– At first glance, it looks like an ordinary, TV antenna tower, Anywhere, USA.
Thirty-five miles of dusty desert from the nearest highway, it telescopes almost a hundred feet up in the air, topped with cameras and radars.
Nearby, a security man from Pinkerton National Detective Agency sits in a dark-colored SUV, watching movies on a laptop. Two giant tanks of water apparently for wandering migrants have been placed close at hand by the human rights group Derechos Humanos. The hum of a gas generator is the only sound among the mesquite bushes and paloverde trees.
The tower is one of nine surveillance turrets strung across 28 miles of Arizona border north of Sasabe that are supposed to communicate coordinates and images of moving figures to remote centers and laptops in border patrol vehicles.
They are part of "Project 28," a Department of Homeland Security initiative meant to test the viability of a "virtual fence" – a high-tech, possibly more effective alternative to the fencing the US is erecting across hundreds of miles of the southwestern border. The idea that radar towers could help fill in the current gaps in the physical wall, and the technology could even be transported to problem areas at will appealed to the DHS.
President Bush touted the project in May 2006 as "the most technologically advanced border security initiative in American history."
The excitement over the possibility of securing the border through all-seeing electronic eyes has ebbed slightly since. Technological snags in the project, which was contracted out to Boeing for $20 million, have raised doubts among some congressional investigators about the workability of the entire venture.
Nevertheless, Homeland Security officials say that despite the hiccups the project is on track.
A high-tech alternative
At the top of the tower, a rotating radar dish detects moving figures on the border and radios the coordinates to an integrated camera. The camera then focuses in on the targets.
Agents sitting at a remote command post look at these images on monitors to determine if the moving figures are human – as opposed to a roadrunner or other desert critter. They then radio the coordinates to laptop computers placed in the cabins of roving border patrol cruisers. Ostensibly, the agents then swoop into the designated area and catch the illegal border crosser.
Project 28 is an early test of various new ways to coordinate radar, cameras, and on-the-ground apprehension being developed under a broader umbrella known as SBInet – a DHS initiative launched in 2006 to examine technological alternatives to fencing along unspecified stretches of the border.
In early trials from July to December 2007, technical problems and other snafus led to media reports that DHS and the Boeing company might mothball the project. Problems included software glitches, wind and rain that affected the camera image quality, and radar that had trouble distinguishing sage-brush from camping migrants or animals.
Boeing officials publicly admitted that the effort has been more challenging than they anticipated. The project, which was supposed to be handed over to border patrol in June 2007 was not accepted till December. At a recent congressional hearing, Richard Stana, Homeland Security and Justice Director for the General Accounting Office said that Project 28 "did not fully meet the user needs."
But DHS officials have moved to dispel the notion that the project was stalled or scrapped. They say the technology is still in use, that it is being tested to improve various designs and capabilities.
It has been responsible for catching more than 2,400 migrants in the desert testing areas, the DHS claims.
DHS withheld part of the $20 million original funding until Boeing made the necessary corrections, according to DHS press secretary Laura Keehner. "As good stewards of the taxpayers money, DHS delayed acceptance of P-28," says Ms. Keehner. "After a period of operational testing, additional deficiencies were identified and subsequently corrected to the department's satisfaction."
DHS has requested $775 million next fiscal year to continue to develop and deploy such technology. "There are some things we want to improve and there are some things that probably it turns out we don't really need," DHS Secretary Michael Chertoff said in a press conference in Washington Feb. 28. "But I envision we will use this design in other parts of the border."
Congress raises questions
Some congressional investigators have warned that if they judge the system to be underperforming in their eyes, they may urge ending the project.
Many were not impressed with the shadowy footage taken in late February in which Project 28 cameras tracked three large groups of immigrants crossing the border before relaying the images to a command post in Tucson 70 miles away.
"Project 28 was supposed to be an example of how we could use technology to secure the border. The lesson is we can't secure 28 miles of our border for $20 million," said Rep. Bill Pascrell Jr. (D) of New Jersey, House Homeland Security Committee member at a hearing Feb. 27. "After so many years of promises and tests and millions of dollars spent, we are no closer to a technological solution to securing the border. This is unacceptable."
Another possible problem, experts say, is that the radar is easily foiled by terrain that is not flat. And operator training appears to be important if the system is to be effective.
"Parts of Project 28 hold much promise if you can nurture experienced operators who can detect migrants and then guide other agents to intercept them," says T.J. Bonner, president of the National Border Patrol Council, which represents more than 12,000 United States border patrol agents. "The radar doesn't give you depth perception, and the same can be said of the cameras especially at night. So it's easy for contract employees who are trying to sector in the precise locations to be way off."
Some investigators of the US Government Accountability Office, the investigative arm of Congress, have said that the initial investment was too paltry to expect significant results and that Boeing was not given enough time to complete the fairly complex project.
The company has also received $64 million for a new contract, according to Keehner, to develop new command-and-control software, improved identification capability, and greater range.
After testing, she says, the new hardware and software are intended to be installed in two locations, one in Texas and one in Arizona.