Workers used to take the day off for Chinese New Year at the Hong Kong International Terminals (HIT), the flagship outpost of shipping giant Hutchison Port Holdings in the former British protectorate. Now with the U.S. economy bubbling again and the global shipping business on the rebound, the busiest freight terminal in the busiest port on Earth goes full-tilt 365 days a year, 24 hours a day. From wrinkle-free pants to portable DVD players, the Hong Kong seaport moves 560,000 truckloads of merchandise to the U.S. each year.
Starting in May, 2004, every single truck bringing a load to HIT will be screened for radioactive material and suspicious cargo using a new portable scanner built by Science Applications International. In the past, time constraints prevented 100% coverage. Older scanners required two to three minutes to accomplish their task.
The new version takes only 10 seconds to scope out a truck, from front bumper to tailpipe -- a delay that's barely noticeable at HIT, where trucks have to make a hairpin turn to enter the terminal queue. Called a VACIS (vehicle and cargo inspection system), the device also combines two formerly separate scanning techniques. Gamma rays look for suspicious images, while radiation detection tracks radioactive signatures.
Installing enough scanners to bring 100% screening capability to U.S. shipments traversing the world's 60 largest ports would cost $500 million to $600 million, less than the cost of a week's operations in Iraq. Extending the program to cover air cargo and other key means for moving goods would cost several billion more. Yet such a capability could provide far better visibility into U.S. transportation systems and their vulnerabilities to terrorist attacks.
"When you have no system of tracking things, then you end up with a crisis of confidence and a Draconian response if something goes wrong," says Steven Flynn, a transportation security expert with the Council on Foreign Relations in New York City. "If you can at least identify where something came from with certainty, we won't have to shut down the Ambassador Bridge from Toronto because a box blew up in Long Beach."
Sounds great. But Flynn and a growing chorus of critics say the Bush Administration has misplaced priorities in the war on terror, choosing to emphasize boots in the desert over far more economical and effective technology solutions to prevent terror closer to home. The federal government has spent only hundreds of millions on cargo protection and a fraction of that on VACIS programs. Even in instances where the Homeland Security Dept. has indicated that it'll pour big bucks into new programs, critics wonder if those efforts are part of a cohesive whole.
"I don't think we're much safer as a result of technology," says James Lewis, director of technology policy at the Center for Strategic & International Studies, a Washington (D.C.) think tank. "We have tons of gadgets and pilot projects, but we haven't tied them together for real intelligence."
Witness the $15 billion U.S.-VISIT program, an effort to identify foreign visitors as they enter America via land, sea, or air, and to track their whereabouts inside the country. Homeland Security is expected to assign the contract for the program, which will meld biometrics with cutting-edge computer networks, to one of three contractors within the next few months.
However, the program is raising more questions than it answers. Namely: How will that information be shared among law-enforcement bodies, from the FBI down to the local gumshoes? And how will U.S.-VISIT tie into other government networks that still don't talk to each other, such as the highly isolated systems at the State Dept.? "If the system is able to combine the watchlist and disperse the information to the appropriate agencies, it's good. If it's just compiling information, it's a waste of money," says Representative John Mica (R-Fla.), chairman of the Aviation subcommittee of the House Transportation & Infrastructure Committee.
So far, Homeland Security hasn't revealed the exact specifications of the project, but even in existing efforts, top-to-bottom sharing has proven problematic. At the same time, critics fear that intelligence efforts may go awry by a growing reliance on technology to do analysis work still better suited to humans.
While the military and intelligence agencies still face tremendous shortages of Arabic-speaking agents, efforts to mine huge vaults of public and private data with software designed to finger likely terrorists has sucked up an inordinate amount of political capital from top Homeland Security officials. Yet those predictive data-mining efforts, such as the now-defunct Total Information Awareness System championed by former Admiral John Poindexter and the MATRIX system, which connects disparate state-level information systems, remain highly controversial and largely unproven.
Skeptics say they simply won't work or will create so many false alarms that the entire system will be worthless. Says Bruce Schneier, technology and security expert and author of Beyond Fear: Thinking Sensibly About Security in an Uncertain World: "Like all crime, terrorism varies. You never know what's going to be important. And there are far fewer examples of terrorism to compare against than, say, auto theft. But data-mining assumes you know what you're looking for. Computers aren't good at that. That's a job for human agents. Unfortunately, hiring more agents doesn't play as well in the press as a $20 billion security project."
True, the U.S. and its allies are spinning out a raft of interesting pilot programs such as the Hong Kong effort and a walk-through explosive-detection portal at a highly trafficked Amtrak rail station in New Carrollton, Md., just outside the Beltway. Homeland Security also is rolling out a new sophisticated breed of sensors to detect biological threats and chemical attacks in the Washington (D.C.) subway system. And it's taking steps toward rapid-fire adoption of standards for communications gear and radiation-detection equipment that will ensure federal, state, and local emergency personnel are using tools that work seamlessly together.
LOOKS LIKE A NIKE.
Indeed, thanks to computer-system upgrades, the FBI can finally send images or audio files over its own networks, and the G-men are using modern data-analysis techniques to parse leads, something that wasn't an option in the pre-September 11 era.
Without a doubt, the digitization of America's borders is starting to pay dividends, as well. The application of radio-frequency identification technology -- smart, tamper-proof tags holding identifying information -- to mark trucks and cars that frequently pass over the border has helped cut down the grunt work of patrol agents stationed in the busy California crossing, just south of San Diego.
Or take the software that accompanies the VACIS machine in Hong Kong and creates a database of shipment footprints to benchmark images for future comparison. "If the software sees 12 containers of Nike sneakers, then it knows what they should look like. If something marked as Nike sneakers doesn't match the image, then an alarm sounds and you can bring in the human inspectors," says Flynn.
TOUGH TO WEAVE.
Adds Charles McQueary, Undersecretary for science and technology at Homeland Defense: "We have 7,500 miles of borders between Canada and Mexico. It would take 6.6 million people to cover those borders. That's a ridiculous number. It illustrates the solutions to security problems cannot only be people-oriented."
Indeed, the Bush Administration has committed some real dollars to funding tech as a part of the $39 billion Homeland Security budget. McQueary's unit has grown from 6 to 200-odd people with a budget of close to $1 billion to use as seed funding for technology projects around the country. At the same time, the federal government has handed out $20 billion to states and localities to help them upgrade their own security efforts.
However, weaving all of these pieces together, from radios for local cops to biosensors for subway systems, into a broader plan to safeguard America is proving a far more difficult task and one that will likely take years to play out.
By Alex Salkever, Technology editor for BusinessWeek Online