On the afternoon of October 7, 2001, the first day of the war in Afghanistan, an Air Force pilot named Scott Swanson made history while sitting in a captain's chair designed for an RV. His contribution to posterity was to kill someone in a completely novel way.
In the moments leading up to the act, Swanson was nervous. He sat in a darkened trailer tucked behind a parking garage at CIA headquarters in Langley, Virginia, remotely piloting a Predator drone over Kandahar, 6,900 miles away. Nearly everything about his rig had been cobbled together and hastily assembled. The Predator itself, one of just a handful in existence, was flying about 250 pounds heavier than usual. And the satellite communications link that connected Swanson to the aircraft would periodically shut down due to a power issue, which software engineers in California were frantically trying to patch.
When the order came through to take the shot, Swanson pulled a trigger on his joystick. A little more than a second later, a Hellfire missile slid off an aluminum rail on the Predator's wing and sailed into the Afghan night.
Swanson's target was a pickup truck parked outside a compound thought to be hiding Mullah Omar, the supreme commander of the Taliban. The missile killed two unidentified men believed to have been his bodyguards. It was the first time a US drone had fired a weapon in combat. It was the first time a modern drone had ever killed a human being.
Fourteen years later, the drone is the quintessential weapon of the American military, which now boasts roughly a thousand Predator pilots. At any given moment, scores of them sit in darkened trailers around the country, staring at the bright infrared camera feeds from drones that might be flying over Afghanistan, Iraq, Syria, Pakistan, or Somalia. Between August 2014 and August 2015, a single Predator squadron—the 432nd Air Expeditionary Wing in Nevada—flew 4,300 sorties and dropped 1,000 warheads on ISIS targets. By enabling the White House to intervene without committing troops to battle, the drone has transformed US foreign policy.
Indeed, the national security establishment's embrace of the drone has been so complete, it's tempting to assume that this new paradigm of warfare was something dreamed up long ago by senior officials, who methodically plotted their way to it over a span of years and a string of defense contracts. That is, after all, how we got other major weapons like the M1 Abrams tank, the Apache helicopter, and the F-35 Joint Strike Fighter.
But that's not how we got the modern drone. The Predator as we know it—with its capacity to be piloted from thousands of miles away and its complement of Hellfire missiles—wasn't developed with the expectation that entire wars might one day be fought by pilots sitting in trailers. As a matter of fact, most military planners at the time regarded the Predator as pretty much a technological dead end.
The tiny team of engineers and operators behind the program, who rarely speak publicly about their roles as the architects of remote warfare, worked under intense pressure, almost entirely free from the scrutiny of Pentagon acquisitions officers. In a series of breakthrough hacks, they hot-wired together the lethal, remotely piloted Predator over the course of just a few months in 2000 and 2001, in a mad dash to meet the heinous design challenges of a single job: to kill Osama bin Laden before he could commit an act of terror greater than al Qaeda's bombing of the USS Cole in 2000.
The lethal Predator wasn't a production vehicle. It was a hot rod, built for one all-out race against the clock. Of course, in those months before September 11, 2001, none of its designers knew the nature of the clock they were racing against. And most Americans have no idea quite how close they came to beating it.
America's first lethal drone pilot was obsessed with flying from an early age. Growing up in Minnetonka, Minnesota, he joined the Civil Air Patrol at 13, got his private pilot's license at 18, and enrolled in the Air Force ROTC program at the University of Minnesota just after graduating from high school. During the first Gulf War, he flew UH-1 Iroquois "Huey" helicopters. After Iraq, Swanson became a special operations pilot, focusing on sensitive and covert missions. Whenever he was at home base, he would volunteer to help test new Air Force weapons.
In 1997, Swanson was coming up on the end of a two-year mission in Iceland, some details of which remain classified. ("The Icelandic women were amazing" is about as much as he'll volunteer.) Contemplating his next move, he searched a database of Air Force duty openings and found a curious posting that asked for rated pilots to join the Eleventh Reconnaissance Squadron at Indian Springs Air Force Base, near Las Vegas. The two-year assignment was to fly the Air Force's newest aircraft, a little-known bird called the Predator Unmanned Aerial Vehicle.
An avid reader of Aviation Week, Swanson already knew a bit about the unmanned aircraft. Hand-built by a small, idiosyncratic California startup called General Atomics Aeronautical Systems, it had been used in the Balkans for surveillance since 1995. But it was not well loved by the defense establishment. The Predator was unarmed, couldn't fly in bad weather, and could only be operated within a 500-mile range of the pilot. In 1997, an evaluation by the Defense Department found that it suffered mechanical failures in a staggering 12 percent of missions.
To most Air Force pilots, the idea of operating a drone would be a nonstarter. Pilots fly in planes. But Swanson had always been interested in tinkering, technology, and experimental weapons. (As a teenager, he once used a homemade batch of cellulose nitrate to fire a projectile through the door of an abandoned car.) And as a special operations pilot, he grasped the Predator's surveillance capability right away. "It kind of clicked," he says.
So Swanson signed up with the Eleventh, and before the year was out he was in Taszár, Hungary, flying surveillance drones over Bosnia on a four-month deployment—the beginning of a years-long career with the Predator.
It was also in Taszár that the Predator caught the eye of another figure who would be crucial in its development, a senior Defense Department officer who was among the first to recognize the aircraft's potential. This past spring, I made my way to the Pentagon to meet him. (For security reasons, he declined to be named.)
Sitting in his windowless office with a short public affairs staffer and a very tall security officer, the official—whom I'll call Marshall—told me about that first time he saw the Predator in action in Hungary. "I was blown away," he says. "It flies at 70 miles an hour with a TV camera, but it can stay there forever." Marshall could see that it represented a strategic breakthrough comparable to that of the World War II codebreakers at Bletchley Park. From then on, he became a Predator evangelist, providing political cover and money when the project faced a roadblock. As I looked around Marshall's office, I noticed several bottles of a wine called Predator Old Vine Zinfandel sitting on a bookshelf.
In 1998, Marshall helped see to it that the Predator program was handed over to a tiny outfit within the military that would essentially improvise the genesis of modern drone warfare: an entity known as Big Safari.
A highly secretive Air Force skunkworks based in Dayton, Ohio, Big Safari specialized in modifying standard Air Force aircraft for time-sensitive and highly classified operations, sometimes even for use in just a single mission. In 1961, for instance, when Nikita Khrushchev boasted that he was about to test the largest hydrogen bomb ever built, Big Safari had just five days to retrofit a Boeing KC-135 to carry a small lab's worth of sensing equipment—shored up with two-by-fours—to snoop on the enormous detonation.
"We generally didn't do anything from scratch," says retired colonel Bill Grimes, Big Safari's director from 1985 to 2002. "We took existing hardware that was maybe for one purpose and adapted it to a completely different one for our needs." Like at a tech startup, Big Safari's teams were small and horizontal. Expediency, agility, and thrift were essential. "The most important thing was to get something useful to the war fighter quickly," Grimes says.
Big Safari set up its Predator office inside the General Atomics factory in San Diego, where the drone was made. And in the spring of 1999, during the Kosovo War, they got their first major chance to tinker with it. The Air Force came to Big Safari looking for a new way to steer laser-guided bombs dropped by jet fighters. US pilots wanted to stay above the range of Serbian antiaircraft fire, but their jets' laser designators—devices that beam pulses of light onto targets to guide missiles toward them—could not penetrate the region's heavy cloud cover. Big Safari's idea was to bolt a helicopter's laser designator onto a Predator. That way, the drone could stay below the clouds, in harm's way, and paint laser bull's-eyes on the ground for the jets high overhead.
In a typically lightning-fast turnaround, Big Safari had a modified Predator ready to be airlifted to the battlefield within 45 days. And its pilot—both in preliminary testing and on the ground in Kosovo—was none other than Scott Swanson.
Ordinarily, before a modified military aircraft is dispatched into combat, it has to pass through a lengthy vetting process that can take years. But Big Safari liked to deploy its creations before they were fully polished. The team referred to this as "the 80 percent solution" (because sometimes the last 20 percent of a job takes the longest). It was like releasing the beta version of a piece of software, says Brian Raduenz, then the commander of Big Safari's Predator detachment. "We would need to get it out there, get it into the hands of the guys doing the job, and then pay close attention to what they had to say about how it was working."
The rest of the Air Force was naturally allergic to this approach. At one point, authorities at Air Combat Command—an entity that had jurisdiction over all the Air Force's Predators—ordered Grimes to relay all his communication with Predator test pilots through the command's headquarters in Virginia. Unwilling to play a game of telephone, Grimes just gave Swanson a secure line so he could report back to Big Safari on the sly.
In Kosovo itself, Swanson participated in just one strike before the war ended. But by then, the pilot and his colleagues at Big Safari could tell they were onto something; a drone that could pinpoint targets was no joke. "We knew it was the future," Swanson says. And that future was about to come at them in a rush.
Big Safari's work on the Predator really took off when the group was enlisted in a high-stakes manhunt. In 1999, the CIA began to focus intently on Osama bin Laden, who had claimed responsibility for the US embassy bombings in Kenya and Tanzania in 1998. Intelligence reports indicated that bin Laden was planning further attacks. The agency wanted to put eyes on the al Qaeda leader and possibly target him, so it went looking for a covert way to get a high-powered camera over Afghanistan. The agency and the Pentagon considered several options, including a bizarre plan to mount a giant telescope on the side of a mountain. But after dispatching a group of officials in July 2000 to Indian Springs for a demonstration by Swanson, the CIA settled on the Predator.
First off, Big Safari had to figure out a way to sneak the Predator into Afghan airspace. Between maintenance crews, pilots, and field officers, it took several dozen people on the ground to sustain the operations of a single drone. According to Richard Whittle, whose book Predator authoritatively recounts the drone's history, the ground control station and satellite terminal were too large to conceal anywhere within a 500-mile radius of Kandahar. To make the operation truly covert, they would need to separate the drone from those controlling it by several thousand miles—by situating the command center at Ramstein Base in Germany. Ginger Wallace, an Air Force intelligence officer who was assigned to work on the project, thought the idea was ludicrous. "There's no way," she remembers thinking. "We can't really do that."
The guy who figured out how to do exactly that—how to wage war from thousands of miles away with a few clever modifications—was known among his colleagues in Big Safari as the Man With Two Brains, for his freakish intelligence. Without him, Grimes tells me, "it would not have happened."
An independent contractor who started working on the Predator in 1994, the Man With Two Brains almost never gives interviews. He spoke on condition of strict anonymity. At the beginning of our conversation, which I was allowed to record with a pen and paper only, I was scanned with a small black device, for a wire.
The basic premise of his remote control system, called split operations, was simple. A small, covert team of General Atomics contractors would post up at an airfield somewhere in a country bordering Afghanistan (the location of the site remains classified). There, they would launch the drone using a traditional line-of-sight remote control link. Once the drone was airborne, an onboard antenna would connect to a commercial satellite, which would relay the link to the ground control station hidden inside Ramstein Air Base, where Swanson, Wallace, and the rest of the operations team—working in secrecy—would control the drone as it scanned the desolate Afghan desert for the CIA's target.
True to Big Safari hacking tradition, the system did not require any significant new technology. But it did pose certain creative challenges. For instance, the plan required an antenna in Germany powerful enough to pick up a distant satellite signal—and the only option was a 36-foot "big-ass dish" located, appropriately enough, at Air Combat Command headquarters in Virginia. A team of contractors dismantled and made off with the satellite dish in a single night. By the time one of the lower-level staffers who managed the dish discovered it was gone and began circulating angry emails demanding its return, it was already en route to Germany, as were Swanson, some General Atomics contractors, and a joint CIA and Air Force operations team.
The team found what they were looking for during one of the Predator's very first split operations missions in early September 2000. Swanson was circling over Tarnak Farms, a walled compound near the Kandahar airport where bin Laden—or UBL as the team called him, referring to the alternative spelling, Usama—was thought to be living. Jeff Guay, an Air Force master sergeant on the team, was controlling the drone's camera. Sure enough, a man in white, surrounded by an entourage, soon emerged on their screens.
"When UBL walked out of that one building," Swanson says, "the way he appeared much taller than everybody, the people were deferential around him, the way he was dressed, Jeff and I just looked at each other and it's like, 'Yeah, that's got to be him.'" Swanson assumed a cruise missile would be dispatched in the direction of bin Laden while the Predator loitered overhead to make sure he stayed put. The team had been instructed to continue circling for as long as necessary, even if that meant running out of fuel and crashing.
But for reasons obscure to the team, no strike was ordered. With Swanson gripping his joystick, unable to do anything but stare, America's final chance to kill Bin Laden before September 11 slipped away.
It was clear: If the Predator had been armed, Swanson could have done the killing himself. And sure enough, the flight over Tarnak Farms kicked into high gear a project that had been quietly under way for months. Air Combat Command had decided to look into arming the Predator in 1999. When they put Big Safari on the case, Grimes convened a gathering of engineers and weapons specialists at the Big Safari office in Dayton. On the first day of what would become a two-day meeting, he noticed that some of the engineers were laughing at the proposition of mounting a missile on a motorized glider. "We identified those who were leaning forward, who felt like this could be done, and I privately got ahold of them and invited them the next day," he says. "The remainder were totally unaware of the second meeting."
Grimes and his team briefly considered packing the Predator with explosives and flying it directly into its targets, but a projectile that chugged along at highway speed was too slow to reliably surprise anyone. Big Safari needed a weapon that was small enough to fit on the Predator's delicate wings but powerful and precise enough to destroy a car or a person from high in the air.
Eventually they settled on the Hellfire, the Army's low-altitude, laser-guided helicopter missile. But the technical challenges of taking an antitank weapon designed to be fired from no higher than 2,000 feet and converting it into an antipersonnel missile that would be shot from above 10,000 feet were considerable. Among other things, the Predator would need a new forward-looking infrared camera, the team would need to recode the guidance systems on each missile, and someone was going to have to figure out how to give an armor-piercing munition the kind of grenade-like, shrapnel-spewing blast that would be effective at killing humans. "Even with a Big Safari mentality, that's a big-ass project to get done," Swanson says.
The team certainly didn't lack for motivation, however. In October 2000, just six weeks after the crew had first set eyes on bin Laden at Tarnak Farms, al Qaeda carried out its attack on the USS Cole in Yemen, killing 17 sailors. Big Safari had already proven it was possible to get a Predator within striking distance of the al Qaeda leader. Now their goal was to get a shot at him with an armed Predator before the next winter. When the CIA approved the idea of a lethal Predator and put its weight behind the program, the project went into overdrive. "You could see and tell with the energy of the team that you had a real no-shit goal with this," Swanson says. "We're gonna arm this thing and go hunting."
The Hellfire program's deadline was set for September 1, 2001. And everything was on schedule until a new roadblock shot up—a political problem that would inspire Big Safari's most historically significant technological hack.
In the summer of 2001, the German government decided that it would not permit the US to operate its newly armed Predators from Ramstein. So the CIA's deputy counterterrorism chief convened a briefing to announce that the effort to deploy an armed Predator to Afghanistan in search of bin Laden would be tabled until they could figure out somewhere else to base their operation. When the floor was opened for questions, the Man With Two Brains says, he raised his hand. He had an idea.
For years he had told Grimes and others at Big Safari that it would be technically feasible to operate Predator drones around the globe from within the US. He called his concept remote split operations. Now he realized that such a system wouldn't just make deployments easier, it would solve the agency's legal conundrum. The idea was to use the military's existing fiber-optic network to put 4,000 miles between the drone pilot, who would now be in the US—unaffected by Germany's laws—and the big-ass satellite dish, which would still be located at Ramstein.
This time, the challenges were technical: The Man With Two Brains had to find a way to package the various kinds of data traveling between the drone and the operators—flight commands in one direction, data from the camera and the drone's other sensors in the other—and shuttle them across the Atlantic without creating a lot of lag time.
To package the data, the Man With Two Brains turned to something called a multiplexer, a fairly cheap commercial device that Internet companies were using to bundle various kinds of files, like MPEGs, into fiber-optic-friendly packets for streaming. He split an encoding device in the Predator's existing satellite link system in two and placed each half on either end of the military's 4,000-mile undersea fiber-optic cable. Discarding a modem, he installed in its place two multiplexers, which encapsulated the data traveling in both directions.
Figuring out how to minimize lag time, or latency, was an equally devilish challenge. Anyone who has tried to have a conversation over a laggy Skype connection will have a sense of the problem. CIA and Air Force engineers had run their own calculations and determined that the new system would increase the total latency to five seconds, too much to safely operate a weapons system.
The Man With Two Brains wanted three months to complete the entire project; he got six weeks. Working in a lab in Washington, DC, the contractor spliced fiber-optic cables, soldered switchboards, and created a variety of loops and circuits that more resembled a Rube Goldberg machine than a device that would enable the killing of a human being from half a world away.
Once he completed the hack—on schedule—he traveled to Southern California to see how it would work on a real Predator. On the first day of flights, the remote split link passed various stress tests. The Man With Two Brains planned to conduct further tests the next day—which was September 11, 2001.
Time slowed down for millions of Americans that morning, but for the Predator team it sped up. The remote split system was promptly approved for operation by the CIA; in Alabama, a batch of Big Safari's modified Hellfires was loaded into a transport pallet bound for Afghanistan; and Scott Swanson packed his bags for Langley, where the team's ground control station—the darkened trailer—was waiting by the CIA parking garage. Major Mark Cooter, the operation's director, started making calls to the other members of the group that had flown the Predator over Tarnak Farms, telling them it was time to get the band back together.
CIA electrical engineers had set up a sleek-looking control console in the team's trailer at Langley, but it didn't work, so Cooter instructed his own team to rip it out and replace it with a more functional setup, partly held together with zip ties and Velcro. On September 17, the team fired up the remote split link for the first time. Everyone watched as Swanson moved his joystick; 1.3 seconds later, the Predator responded. The addition of 4,000 miles of fiber-optic cable had increased latency by only 200 milliseconds, round-trip.
The 80 percent solution wasn't perfect, of course. "There were glitches," Swanson says. The data link, which shared a satellite with several cable television networks, would drop out unexpectedly, as happened on October 7, when a white-knuckled Swanson killed Mullah Omar's two presumed bodyguards. During another strike, the Predator's communications system went into a reboot at the precise moment Swanson launched one of its Hellfires at a radar site. In the first three months of the war, the team lost at least two Predators due to malfunctions. "Oh yeah, it was a duct tape war," Marshall says.
Bit by bit, the team racked up some early successes, including a strike that killed Mohammed Atef, the military commander of al Qaeda. Word of their exploits spread, and elite forces on the ground would specifically request air support from the Predator team, which had been given the codename Wildfire. The team framed a copy of a 2000 Defense Department report that had declared the Predator a failure and hung it on a wall next to a list of what Marshall called their greatest hits. That October, an unnamed official said in a Pentagon briefing that theater commanders were "begging for more Predators." In a December 2001 speech, then-president George W. Bush singled out the Predator as a harbinger of the military's future. "It is clear the military does not have enough unmanned vehicles," he said. In the space of three years, Big Safari had transformed an albatross on the verge of extinction into a lethal bird that was now being hailed as the chief weapon in the War on Terror.
And they kept hacking. Using consumer electronics, the Man With Two Brains figured out how to transmit the Predator's live feed to AC-130 gunships and, later, to ground forces. The team even figured out how to channel a cable television feed into the Predator's video dissemination system, essentially turning the drone into a flying TV antenna: That way, forward-operating Special Ops teams could watch NFL games and movies during their downtime.
Swanson, who now works as a consultant out of Antigua, Guatemala, has a stiff crop of red hair that turns to gray at the sideburns. His eyes are small and intense, and he chooses his words with the concentration of someone who knows a lot of secrets.
Today the Big Safari team members don't have much to do with the Predator. They're mainly retired or doing other things, while the national security establishment that once disparaged the drone has thoroughly embraced it. The Predator has ushered in a more precise era of warfare. It has also inspired new kinds of nightmares for those who live under drones—and those who fly them.
In the summer, Swanson Skypes me from Antigua. During those first missions, he says, he was struck by the intimacy of this new form of warfare. "You're watching these people coming and going," he says. "You're watching them go out and take dumps or pees in the middle of the night.
"I'm not saying you ever really bond with the target," he goes on. But you dwell on them for dramatically longer than with any other weapons system, he says. His pauses begin to draw out.
I ask how it feels to have participated in the creation of the Predator. He mentions a recent drone strike that killed Nasir al-Wuhayshi, al Qaeda's second-in-command. "I feel proud to have been part of the team that brought that forward," he says.
What about when a strike misses its target or is used for ill? That has less to do with what the Predator can and cannot do, he says. "That is just the ugly nature of war. And yeah, there's always a little twinge of regret with that." Swanson pauses again. "The world is not black-and-white," he says. "It's shades of gray presented to you in an infrared image."
Arthur Holland Michel (@writearthur) is codirector of the Center for the Study of the Drone at Bard College.