(Bloomberg Businessweek) -- Tony Sedgwick steers his red Nissan pickup to the edge of his vast Arizona ranch in the Sonoran Desert, unlocks a cattle gate, and continues rattling south along a dirt road until he reaches the U.S.-Mexico border. He climbs out of his truck and follows the undulating line of towering vertical steel beams as the ground slopes down into a dry riverbed. Here, the beams give way to crisscrossing shoulder-height iron bars. The white-haired cowboy removes his hat, hikes up his Wranglers, scissors over one iron bar, and ducks under the next. “I mean, I’m a 66-year-old man, and I have no trouble going through this fence,” he gripes. “You can see the senselessness of this.”
After ambling around for a half-hour, Sedgwick spots a white-and-green U.S. Customs and Border Protection (CBP) truck bumping up and down in a cloud of dust. It slows, and Sedgwick tips his hat and waves. The agents nod and drive on. He explains that we probably tripped one of the hundreds of sensors buried in secret locations under his pebble-specked ranch. Sedgwick points to another possibility on a hilltop about a mile away. “You see that little tower there?” he asks. A slender latticed edifice pokes into the blue sky, radar antennas and cameras affixed to the top.
This is the southwest border as it exists today. It’s an open wound to President Trump and a significant contingent of his supporters, one that was freshly salted on March 23. The $1.3 trillion budget passed by Congress that day didn’t include the $18 billion Trump’s administration requested for a concrete wall. Instead, it provides $696 million to replace old fences and $641 million to build new ones in areas where there aren’t any currently—and solid concrete is prohibited.
In response to the budget defeat, Trump announced angrily that he’ll deploy the National Guard to police the border. He’s followed up by signing a presidential memorandum to authorize the deployment of as many as 4,000 soldiers. Strictly from a geographical perspective, it’s obvious why the dividing line between Mexico and the U.S. would drive a nativist a little mad. It occupies 1,954 miles of desert, mountains, cities, and valleys from the Pacific Ocean to the Gulf of Mexico and features a hodgepodge of iron bars, barbed wire, concrete blocks, sand levies, stone obelisks, rotting piles of logs, and plenty of wide-open land. An army of about 20,000 Border Patrol agents guards its length on foot and horseback, by all-terrain vehicle and truck, day and night, 365 days a year.
Layered unevenly into this tangle is an electronic perimeter of surveillance towers, 12,000 underground motion sensors, cameras that can spot a jack rabbit hopping through cactuses miles away, long-range radars mounted on distant towers, and Predator B drones whose advanced radar can detect footprints in the sand. Long before Trump signed off on the budget, those closest to him, including his chief of staff, John Kelly, were steering him away from his “big, beautiful wall.” There’s bipartisan interest in shifting toward more electronic surveillance. Democrats see technology as the wall’s lesser evil, despite concerns for the civil liberties of people living within its range. Many Republicans in Congress, especially those from border states, say electronic surveillance promises greater security and, at least in theory, is more cost-effective. The new budget is a win for the tech believers, allotting about $400 million for border technology, including about $50 million for new towers and $20 million for more ground sensors. “Buy more surveillance equipment!” isn’t a catchy rally chant, but it looks like reality for now.
Contractors and consultants have been positioning themselves to cash in. Peter Thiel, the PayPal Holdings Inc. billionaire, has raised money for a virtual wall startup founded by Oculus VR creator Palmer Luckey. Called Anduril Industries, after a powerful sword in The Lord of the Rings, the company says on its website that it’s hiring engineers. Anduril has already paid Heather Podesta and her lobbying company, Invariant, $80,000 to keep tabs on border security funding. The U.S. Department of Homeland Security (DHS), which opened a small satellite office in Silicon Valley in 2015, recently solicited proposals for a miniature facial-recognition drone that could be operated by a touchscreen embedded in the sleeve of a border agent’s uniform, as well as for energy-harvesting fabrics and 3D mapping technology. Military contractors whose technology was designed for overseas conflicts are investigating how their wares might be deployed along the border. Already, aerostats (a kind of tethered blimp) used to guard forward operating bases in Afghanistan are watching remote sections of desert, and wheeled “MARCbots,” tested on the battlefield in Iraq, are scouring smuggler tunnels.
Representative Will Hurd, a Republican whose district encompasses 800 miles of the Texas border, says sensor technology has come so far and gotten so cheap that the border should be blanketed with cameras, radar, and fiber-optic cable. “That is not a Star Trek scenario,” he says. “That’s something we should be able to do today.”
Yet for all its promise, surveillance technology has become a Bermuda Triangle for border security. The government has devoted a half-century and billions of dollars to creating a virtual wall, but political leaders, America’s biggest companies, and laboratories filled with rocket scientists have failed to deliver one that works.
As early as the 1940s government agents introduced a fleet of radio-equipped gyrocopters and built a network of radio transmitters and observation towers to fortify the barbed wire fences built to keep out cattle, immigrants, and bootleggers. But the first major technological experiment in border surveillance grew out of the Vietnam War. In 1970, a Department of Defense engineer traveled to San Diego to see if the seismic and magnetic sensors the Army was deploying to track Viet Cong along the McNamara Line could identify migrants on the border. The government ultimately installed 177 ground sensors, says Iván Chaar-López, whose dissertation at the University of Michigan at Ann Arbor focuses on technology along the border. Over the next half-century, officials came to envision what Chaar-López calls a “system of systems,” in which physical barriers, patrol agents, and technology work in sync.
At one point the government deployed a gleaming new Ford Bronco outfitted with electronic surveillance equipment that bounced along stretches of the border. It stood out to the very smugglers the U.S. was trying to bust. One morning agents emerged from their motel room to find the Bronco missing. Soon they tracked it down—sunk in the middle of the Rio Grande. “The message was clear,” retired U.S. Air Force Colonel Bill Grimes recalls in his book, The History of Big Safari, about secret surveillance missions. “We know who you are and what you are doing.”
By the mid-1980s, with oil prices tanking and the Mexican economy sinking into recession, more than a million people a year were caught along the border. In 1993 the U.S. constructed fences—some made of military-surplus steel used for makeshift runways—along 14 miles near San Diego, the busiest point of entry. The following year, Congress, in a rare bipartisan moment, funded Operation Gatekeeper to pay for additional border agents. “We are a nation of immigrants, but we are also a nation of laws,” President Clinton said to a standing ovation in his 1995 State of the Union address. “We must do more to stop them.” Agents got night scopes to see in the dark, seismic sensors, and an electronic fingerprinting system.
The government also considered dropping invisible dye on immigrants from helicopters to track them and blaring the sound of barking dogs over loudspeakers. Immigration officials asked Sandia National Laboratories, one of the nation’s foremost defense research labs, to recommend other strategies. Officials sporadically plucked advice from the resulting three-volume, 695-page report, not heeding, for instance, the recommendation against investing in expensive advanced technologies that in Sandia’s analysis typically “did not provide benefit commensurate with cost.”
With the new fences and technology, migrant crossings did plummet in San Diego; instead, more people risked crossing in the deserts of Arizona. According to the Migration Policy Institute: “The Tucson morgue recorded an average of 18 migration-related deaths per year in the 1990s, while in the 2000s it saw almost 200 per year.” Next, Congress plowed almost a half-billion dollars into a series of ever more ambitious surveillance programs, such as the America’s Shield Initiative and the Integrated Surveillance Intelligence System, known by the unfortunate acronym ISIS. These efforts were plagued by technical problems. Some cameras couldn’t pan and hold steady, while insects chewed through components. About 90 percent of ground sensor alerts were false alarms.
After Sept. 11, President Bush marshaled yet another push. The bipartisan Secure Fence Act of 2006 directed the DHS, which had been formed shortly after the attacks, to erect 700 miles of double-layered fencing with room for patrol cars to drive between. (Congress later scaled back the plan.) And Bush introduced the Secure Border Initiative, a multibillion-dollar program to create a bespoke virtual fence of 1,800 towers equipped with cameras and sensors lining the entire 6,000 miles of borders with Mexico and Canada, which would relay data to a central location in Washington, D.C. “We’re launching the most technologically advanced border security initiative in American history,” Bush said in an Oval Office address.
Boeing Co. won the contract, promising to detect 95 percent of illegal border crossings. Almost immediately, the project fell behind schedule and went over budget. Worse, it barely worked—sensors confused raindrops or leaves blown in the wind for people, an official from the U.S. Government Accountability Office told 60 Minutes. During a congressional hearing, Joseph Lieberman, the Independent senator from Connecticut, said he was frustrated by the government’s inability to “find that mystical point where parallel lines finally meet. It’s always just over the horizon, but you never actually get there.”
More than $1 billion later, in 2011, Homeland Security Secretary Janet Napolitano canceled the program. The technology ended up covering only 53 miles along the Arizona border. The government is decommissioning the towers and in the meantime has paid roughly an additional $200 million to maintain the program. Boeing has received almost 40 percent of that.
It took an economic shift to accomplish what the virtual fence could not. The Great Recession saw out-of-work migrants return home, and an improving Mexican job market has kept them there. From 2007, when the recession started, to 2011, the number of apprehensions along the border fell more than 60 percent. Last year fewer people crossed than at any point since the early 1970s.
For Trump, these statistics are no reason to claim success in sealing out the “drastic illegal activity” he says is putting the border in crisis. Hard-liners look to one country, and one country only, as a model of success: Israel. Making a pilgrimage to Jerusalem has become a rite of passage for lawmakers interested in border security.
In 2002 bulldozers began uprooting olive groves and pomegranate trees to make way for a system of fencing and barriers intended to prevent suicide bombers from crossing over from the West Bank. Israel’s largest private defense contractor, Elbit Systems Ltd., helped design and build what Israel today calls the “smart fence,” parts of which circumscribe Jerusalem. In some places the smart fence is a steel-mesh structure topped with razor wire and layered with integrated technologies including sensors, radar, and cameras, augmented by drones. In other places it comprises an invisible network of underground wireless sensors, each with a unique IP address.
“Think of it as IoT, the internet of things,” says Haim Delmar, an Elbit senior vice president. “Every sensor has its own logic, it knows where it is, it knows where the other sensors are. They talk amongst themselves to create an understanding of what’s happening.” Data points picked up by the sensors, such as changes in magnetic fields, temperature, and vibrations, are fed into an algorithm that’s grown advanced enough to distinguish between an intruder and an animal or a bush shaking in the wind.
As Israel was isolating Jerusalem, the CBP was formulating its latest modernization scheme. Mindful of the Boeing debacle, the agency required that its new system, called the Arizona Border Surveillance Technology Plan, rely on integrating the feeds from off-the-shelf technology and aim to relay information to local stations rather than Washington. The plan was nothing close to Israel’s skin of sensors and advanced capabilities. “Baby steps,” says Jeff Gwilliam, a CBP deputy program manager. But it did draw on Israeli expertise.
In February 2014, armed with a proven track record, Elbit’s independent U.S. division, Elbit Systems of America LLC, beat out America’s largest defense contractors for a $145 million deal to build the most expensive piece of the Arizona plan—the integrated fixed towers (IFT). These 80- to 160-foot-tall structures were designed to carry daylight and infrared cameras as well as radar capable of spotting targets as far as 7.5 miles away, letting agents see if someone has a backpack or long-barreled weapon and track them as they move through bramble.
Elbit has since completed 43 towers in Arizona, including the one on Sedgwick’s land. Raanan Horowitz, chief executive officer of Elbit Systems of America, foresees an opportunity to sell the government even more advanced technology, such as foliage penetration radar, and to add the kinds of sophisticated intelligence gathering and listening capabilities Israel uses. “It’s not just a bunch of gadgets.” Horowitz says.
But America’s border is five times longer and far more varied in its topography than Israel’s. And there the goal is to prevent all border crossings—with deadly military force if necessary. “In Israel, the main intruders are terrorists,” says Gabby Sarusi, an electro-optic engineering professor at Ben-Gurion University of the Negev and a co-founder of a company called SabraFence Technologies that develops sensors for Israel’s smart fence. “In the U.S., the main intruders are either smugglers or people that want to live in the United States.”
Near the crossing from Nogales, Mexico, into Nogales, Ariz., U.S. Border Patrol agents receive the data transmitted from seven of Elbit’s surveillance towers in a cinder block building that once housed a clothing factory—the type of long-gone manufacturing work Trump says he wants to repatriate. The command and control center is at the end of a windowless corridor. “This room,” says Public Information Officer Jake Stukenberg, “will go from zero to 100, like”—he snaps his fingers.
One wall is filled with dozens of monitors streaming footage from the seven towers and other cameras. The towers are designed to withstand winds of 10 miles per hour, but on this spring day, the wind is moving at 20 mph and gusting even faster. Red outlines flicker on screens lining the wall, indicating movement. A half-dozen men and women sit in front of workstations, each watching several monitors, panning around the desert miles away. Some of the images are crisp; others quiver enough to make someone beg for Dramamine.
The system depends on the human eye. When there are lots of alerts, the agents must quickly decipher what’s happening. “The camera doesn’t know,” Stukenberg says. If we had triggered a sensor on Sedgwick’s ranch, an alarm would have gone off at the Border Patrol station. From there, the agents manning the computers might have pivoted to a camera mounted on an IFT tower and zoomed in to see who was there. It’s also possible agents were just cruising by on a routine patrol. Whether it took a half-hour to get to us because we were seen and identified as a low priority or they just stumbled upon us, the Border Patrol isn’t telling. “That’s the tricks of the trade,” says John Mennell, a CBP spokesman.
Last year the Government Accountability Office said the CBP is getting better at procuring and deploying surveillance technology. Still, while the IFT system was under budget, it fell behind schedule, faced a funding shortfall, and didn’t function consistently. “The agents are very impressed and very satisfied and highly appreciative of the operational tool they now have,” the CBP’s Gwilliam says. “Our system is so sensitive sometimes it can pick up birds.” The GAO later reported that the DHS can’t definitively say how technology has helped spot illegal crossings. A database tracks which systems assist agents in apprehensions, but the reporting is unreliable. Agents in Texas’ Rio Grande Valley, for example, have credited IFTs with helping in almost 500 apprehensions. That’s even though there isn’t a single tower in the state. The parallel lines apparently still aren’t converging.
To contact the authors of this story: Lauren Etter in Austin at letter1@bloomberg.net, Karen Weise in Seattle at kweise@bloomberg.net.
To contact the editor responsible for this story: Miranda Purves at mpurves5@bloomberg.net.
©2018 Bloomberg L.P.
