The Pentagon's Brain
“The nature of war is changing,” Scales wrote that same fall in Proceedings magazine. “Fanatics and fundamentalists in the Middle East have adapted and adopted a method of war that seeks to offset U.S. technical superiority with a countervailing method that uses guile, subterfuge and terror mixed with patience and a willingness to die.” Scales warned that this new kind of warfare would allow the weaker force, the insurgents in Afghanistan and Iraq, to take on the stronger force, the United States, and win. Since the Israeli War of Independence, Scales wrote, “Islamic armies are 0 and 7 when fighting Western style and 5 and 0 when fighting unconventionally against Israel, the United States, and the Soviet Union.”
The Pentagon moved forward with DARPA’s idea to bring anthropologists into the Iraq war, and McFate garnered exclusive permission to interview Marines coming home from Iraq. In July 2005 she authored a paper in Joint Force Quarterly, a magazine funded by the Department of Defense, titled “The Military Utility of Understanding Adversary Culture.” In it she stated clearly her opinion about what had gone wrong in Iraq. “When the U.S. cut off the hydra’s Ba’thist head, power reverted to its most basic and stable form—the tribe,” wrote McFate. “Once the Sunni Ba’thists lost their prestigious jobs, were humiliated in the conflict, and got frozen out through de-Ba’thification, the tribal network became the backbone of the insurgency.” As an anthropologist, McFate believed that “the tribal insurgency is a direct result of our misunderstanding the Iraqi culture.”
Soldiers in the field had information, McFate said, but it was the wrong information. “Soldiers and Marines were unable to establish one-to-one relationships with Iraqis, which are key to both intelligence collection and winning hearts and minds.” McFate issued a stern warning to her Pentagon colleagues: “Failure to understand culture would endanger troops and civilians at a tactical level. Although it may not seem like a priority when bullets are flying, cultural ignorance can kill.”
McFate was hired to perform a data analysis of eighty-eight tribes and sub-tribes from a particular province in Iraq, and the behavioral science program she was proposing began to have legs. At DARPA, Bob Popp was enthusiastic. “It was not a panacea,” he says, “but we needed nation rebuilding. The social science community had tremendous insights into [the] serious problems going on [there], and a sector of DoD was ready to make serious investments into social sciences,” he says of DARPA’s efforts.
Arthur Cebrowski died of cancer the following year. The Office of Force Transformation did not last long without him and within a year after his death closed down, but the social intelligence programs forged ahead. Montgomery McFate found a new advocate in General David Petraeus, commander of the Multi-National Security Transition Command, Iraq, who shared her vision about the importance of winning hearts and minds. Petraeus began talking about “stability operations” and using the phrase “culture-centric warfare” when talking to the press. He said that understanding people was likely to become more important in future battles than “shock and awe and network-centric warfare.”
The DARPA program originally conceived broadly by Bob Popp to bring social scientists and anthropologists into the war effort was fielded to the U.S. Army. Montgomery McFate became the lead social scientist in charge of this new program, now called the Human Terrain System. But what did that mean? The program’s stated mission was to “counter the threat of the improvised explosive device,” which seemed strangely at odds with a hearts and minds campaign. Historically, the battle for hearts and minds focused on people who were not yet committed to the enemy’s ideology. The Army’s mission statement made the Human Terrain System sound as if its social scientists were going to be persuading terrorists not to strap on the suicide vest or bury the roadside bomb after all. The first year’s budget was $31 million, and by 2014, the Pentagon would spend half a billion dollars on the program. Unlike in ARPA’s Motivation and Morale program during the Vietnam War, the social scientists who were part of the Human Terrain System program during the war on terror would deploy into the war zone for tours of six to nine months, embedded with combat brigades and dressed in full battle gear. Many would carry guns. So many elements of the program were incongruous, it was easy to wonder what the intent actually was.
“I do not want to get anybody killed,” McFate told the New Yorker. “I see there could be misuse. But I just can’t stand to sit back and watch these mistakes happen over and over as people get killed, and do nothing.” Major General Robert Scales, the keynote speaker at the DARPA counterinsurgency conference organized by McFate, wrote papers and testified before Congress in support of this new hearts and minds effort in Iraq and Afghanistan. In the Armed Forces Journal Scales wrote, “Understanding and empathy will be important weapons of war.” Then he made a bold declaration. “World War I was a chemists’ war,” Scales said. “World War II was a physicists’ war,” and the war on terror was “the social scientists’ war.”
The program quickly gathered momentum. The Human Terrain System was a countermeasure against IEDs, and counterinsurgency was back in U.S. Army nomenclature. In December 2006 the Army released its first counterinsurgency manual in more than twenty years, Counterinsurgency, Field Manual, No. 3-24. Lieutenant General David Petraeus oversaw the manual’s publication. Montgomery McFate wrote one of the chapters. “What is Counterinsurgency?” the manual asks its readers. “If you have not studied counterinsurgency theory, here it is in a nutshell: Counterinsurgency is a competition with the insurgent for the right to win the hearts, minds, and acquiescence of the population.” As it had done in Vietnam, the COIN manual stressed nation-building and cultural understanding as key tactics in winning a guerrilla war.
It was as if the Vietnam War had produced amnesia instead of experience. On its official website, the U.S. Army erroneously identified the new Human Terrain System program as being “the first time that social science research, analysis, and advising has been done systematically, on a large scale, and at the operational level” in a war.
CHAPTER TWENTY-TWO
Combat Zones That See
For the Pentagon, trying to fight a war in an urban center was like fighting blind. From the chaotic marketplace to the maze of streets, there was no way of knowing who the enemy was. DARPA believed that superior technology could give soldiers not just sight but omnipotence. Their new effort was to create “Combat Zones That See.”
In the second year of the Iraq war, DARPA launched its Urban Operations Program, the largest and most expensive of the twenty-first century, as of 2014. “No technological challenges are more immediate, or more important for the future, than those posed by urban warfare,” DARPA’s deputy director, Dr. Robert Leheny, told a group of defense contractors, scientists, and engineers in 2005. “What we are seeing today [in Iraq] is the future of warfare.” While the short-term priority remained the IED, the long-term solution required a larger vision. It was less about locating the bombs than about finding the bomb makers, Tony Tether told Congress in 2005. With Vietnam came the birth of the electronic fence, with a goal of sensing and hearing what was happening on the Ho Chi Minh Trail. With Iraq came the birth of the electronic battle space, with eyes and ears everywhere—on the ground, in the air, behind doorways and walls. DARPA needed to bolster its research and development programs to produce wide-scale surveillance technology for urban combat zones—total surveillance of an area wherever and whenever it was needed. This was the plan for Combat Zones That See.
“We need a network, or web, of sensors to better map a city and the activities in it, including inside buildings, to sort adversaries and their equipment from civilians and their equipment, including in crowds, and to spot snipers, suicide bombers, or IEDs,” Tether told the Senate Armed Services Committee. “We need to watch a great variety of things, activities, and people over a wide area and have great resolution available when we need it.” Through information technology the United States could gain the upper hand against the terrorists in Iraq and places like it. “A
nd this is not just a matter of more and better sensors,” he explained, “but just as important, the systems needed to make actionable intelligence out of all the data.” Director Tether requested half a billion dollars to fund the first phase of development.
The timing was right. Congress had eliminated funding for DARPA’s Total Information Awareness programs in the fall of 2003, citing privacy concerns. But Iraq was a “foreign battle space.” Civil liberties were not at issue in a war zone. “Closely related to this [network of sensors] are tagging, tracking, and locating (TT&L) systems that help us watch and track a particular person or object of interest,” said Tether. “These systems will also help us detect the clandestine production or possession of weapons of mass destruction in overseas urban areas.”
DARPA partnered with the National Geospatial-Intelligence Agency (NGA), a dual combat support and intelligence agency that had been drawing and analyzing military maps since 1939. With the invention of the satellite, NGA became the lead agency responsible for collecting “geospatial intelligence,” or GEOINT, interpreting that intelligence, and distributing its findings to other agencies. The NGA remains one of the lesser-known intelligence agencies. The majority of its operations are born classified.
In Iraq, DARPA and the NGA worked together to create high-resolution three-dimensional maps of most major cities and suspected terrorist hideouts. The mapping efforts became part of a system of systems, folded into a DARPA program called Heterogeneous Urban Reconnaissance, Surveillance and Target Acquisition, or HURT. Entire foreign civilian populations and their living spaces would be surveyed, observed, and scrutinized by the U.S. military and American allies so that individual people—insurgents—could be targeted, then captured or killed. In urban warfare situations, DARPA knew, terrorists tried to blend in among heterogeneous crowds, much as the Vietcong had done with trees on the trail. DARPA’S HURT program was technology designed to deprive terrorists of people cover.
To implement the terrain-based elements of the HURT program, hundreds, perhaps thousands, of defense contractors were dispatched to Iraq, capturing digital imagery along at least five thousand miles of streets using techniques similar to those used for Google Maps. Many details of the program remain classified, including which cities were targeted and in what order, but from Tether’s own testimony Congress learned that thousands of tiny surveillance cameras and other microsensing devices had been discreetly mounted on infrastructure, designed to work like England’s CCTV system. Tether described these surveillance cameras to Congress as “a network of nonintrusive microsensors.” Unclassified documents from the NGA described these sensors as including low-resolution video sensors placed close to the ground to monitor foot traffic; medium-resolution video sensors placed high on telephone poles to watch motor vehicle and pedestrian streams, and high-resolution video sensors placed at an opportune height to capture “skeletal features and anthropometric [body measurement] cues.” The resulting three-dimensional maps laid the groundwork for the first of many Combat Zones That See. DARPA program managers joked that their goal was “to track everything that moves.”
One of the drones in the HURT program was the Wasp, a tiny unmanned aerial vehicle with a fourteen-inch wingspan and weighing only 430 grams, or less than a pound. Providing real-time overhead surveillance to soldiers on the ground, a fleet of Wasps took to the airspace over Iraqi cities and supply routes. The Wasp was one of the smartest drones in the drone fleet in 2005. Powered by batteries, it flew low and carried an exceptional payload of technology packed inside, including a color video camera, altimeter, GPS, and autopilot. The Wasps worked together in the system of systems, bird-sized drones flying in pairs and in threes.
“The [HURT] system can get reconnaissance imagery that high-altitude systems can not,” says Dr. Michael A. Pagels, a HURT program manager who oversaw field operations in Iraq. “It can see around and sometimes into buildings.” Because of the Wasp’s micro size, some could enter into buildings undetected, through open windows and doorways, then fly around inside. The drones’ capabilities were tailored for specific urban combat needs. If two of the Wasps were taking surveillance photos of the same area, their advanced software was able to merge the best of both images in a “paintbrush-like effect,” updating the images captured in near real time, then sending them to small computers carried by soldiers on the ground. At a soldier’s behest, the HURT system could pause, rewind, and play back the Wasp’s surveillance video. This was a key feature if a soldier was hunting a terror cell planting an IED and needed to know what an area looked like three minutes, or three hours, before. The HURT system even had several self-governing features. It knew when one of its drones was low on fuel and could coordinate refueling times to ensure that surveillance was maintained by other drones in the system. The Wasp was also designed to recognize when it was running low on battery power. It could transmit its status to an operator. “HURT is designed to be agnostic,” Pagels says, meaning that if one part of the system goes down, the other parts of the system quickly adapt to compensate for the loss. Mindful of what DARPA called the “chaotic fog of war and the mind-numbing complexity of the urban environment,” the system’s creators aimed to achieve “Persistent Area Dominance.” HURT was part of that domination. With HURT, humans and machines would work together to maintain situational awareness in dangerous urban environments.
Giant unmanned blimps were also involved in surveillance, in DARPA’s Tactical Aerostat program, also called the “unblinking eye.” Originally designed for U.S. border patrol surveillance, these forty-five-foot-long airships were tethered to mobile launching platforms by reinforced fiber-optic cable. The moored balloons were then raised to heights of between one thousand and three thousand feet. They were designed to be compact and portable, able to go up and down before insurgents could shoot them out of the sky. Fiber optics allowed for secure communication between the classified surveillance systems carried inside the blimps and the operators on the ground. The blimps were helpful for keeping watch over increasingly dangerous roads, like Main Supply Route Tampa, a fifteen-mile stretch of road out of Baghdad, and Route Irish, the deadly road to the Baghdad International Airport.
Unclassified DARPA literature reveals that sometimes the system of systems worked. Other times, elements failed. Sandstorms made visibility difficult, and when that happened, terrorists could sneak in and plant their IEDs under cover of weather. When the sandstorm cleared, it was often impossible to distinguish windblown trash from newly planted bombs. Several of the blimps and drones also either were shot down or crashed on their own.
But DARPA’s defense contractors and scientists back home persevered. The system of systems being built by DARPA was long term, and had ambitious, well-funded goals. The ultimate objective for Combat Zones That See was to be able to track millions of people and cars as they moved through urban centers, not just in Iraq but in other urban areas that potentially posed a threat. Cars would be tracked by their license plates. Human faces would be tracked through facial recognition software. The supercomputers at the heart of the system would process all this information, using “intelligent computer algorithms [to] determine what is normal and what is not,” just as the Total Information Awareness office proposed. Combat Zones That See was similar to TIA’s needle-in-a-haystack hunt. It was bigger, bolder, and far more invasive. But would it work?
In Combat Zones That See, DARPA’s goal was for artificially intelligent computers to process what it called “forensic information.” Computers could provide answers to questions like “Where did that vehicle come from? How did it get here?” In this manner, the computers could discover “links between places, subjects and times of activities.” Then, with predictive modeling capabilities in place, the artificially intelligent computers would eventually be able to “alert operators to potential force protection risks and hostile situations.” In other words, the computers would be able to detect non-normal situations, and to notify the humans in the system of syst
ems as to which hostile individuals might be planning an IED or other terrorist attack.
In the winter of 2005, the Washington Post reported that an IED attack occurred inside Iraq every forty-eight minutes. The primary countermeasure was still the electronic jamming device, designed to thwart IED activation by remote control. But these jammers were doing only a little good. In Iraq, coalition forces were up against an electromagnetic environment that was totally unpredictable and impossible to control. Iraq had an estimated 27 million people using unregulated cell phones, cordless phones, walkie-talkies, and satellite phones, and DARPA jammers were failing to keep up. Jammers were even getting jammed: Al Qaeda bomb makers developed a rudimentary radio-controlled jamming signal decoder that the Americans called the “spider.” The U.S. military appeared to be losing control. Despite DARPA’s lofty goals of Persistent Area Dominance through battle space surveillance, in reality the Combat Zones That See concept was collecting lots of information but providing little dominance.
DARPA had dozens of potential solutions in various stages of development. The Stealthy Insect Sensor Project, at Los Alamos National Laboratories, was now ready to deploy. As part of the animal sentinel program, going back to 1999, scientists had been making great progress training honeybees to locate bombs. Bees have sensing capabilities that outperform the dog’s nose by a trillion parts per second. Using Pavlovian techniques, scientists cooled down groups of bees in a refrigerator, then strapped them into tiny boxes using masking tape, leaving their heads, and most of their antennae, poking out the top. Using a sugar water reward system, the scientists trained the bees to use their tongues to “sniff out” explosives, resulting in a reaction the scientists call a “purr.” After training, when the scientists exposed the bees to a six-second burst of explosives, some had learned to “purr.”