Software Glitches Put Virtual Border Fence Months Behind Schedule

By Nathan Burchfiel | July 7, 2008 | 8:23 PM EDT

( - Efforts by the Department of Homeland Security to establish a 28-mile digital fence along the U.S. border with Mexico are four months behind schedule, according to a report released last week by the Government Accountability Office.

Project 28, a $20 million project to secure 28 miles of the Arizona border with radar, sensors, computers and other technology, has fallen four months behind schedule due to software problems, according to the report. Project 28 is the first of several programs aimed at electronic surveillance.

Project 28 is part of SBInet, a $7.6 billion, five-year Secure Border Initiative program aimed at developing a border protection system that mixes security infrastructure, including 570 miles of pedestrian and vehicle fences, with technology-based measures such as radar, cameras, and computers.

The GAO report blames the government contractor, Boeing, for the delays, saying while the company delivered and deployed the project's hardware components on schedule, "the delays are primarily attributed to software integration problems - such as long delays in radar information being displayed in command centers."

"In September 2007, CBP officials told us that Boeing was making progress in correcting the system integration problems," the report said, "but CBP was unable to provide us with a specific date on when Boeing would complete the necessary corrections to make Project 28 operational."

Deborah Bosick, a spokesperson for Boeing's SBInet Program, declined to comment on the report, referring questions to Customs and Border Patrol. A spokesman for CBP did not respond to requests for comment Monday.

Richard Stana, director of Homeland Security and Justice Issues at the GAO, told Cybercast News Service Monday that Boeing is scheduled to begin testing the new system at the end of October and "based on how well that goes, and the pace at which the government decides to accept or not accept the Boeing system, perhaps by the end of the year there could be acceptance, but perhaps not."

"They say they're getting on top of it, but we'll find out in testing," Stana said, adding that once existing software problems are worked out, there will be questions about whether the hardware will be able to stand up to harsh conditions along the border.

"Is the hardware that is going to be placed in the different vehicles, is it going to stand up to the constant bouncing and jolting that you get sometimes out in the field? Are the antennas going to stand up to a good whacking by a tree branch?" Stana asked. "All of these have to be tested and the results have to be evaluated."

Proponents of stronger border enforcement said the report raises questions about the direction of enforcement measures, including whether "virtual fences" are worth the investment.

"They've spent $20 million so far ... on this very small project, relatively speaking," Jessica Vaughan, a senior policy analyst at the Center for Immigration Studies, told Cybercast News Service. "And it's not working yet."

Vaughan said that glitches "are to be expected in any project." She questioned "whether this is the best use of our resources now or perhaps a more traditional fence, such as the one that has worked very, very well in San Diego County, might be" more efficient.

The GAO report also found progress in SBInet's fence-building efforts. CBP met the goal of building 70 miles of more than 500 planned miles of pedestrian and vehicle fencing. But the report warned that costs for building a physical fence may end up being higher than expected.

"Every government contract seems to cost more and take longer than anybody thought to begin with," Vaughan said. "Good border security is expensive and costs money, but relative to the cost of illegal immigration it's clear that it's very necessary to do.

"There's not one-size-fits-all solution, but it doesn't necessarily have to be a high-tech solution to be effective," Vaughan said. "There's not a lot of point in having fancy sensors and radar and surveillance if there's no one there to go catch the people who've reached the border."

But Stana cautioned against writing off surveillance-based border enforcement because of delays in implementing the pilot project, scheduled to be the first of many along various sections of the southern border.

"I think we'll have to wait and see until the testing comes out," Stana said. "If parts of the system don't work as well as they had hoped, they can always make modifications, but I think the testing is going to be key, good, rigorous, real-life testing."

Make media inquiries or request an interview about this article.

Subscribe to the free daily E-Brief.

E-mail a comment or news tip to Nathan Burchfiel

Send a Letter to the Editor about this article.