ARIVACA, Ariz. — As anger over illegal immigration bubbled up nationally, the Department of Homeland Security in 2006 dropped nearly $21 million on what was sold as a state-of-the-art virtual border fence to detect covert crossings and dispatch border agents. The system was heralded as a revolutionary solution to a vexing problem. Crews installed high-tech sensors and camera equipment along a 28-mile segment of border here, south of Tucson and west of Nogales.

Cody Shotwell and Chris Cameron CLICK IMAGE TO VIEW
Learn more about the virtual fence Project 28.

But when Border Patrol agents finally flipped on the switch after months of delays, they found a system littered with problems. It detected rain and blowing leaves as often as it found illegal border crossers. Border agents chased those leads, only to find that their vehicle computer systems failed in the rugged border terrain.

Now, less than a year later, the system is being re-engineered. The estimated cost has jumped to more than $520 million, while plans call for the protection of just 53 miles of the nearly 2,000-mile southwestern border. Congress and the Government Accountability Office have loudly roared their frustration with project delays. As the next nominee to head the Department of Homeland Security, it could be Arizona Gov. Janet Napolitano who holds the key to the program’s future.

Rush for border security

With mid-term elections looming and a national backlash against illegal immigration brewing, President George W. Bush signed the Secure Fence Act in October, 2006. It created a program known as the Secure Border Initiative and directed the Homeland Security department to build 670 miles of fencing along the U.S. Mexico border. It called for 370 miles of pedestrian fences and 300 miles of vehicle barriers. And it set aside $20.6 million for an experimental system designed to electronically detect desert movement, assess the threat and lead border patrol agents to the scene.

The project, officially called SBInet, quickly became known as the “virtual fence.” And it received considerable bipartisan support. Democratic politicians championed the virtual fence’s theoretical ability to stop crossings with fewer environmental consequences than those posed by physical barriers. Republicans liked the extra layer of defense it provided. Both sides saw it as a potentially inexpensive way to secure both international borders.

Arizona Governor Janet Napolitano, a Democrat, even attended Bush’s bill-signing ceremony in Phoenix. Plans called for a pilot project along a 28-mile stretch of border near Sasabe, Ariz., west of Nogales. The sparsely populated area has been a hotbed for human and drug trafficking.

"The lack of user involvement resulted in a system
that does not fully address or satisfy user needs."

DHS and Boeing officials devised a plan to install nine surveillance towers with high-powered cameras, radar equipment and ground sensors. The towers would be controllable from an operations center in Tucson. They would also beam camera images and sensor reports to computers mounted in mobile Border Patrol units.

But problems quickly emerged. Defense contractor Boeing missed its planned June 2007 deployment date. When DHS finally accepted the system eight months late, it fell short of expectations, according to documents from the Government Accountability Office. The supposedly high-powered cameras couldn’t focus beyond five kilometers. They were supposed to be effective at twice that distance. Vehicle-mounted laptop computers designed to receive data from the towers never worked as planned. They proved to be too difficult to operate while driving. Many broke down due to constant rattling in the rugged terrain.

According to GAO documents, Boeing never field-tested the system prior to installation. Nobody consulted the rank-and-file border agents who were supposed to use the new technology. “The lack of user involvement resulted in a system that does not fully address or satisfy user needs,” the GAO’s director of homeland security and justice issues, Richard M. Stana, wrote in prepared testimony for Congress earlier this year.

deanna dent
The moon rises over the small Arizona town of Arivaca. Residents are split in their support of a virtual fence and surveillance towers, which are meant to help capture undocumented immigrants but could also be used to spy on the locals.

Officials blame impossible demands. Amid intense political pressure to secure the border, DHS requested the system be designed and deployed in just eight months. Boeing created the pilot project largely by piecing together existing off-the-shelf technologies.

The communication system between towers and the command center in Tucson was originally intended for law enforcement dispatch. It couldn’t handle the volume of data from cameras, radar units and ground sensors, according to the GAO.

“The preliminary work was rushed, and it showed in the end result,” said P.J. Crowley, a senior fellow who has followed the virtual fence for the Center for American Progress, a left-leaning think tank in Washington. “I’m not surprised that that first project failed.”


Officials involved defend Project 28 as a successful prototype that suffered from bad marketing. They say it was never supposed to work perfectly; it was merely designed to test the concept, and in that regard it was a wild success. “A lot of people I guess don't understand what a prototype is,” said Deborah Bosick,(cq) a Boeing spokeswoman. “It was there to test the concept of technology to help the border patrol do their jobs more efficiently, more effectively and more safely.”

Mark Borkowski, executive director of the Secure Border Initiative, said project designers and the general public had different goals in sight. “Project 28 was really designed to be a learning experience that we frankly misadvertised,” Borkowski said. “We expected to get capability out of it but we really designed it as a learning experience. We thought we were doing the right thing but we didn't explain that very well.”

Borkowski took control of SBI in late September, two weeks after a scathing GAO report criticized the program’s delays. Previous SBI director Gregory Giddens was re-assigned as executive director of facilities management and engineering for U.S. Customs and Border Protection.

Now, what was originally sold as a $20.6 million project to test high-tech equipment along 28 miles of border has turned into a $500 million attempt to secure 53 miles, according to documents submitted to Congress. Another estimated $20 million was allocated to demonstrate the technology on the northern border with Canada. The number reflects money spent through September 2008 when the federal fiscal year ended. It includes estimated costs for some elements of the project.

Borkowski said planners always knew the project would cost a lot more than the $20.6 million allocated for Project 28. Officials used initial money to decide whether the concept was feasible and how far to take it, he said. The total $520 million price tag includes costs for project management, logistics, support and software development—elements required to deploy the project over a larger area and timeframe, Borkowski said.

The Project 28 contract also masks the true cost of the program, he added. DHS and Boeing agreed to a “firm-fixed” contract, meaning the defense contractor agreed to deliver a virtual fence system for $20.6 million regardless of its true cost. “We know what we paid Boeing,” Borkowski said. “But we don’t know what it actually cost Boeing…Maybe Boeing made it all work for $20 million, but they probably didn’t.”

Boeing officials won’t talk publicly about the true cost of Project 28, but Roger Krone, president of Boeing Network and Space Systems, told a congressional panel in October 2007 that the company had spent “over twice the contract value.”

Napolitano goes to Washington

Program observers are now looking at Napolitano’s history to gauge the future of SBI and other border security programs. She has been outspoken in her opposition to physical fencing, saying famously, “You show me a 50-foot wall and I’ll show you a 51-foot ladder.” She has been a strong proponent of comprehensive immigration reform that would strengthen border security while creating new options for legal immigration.

Napolitano has sharply criticized federal border policies under Homeland Security Secretary Michael Chertoff. But the governor has publicly supported the virtual fence and repeatedly criticized project delays. In a February 2007 speech to the National Press Club in Washington, Napolitano said high-tech equipment “will work far better than any 10 or 20 or 50 miles of wall.”

Later that year, before meeting with Chertoff at a border summit in Mexico, the governor wondered why the virtual fence “seems to be virtually missing” and chided the DHS chief for the system’s delays and shortfalls.

As President-Elect Barack Obama’s pick to succeed Chertoff at DHS, Napolitano could soon have direct control over federal border policy, including the programs she has criticized. Officials from her office did not return a call seeking comment on her plans as DHS secretary.

An ever-watching eye

Here in Arivaca, 12 miles north of the border, some residents say they are under the constant watchful eye of an unwanted intruder. Barely two miles from Arivaca Road, the main street through this secluded ranching community of about 2,400 people, a chrome post has sprouted near the dusty base of a small mountain range. A red light blinks as a camera spins 360 degrees. The tower is about 150 yards from the nearest home. A smattering of cattle and some thin desert brush separate them.

View a slideshow of photos of the border fence project.

“Most of us think they put it there because it could also be used to surveil the town,” said Peter Ragan, a construction contractor who moved to Arivaca from Phoenix five years ago. “It does look at the town. It does look at people’s property.”

Ragan and others in Arivaca say the towers were imposed with little warning. And they complain that DHS officials gave them little opportunity to share their concerns before the towers powered on. “Any time you’re under the eyeball of anybody you're not really happy about that,” said Roger Beal, who owns the “Mercantile” general store on Arivaca Road.

Ragan and Beal said Border Patrol agents did improve communications after the Project 28 shortcomings. Officials assured residents that strict policies would prevent agents from training cameras inside private homes, but locals remain skeptical. “A lot of people have a lot of reservations about the towers here,” Ragan said. “There are a lot of people here who are naturally more suspicious of big government. They don’t trust government much to begin with.”

Citing privacy concerns, civil rights groups have unsuccessfully lobbied Congress to shut down the cameras. “It's not the case that just because you're near the border you give up your right to privacy,” said Tim Sparapani, chief legislative council for the American Civil Liberties Union in Washington. “The Department of Homeland Security can't simply eviscerate the Fourth Amendment.”

Looking ahead

Beal and his Arivaca neighbors may have to endure the hated surveillance towers for the indefinite future. DHS and Boeing are preparing to install new towers during the first quarter of 2009. Construction originally planned for summer 2008 was delayed by confusion surrounding the need for environmental assessments, according to a GAO report.

This map shows where the Department of Homeland Security has installed or plans to install technology-based border fences in Arizona.

Boeing and government officials say the new construction reflects lessons learned from their mistakes in Project 28. They slowed the design process and are testing the equipment in actual desert conditions at a facility in Playas, N.M. Satellite communication systems were dropped in favor of faster microwave technology. Instead of designing elements concurrently and hoping they’d eventually work together, engineers this time built pieces individually.

“That should give us high confidence that when we put this out in the field it should work,” said Borkowski, the program director. “Is it 100 percent confidence? No. We'll find some things. But the likelihood of them being significant is not nearly has high.”

Construction is expected to take about six months, Borkowski said. Once completed, officials will continue deploying the technology across the southwestern border, he said.