A continental security job

Whether they are monitoring activity at bustling airports and seaports or covering vast, remote regions, government agents charged with protecting U.S. borders struggle to meet the increasing demands of the times. Lingering fears from the Sept. 11, 2001, terrorist attacks continue to shape those demands.

One solution is to increase the number of employees. Homeland Security Department officials announced in March, for example, that they were assigning more than 500 additional Border Patrol agents to help stem the flow of illegal immigrants across the Arizona desert region.

Congress had included plans to add 10,000 border agents in five years as part of the National Intelligence Reform Act signed into law late last year, though President Bush's 2006 budget proposal included funding for only a little more than 200.

But adding more employees is unlikely to completely solve the problem. In a December 2004 report on DHS' major management challenges, the department's inspector general said no matter how many Border Patrol agents are added, they can't effectively monitor some border regions.

So to improve border protection, DHS officials will rely on technology programs to help extend border agents' reach. The America's Shield Initiative (ASI), formerly known as the Integrated Surveillance Intelligence System, is one such program. It's a major component of the government's technology-intensive smart borders strategy.

In testimony to the House Appropriations Committee's Homeland Security Subcommittee in March, Robert Bonner, commissioner of DHS' Customs and Border Protection, said ASI is an important tool because it allows Border Patrol agents to remotely monitor the border. They can then respond to specific illegal crossings rather than patrol an entire area adjacent to the border.

"By contrast, Border Patrol operations without ASI support are not only less effective, they are more resource-intensive and less safe for Border Patrol agents," he told the House panel.

Speeding trade

The use of technology to support border agents depends on their particular security needs.

At seaports, the basic problem is moving high volumes of cargo and people through relatively small areas, said Rod MacDonald, acting assistant commissioner of CBP's Office of Information and Technology. The ideal solution would allow port officials to collect and review shipment data before cargo arrives so they could more quickly sift possible threats out of legitimate traffic.

Radio frequency identification (RFID) technology has an increasingly significant role at seaports because it can provide a hands-off way for border agents to quickly determine a particular cargo's contents, its origin and destination, and even whether someone has tampered with the container.

The technology is also useful at land borders, where agents have set up RFID readers about 100 feet from border inspection booths, MacDonald said. From those scans, border agents can receive information about a truck's contents and compare that with a list of items that should be on the truck. They can target suspicious trucks for closer inspection before the vehicles reach the border.

"Given the volume of trucks coming into the country each day, that means considerable savings in time and effort," MacDonald said.

The next part of this strategy is to develop technologies that allow finer analysis of the data and better recognition of potential problems. The primary needs are new rules engines, link analysis and pattern-recognition software, which would enable DHS' limited resources to more accurately identify potential threats, he said.

"The worst case would be where analysts would have to look at every single manifest coming into the country," MacDonald said. "The force multiplier there is the much smaller subset of the data than would otherwise have to be considered."

Covering ground

To protect the United States' long borders with Canada and Mexico, agents need technology that would help small numbers of them cover large, often inhospitable terrain. Despite those challenges, they must be able to spot people or vehicles that could pose a problem.

During the past few years, DHS has developed video technologies to simplify an agent's task. Video cameras strategically placed along the borders have helped agents secure large portions that they would otherwise have no hope of covering.

Although video technology has improved, officials have a better understanding of its limitations based on their recent, hands-on experience with the technology, said Alan Lipton, chief technology officer at ObjectVideo. At one time, the expectation was that video could significantly reduce border agents' workload, but officials clearly need brains with the video eyeballs, he said.

"Throwing some kind of intelligence onto video streams is an absolutely essential piece of the business now," Lipton said.

For example, Homeland Security Advanced Research Projects Agency officials recently awarded contracts to study an emerging field called automated scene understanding; ObjectVideo won two of those contracts.

Border agents need automated intelligence that can recognize a problem from video feed and then alert them to threats.

ObjectVideo's video-analysis algorithms are based on artificial intelligence technology called "computer vision," which compares all images in a camera's view against standard rules. If images violate those rules, the intelligent video system sends an alert by phone, pager or e-mail, or to a central alert console.

Future developments will likely include layering video with other sensor data to provide more comprehensive monitoring information.

VistaScape Security Systems, for example, produces a vision system similar to ObjectVideo's, but it can also integrate with Global Positioning System technology.

In restricted spaces, you may not need that level of location sensing because the geography is well known, said Glenn McGonnigle, VistaScape's chairman and chief executive officer.

"But in wide-open spaces, where you are dealing with longer-range cameras, you want to have the ability not just to detect objects but to correlate them with others," he said. "In that situation, it's important to know both what you are looking at and where it is."

Digital Infrared Imaging, a company specializing in thermal imaging, manufactures systems that combine pan-and-tilt closed-circuit televisions with infrared capabilities, producing a single image that incorporates the two technologies. At night, the value of infrared is obvious. But even during the day, agents can refine an indiscernible image by adding an infrared dimension.

Seth Ellis, the company's CEO, said Digital Infrared Imaging is now offering this product for border-protection applications.

And DHS officials are studying whether they can boost those types of sensors with unmanned aerial vehicles (UAVs). Used extensively by the military in Afghanistan and Iraq, UAVs were recently evaluated as part of the first stage of the Arizona Border Initiative. The study tested the vehicles' ability to work with ground-based video systems.

But the future direction of UAVs for border protection remains unclear, MacDonald said. There's still a lot to be learned, such as whether unmanned drones programmed for autonomous flight are more effective than vehicles piloted by remote control. Officials are also considering the different sensor packages that such vehicles could carry.

Those technologies, in combination with radar and ground-based motion detectors, seem to form the beginnings of a formidable protective barrier. The problem is that most of those technologies are in the development or evaluation stages and have not been thoroughly tested.