If Project Hostile Intent is the Department of Homeland Securitys idea for using technology to catch terrorists before they can board aircraft, the Transportation Security Administrations (TSA) Screening Passengers Through Observation Techniques (SPOT) program is the shoe leather version. With Project Hostile Intent, the DHS is essentially trying to automate what TSA staff are already doing in airports today (see this week's story, Big Brother Really Is Watching). But SPOT has yet to result in the arrest of any terrorists at least publicly. Why is that?
Project Hostile Intent is supposed to use technology arrays of audio, visual and infrared sensors - to detect terrorists by monitoring the sound of a persons voice, the gestures they make and even "micro facial expressions" that could indicate that a subject may be acting suspiciously. Trained security staff already look for those cues every day in airports and border checkpoints.
The TSAs SPOT program has been in place since 1993, although not at all airports. In that program, trained security staff watch for indicators of suspicious behavior. We are looking for changes in demeanor when people start to submit to the security process, says Carl Maccario, TSA program analyst.
SPOT is based on techniques originally developed by Israeli security forces. The idea is to focus on people, rather than weapons. Its the same as what the Israelis do, except that we dont detain and interrogate people at the checkpoint, Maccario says. Instead, TSA personnel approach persons exhibiting suspicious behavioral patterns and engage them in conversation to see how they react.
But theres another difference between the Israeli system and SPOT. In Israel they racially profile, he says, while the TSA watches only for behavioral characteristics that might indicate deception.
"The Israeli concept says if we identify a person we define as a high risk passenger and we search him at a completely different level, that ensures that he cant carry anything that would be a threat," says Rafi Ron, president of New Age Aviation Security in MacLean, Va. Ron was formerly in charge of security at Ben Gurion airport in Tel Aviv. His firm developed a technique called Behavior Pattern Recognition, which focuses on behavioral clues and was used at Logan International Airport following the 9/11 attacks. "Concealment of intentions or items creates certain behavioral patterns that are very typical," says Ron. SPOT was designed with a similar goal in mind.
Since SPOT launched, TSA has trained 2,000 employees to use the techniques when watching and interacting with the public. Since its inception, the program has generated 42,604 referrals (meaning people who met a threshold and were referred for additional screening) and 3,092 referrals to law enforcement organizations. Of the latter, 278 were arrested and 442 are under continuing investigation. From January through August of 2007, SPOT yielded 181 arrests. However, not one terrorist or terrorism suspect has been publicly identified as being detained as a result of the program.
Maccario says that doesn't mean that the program has been a failure. I asked him why no terrorist suspects have been caught using the program. "We [TSA] dont know that, he says. We know that there have been many people referred to other agencies that are in ongoing active investigations. Thats all I can tell you."
Maccario thinks the program has other benefits as well. "We may not know if the people SPOT caught in the country illegally, using fake passports/IDs or smuggling money or drugs were doing so to assist with a larger plot. But it's clearly an effective means of identifying people engaged in deception and criminal activity," he says.
As for preventing terrorism, he says the presence of trained security personnel also has a deterrent effect that "can't be overstated." Ron also thinks such programs can work if implemented correctly. He says the program has been successful in Israel, and not just at airports. If you look at the Israeli experience, terrorists have been attacking shopping malls for years. None of them have succeeded in getting into a mall. All of them exploded outside when they realized that there were guards at every door. Once the guard established eye contact with them and identified them as suspicious individuals, they blew themselves up.
But the system in place in the U.S. is very different. Protecting U.S. airports is a much larger scale operation, and its impossible for humans to watch every person, much less engage in a conversation with all of them. In Israel they have maybe 20 flights a day, says Bruce Schneier, chief technology officer at security consultancy BT Counterpane. That makes it much easier to interview every passenger. Israelis also interview passengers before theyre allowed to check luggage.
SPOT isnt nearly as thorough. We dont talk to people. We just screen them. Its relatively superficial and therefore we still remain vulnerable, says Ron. And the Israelis have another advantage: They tend to know that the bad guys are Palestinian. In the U.S. you dont know, Schneier says. Shoe Bomber Richard Reid is a prime example.
If suspicious persons can be identified through automated processes, and 10% of the people at an airport were flagged, short interviews by trained personnel could then cut that down to 1% that require further interrogation, says Ron. The interview, he says, is key. The weakest point of the terrorist is the lack of ability to withstand an interview without exposing themselves. Those who fail the interview might require a more extensive search. Those persons, he says, would be searched at almost a forensic level. (Ouch.) But thats a much more rigorous approach than used in the SPOT program today.
Neither Maccario nor Schneier think the process of identifying suspicious persons can be automated. The best machine you can have out there to detect anomalous behavior is another human being, says Maccario. Schneier says the problem is not one that machines arent as good at performing. Our brains were developed to detect intent in other people. Thats why our brain exists to deal with social interactions, says Schneier. The problem of using computers to identify suspicious behavior is very much like speech recognition and artificial intelligence, he adds. Those are all things that human brains are optimized for and computers are not.
If the SPOT program is truly effective and beneficial, one nagging question remains. The government has a strong desire to show results in its war on terrorism. If terrorists had been caught by the TSA during the our years that the SPOT program has been in existence, it seems likely the public would have heard about it.