the shared lineage of the cruise missile and drones proved fateful to the latter, as UAV systems were defined as restricted military technology and subject to limitations on their use within the European theater of war. Perhaps the most consequential development was the transfer of control over drones from the NRO to the US Air Force’s Tactical Air Command (TAC), which sought control of the drone fleet but was ultimately opposed to the huge cost overruns that accompanied drone development. The result was that each tightening of the defense budget led to a reduction in drones as many models were seen as wasteful and unnecessary compared to manned aircraft.
In the 1970s, research continued on a number of high-altitude surveillance drones, with mixed results. Some drones showed remarkable capabilities. For example, the HALE Compass Cope drone could fly for up to twenty-eight hours, but they were astonishingly expensive and could not be put into regular production.117 The Air Force wanted to have a Compass drone model provide nearly continuous coverage of military units in the Warsaw Pact and experimented with a number of Compass models that could achieve that capability. It also experimented with solar-powered drones to allow for greater endurance and intelligence collection capacity.118 In the late 1980s, the Defense Advanced Research Projects Agency (DARPA) funded a number of drones, such as the Boeing Condor and the RQ-3 Dark Star, all of which were designed to fly at high altitudes and at high speeds. Both models were costly failures, yet they yielded some design insights. The Condor drone proved that drones with longer wingspans could conduct long-range surveillance while the Dark Star showed that autonomous flight based on GPS data was indeed possible.119
The US Army wanted to know more about the battlefield as well and devoted its resources to developing drones for tactical use, including the SkyEye and Aquila, both of which were small drones that could be deployed in support of ground units with reconnaissance. Both of these drones reflected the hope that drones could be used for attacking targets on the ground: one prototype of the SkyEye could be equipped with tiny Viper missiles; the Aquila could use lasers on targets up to 7 miles away.120 The problem is that they were either unreliable (SkyEye) or too costly (Aquila) to be put into widespread use. Even the few times when the SkyEye was used, such as in El Salvador in the mid-1980s, it was derided as a “toy plane” by the rebels and had minimal real-world impact.121 The Aquila was a particularly devastating failure, as it cost $750 million over twelve years and only succeeded in seven of 105 test flights.122 Having already incurred this cost, the Pentagon was reluctant to spend another $1.1 billion for 376 Aquila drones.123 Throughout the 1980s, drones were increasingly seen as a boutique technology: an expensive novelty that was unnecessary given the information collection capabilities of satellites and manned aircraft.
Although US drone development stagnated in the 1980s, the desire to know what was happening on battlefields—in effect, to see around a corner or to anticipate danger—was increasing rather than diminishing. In World War II, the emphasis with drones had been on inflicting damage, as evidenced by the retrofitting of drones as missiles or even nuclear weapons delivery systems and the efforts to sharpen the skills of those manning antiaircraft weapons with target drones. The Cold War subtly changed some elements of the way drones were conceptualized; most Cold War drones were seen as a way of learning more about real and potential battlefields rather than directly inflicting harm. Despite the fact that a few prototypes of drones had been equipped with missiles, the majority of Cold War drones were unarmed and devoted to collecting information that could be converted to intelligence and put to use in the superpower competition.
The problem with US Cold War drones were that they were directed mainly at the superpower competition and were both highly sophisticated and expensive. The predominant focus of the United States was on conducting surveillance of the Soviet Union and to a lesser extent China. This meant a natural emphasis on high-altitude surveillance drones and satellites, as opposed to medium-altitude or other tactical drones. By contrast, other countries were beginning to look at the information-collection capabilities of smaller drones differently, seeing them as a low-cost, disposable mechanism to be deployed in great numbers to get information on more immediate threats.124 Israel was among the first to demonstrate the value of this alternative approach. After Egypt shot down two Israeli F-4 Phantom reconnaissance jets in early 1970, Israel started looking into drone technology for its ongoing battle with Egypt over control of the Sinai. Israel came first to the United States to investigate buying the Compass Arrow drone, but President Richard Nixon ordered it destroyed rather than sold for export.125 Israel later purchased Lightening Bug–style drones from the United States which it then used for surveillance and to patrol the Suez Canal zone.126 By 1978, Israel had created its first drone, the Mastiff, a small reconnaissance drone with a 13-foot wingspan that could be flown over battlefields and also work as a decoy to draw fire and identify weaknesses in another country’s defenses.127 By 1982, they had managed to get Syrian air units in the Beka’a Valley to fire on Israeli drones in mid-air and exhaust most of their SA-2 missiles, thus allowing Israel to strike against those units decisively. When Israeli forces invaded Lebanon in 1982, they also deployed drones in advance to spot Hezbollah forces hiding in civilian population centers.128 Notably, Israeli drones were pitched at a tactical level, and they were successful in drawing out information about the signals and emissions of particular missiles, which in turn allowed the Israelis to develop countermeasures.129
The US military took notice of Israel’s use of drones to adjust artillery fire after the bombing of the US Marines Corps’ barracks in Lebanon in 1983.130 Two US Navy captains were so impressed by the Mastiff’s ability to support artillery fire that they recommended that the United States buy Israeli drones. It was clear that Israel had come up with tactical drones that were more nimble and effective than the costly, high-altitude drones that the United States spent more than a decade developing.131 Although the United States has a strong preference against purchasing drones from foreign suppliers, the army and navy made an exception in this case and purchased a descendent of the Mastiff, the Pioneer drone, in 1986. The Pioneer is a medium-altitude (15,000 feet maximum) sleek gray drone with a 16-foot wingspan.132 Equipped with infrared cameras, the Pioneer was used for reconnaissance and aerial spotting of targets to be hit with conventional fire. This model not only showed the value of tactical drones against insurgents, terrorists, and other forces, it also revived interest in the technology at a point when it was collapsing due to failures with high-end drone prototypes.
The Pioneer drone became the first unmanned aircraft to be used in war since the Vietnam War. In Operation Desert Storm in 1991, some Pioneer drones helped to track Iraqi armored vehicles and provided artillery spotting for a US Marine artillery assault on a small Kuwaiti island held by Iraqi forces.133 In the latter instance, Iraqi soldiers surrendered to the drone, appearing on video at the Pentagon waving white flags in order to avoid yet another artillery bombing.134 Across the war, the Pioneer drones flew 330 missions, spending about 1,000 hours in the air.135 While they mainly flew in uncontested airspace, the Pioneer drones were able to provide limited, real-time video of events on the ground, leading many US military officials to declare them an unalloyed success. Pioneers were also used to confuse and disable Iraqi air defense systems at the beginning of the air campaign, but their actual tactical value was limited in most other respects.136
One of the reasons why drones so often failed is that they had low endurance and tended to crash at a high rate.137 Pioneer drones crashed regularly, with more than a dozen lost to mechanical failure or operator error.138 While the US military was experimenting with the Pioneer drone, an effort was underway elsewhere in government to build drones that could address these problems. Abraham Karem, an Israeli engineer and aviation enthusiast, was fascinated by the possibility of drones used as a decoy to reveal the capabilities of air defenses but also to defend land borders against invading armies.139 Having experimented with drones during his career at Israel Aerospace Industries (IAI), Karem left the company to start his own company, Leading Systems Incorporated, in 1974.140 After a few years of tinkering with drone models, and building rough prototypes in his garage, Karem proposed building a low-endurance drone to DARPA. While DARPA had been involved in drones in the 1960s, it stopped research and development for most unmanned aircraft for years, experimenting only with HALE drones that were kept aloft with cutting-edge motors driven by nuclear or solar power.141 Convinced of the value of Karem’s designs, DARPA funded his research between 1984 and 1990. The result was the Amber drone, which was medium-sized—15 feet long, with a wingspan of 28 feet—and had a liquid-powered piston engine. Made of lightweight, composite materials and equipped with a sophisticated hardware and software package,142 it was capable of flying for 38 hours or more—most of its competitors could manage no more than 12 hours.143 It was, in the words of a senior defense, official “a marvelous aircraft.”144
Yet the fact that the Amber drone had fans was not enough. It was an important technological advance on drone technology but it was operating in a political climate hostile to drones. In 1988, the Pentagon reorganized its drone research and development under a new bureaucratic entity, canceled the Aquila drone, and halved the research budget for drone technology.145 Recognizing that US funding was drying up, Karem developed a variant of the Amber, the Gnat 750, for foreign sales. Facing declining US interest, the Amber program was canceled in late 1988, and Leading Systems International drifted toward bankruptcy. Sensing an opportunity, General Atomics, under the leadership of Linden and Neil Blue, purchased Karem’s company and its surviving arsenal of Gnat 750 drones in 1991. The Gnat drones languished until 1993, when the CIA purchased a number of them for battlefield reconnaissance in Bosnia. During that brutal civil war, Serbia’s military forces laid siege to Sarajevo and shelled its citizens on a regular basis despite a “no fly zone” and UN presence. Frustrated with the inability of U-2s and satellites to get reliable imagery of the territory and position of forces due to cloud cover, the CIA deployed Gnat 750s to collect imagery of tanks, surface to air missile sites, and other targets of interest in Bosnia in 1994.146 The Gnats’ record was not wholly successful, as the video feed was shaky and subject to breaks and interruptions, and many planned missions had to be aborted because of bad weather, but it was sufficient for the military and CIA to be interested in an upgraded GNAT 750. This upgraded model was called the MQ-1 or, more popularly, the Predator.147
In many respects, the birth of the Predator can be thought of the birth of the drone age itself. Although the United States was also pouring money into building a high-altitude surveillance drone, the Global Hawk, it was the Predator that captured the attention of policymakers. More than just the most successful model since the Lightening Bug, the Predator functioned almost like a proof of concept of the capabilities and promise of drones. Even on its own terms, the Predator drone was a significant improvement over the Gnat 750 model. It was more capable of carrying a payload due to its longer wing span and fuselage; it also had a fuel capacity allowing for 24 hours of flying endurance.148 While this did not break the upper limit of the Gnat 750’s flight time, it was far more reliable and stable. It was also significantly quieter than the Gnat—once described by a journalist as a “lawnmower in the sky”—which had significant advantages for reconnaissance and support of combat troops.149 Perhaps more crucial was what was “under the hood”: the Predator had a large snub nose which contained integrated satellite communications that allowed communication via voice, digital imagery, and text messages with a ground station as opposed to a mothership.150 This allowed the aircraft to be flown and commanded from 400 miles away.151 Since it was connected to satellites, Predator drones could also now beam back images to ground stations and to the desks of officials in the Pentagon and intelligence agencies. CIA Director James Woolsey was even able to watch foot traffic over the damaged bridge in Mostar from his office in Langley, Virginia.152 The Predator functioned, for the first time, like a “television camera in the sky.”153
The Predator could only be used in this way because of a series of cascading technological innovations in the late 1990s. The information revolution had resulted in vast increases in processing power for computers and data imagery as well as satellite and other forms of electronic communication. In many respects, this was a natural result of the famous observation by Intel founder Gordon Moore that the number of transistors on a computer chip—and hence its processing power—doubles every two years while the cost of making the chip plummets. In other words, the growth of computer processing power is exponential. As Peter W. Singer has argued, this speed of change has been replicated in other forms of technology, including wireless capacity, optical storage, and internet bandwith.154 Although the speed of this change has recently slowed, leading some to wonder whether Moore’s law will continue to apply, the change in data collection and transmission capabilities transformed in a short period of time what drones like the Predator were capable of doing.155 The early evolution of the Predator capitalized on existing investments and encouraged the Pentagon to go even further with research and development into the transmission of data and imagery. It also encouraged other crucial adaptations, such as affixing a laser designator to spot and transmit the coordinates of a target to the nose of the drone.156
Among the most important developments was the rise of GPS as a navigation software. The earliest drones had to be within the line of sight of the operator, either on the ground or on an accompanying mothership, for the remote control to work. This almost always exposed the operator to some degree of risk, as the discussions of Operation Brass Ring showed clearly, and it imposed some natural limits on how far drones could fly. In 1983, the possibilities for drones expanded through a crisis and a little-heralded announcement by President Ronald Reagan. On September 1, 1983, a Soviet fighter plane shot down a Korean Airlines Flight 007 over the Sea of Japan. The plane had been traveling from New York to Seoul and among the 269 passengers and crew killed were 62 US citizens, including a member of the House of Representatives from Georgia. President Ronald Reagan was infuriated and weeks later announced that the US government would make the Navigation Signal Timing and Ranging Global Positioning System (NAVSTAR), later known simply as GPS, freely available to help civilian airliners avoid Soviet airspace.157 One of the earliest to notice the benefits of GPS for drone technology were Neil and Linden Blue. They came to the conclusion that GPS technology could be used to enable planes to fly unmanned as cruise missiles toward targets either against the Contras in Nicaragua or even potentially in a war with the Soviet Union in Europe. As Neil Blue remarked, “you could launch them from behind the line of sight so you would have total deniability.”158 Ironically, this conceptualization of drones echoed the ambitions of Elmer Sperry and others that unmanned aircraft could be guided along a set path to destroy a target. GPS made it possible to extend the distances involved and to envision a world in which drones run along waypoints set on routes hundreds, even thousands, of miles long. Although this was not well understood at the time, the linkage between drones and GPS gradually made it possible for unmanned aircraft to pinpoint and convey target locations around the world with remarkable accuracy.
Building off these technological developments, the Predator represented an important but imperfect advance on previous drone models. Its early record was mixed. Early in the mission in 1995, at least one Predator crashed and was recovered by Serbian forces. More Predators crashed due to weather and problems with de-icing. Early Predator drones had problems conveying images due to limitations of the UHF antenna or to the fact that the images were compressed for transmission, rendering them grainy.159 In Kosovo in 1999, Predator and US Army Hunter surveillance drones were deployed to collect imagery of Serb forces on the ground, to detect sound, and to collect electronic signals.160 One estimate suggests that between 1996 and 1999, only half of Predator operations in the Balkans were actually completed due to enemy fire or operator error.161 Even when they found the targets, Predator drones tended to struggle to destroy them because there was a costly time lag between spotting a target with a laser designator, conveying the information to an armed, manned aircraft, and firing a missile. One study found that only three of the nearly 300 Serb vehicles located by Predator drones had been destroyed due to this time lag.162 In a scathing internal report in 2001, the Pentagon’s director of operational tests and evaluation Thomas P. Christie wrote that the Predator was not “operationally effective” because of poor target location, accuracy, and limits to communication imposed by relatively benign weather such as rain.163
By early 2001, the Predator was at a crossroads. It was still criticized as ineffective, with some in the Pentagon convinced it was expensive and wasteful, especially given that manned aircraft and satellites could perform many of the same functions. But unlike previous drone models it had built an ardent constituency of military, intelligence, and other government officials who were addicted to the imagery coming from its operations. Enticed by the possibility of learning more about their enemy and the contours of the battlefield, high-ranking military and government officials began to push for even more Predator flights over war zones, and in some cases began to call pilots to make special requests for obtaining specific images.164 Some inside the military dismissed this as merely “Predator porn,”165 but others began imagining new possibilities for these drones. In 2000, General John Jumper began to lead an effort to arm Predators with Hellfire missiles.166 By mid-2001, the United States had conducted its first successful tests of an armed Predator drone in the desert in China Lake, California.167 As the September 11 attacks loomed, the United States was almost ready to begin arming Predator drones for the hunt for Osama bin Laden. It took only that attack for drones to be quickly converted into a weapon against terrorist organizations.
For decades, advocates of drones experimented with the technology and attached a number of purposes to them. Early drones were viewed as crude cruise missiles, target decoys, and reconnaissance crafts, sometimes at the same time. The goals for the technology grew larger over time; as capabilities improved, the calculation of the risks involved in deploying drone technology changed as well. By the end of the twentieth century, the technology behind unmanned aircraft had matured enough to realize the dream that US planners had at the time of Joseph P. Kennedy Jr.’s death. The Predator drone finally allowed the United States to fly and to strike long-distance enemy targets precisely and by remote control. What the planners could not anticipate at the time was how the remote control of unarmed and armed drones would dramatically transform warfare—and indeed the military itself—in the twenty-first century.