A recent pilot test that incorporated automatic target recognition (ATR) algorithms on whole body imagers deployed at some of the nation’s airports was successful, making it more likely that the anomaly detection capability will begin to be installed soon, according to the nation’s top transportation security official. The pilot tests at three airports earlier this year with the ATR algorithms met with “good success” at three airports, John Pistole, administrator of the Transportation Security Administration (TSA), tells a House panel. “By that I mean our expected rates in terms of detection and false positives and throughput,” Pistole says. “Some of the basic criteria have all been positive and so for about half of those machines, 240-plus, the plan is, assuming a couple more things are done in the next week or two, that we would modify those through the rest of this [calendar] year.” Pistole is referring to the anomaly detection algorithms developed by L-3 Communications [LLL] for its ProVision Advanced Imaging Technology (AIT) systems. TSA has deployed 490 AIT machines, including the Secure 1000 developed by OSI Systems [OSIS] Rapiscan division. Pistole says that Rapiscan hasn’t completed development of its ATR algorithms although they are in lab testing and he expects pilot testing to begin this fall. Currently the AIT systems provide security officers with a graphic image of the person being scanned. The ATR feature is meant to eliminate the privacy concerns by presenting a generic outline of a person and then highlighting an area or areas on the individual to show a security officer where a potential threat may be hidden. Pistole didn’t say what TSA’s procurement plans are for 500 AIT machines that Congress appropriated funding for in the FY ’11 budget. Industry officials believe the agency is waiting until the ATR algorithms are ready before obligating those funds. Pistole tells the House Homeland Security Subcommittee on Transportation Security that the AIT technology is “the best” there is currently to detect non-metallic threats hidden beneath a passengers clothing. He also says that the body imagers are increasing passenger throughput at checkpoints. However, Pistole says, throughput at checkpoints is being negatively impacted by the increase in passenger carry-on bags, which reflects many airlines’ decision to charge passengers fees for checked bags, and because passengers are stuffing more items into their carry-on bags, making it more difficult for TSA security officers to screen those bags. “It’s taking longer to resolve those issues than it is the passengers themselves,” Pistole says. Pistole says that the use of AIT machines isn’t a panacea to checkpoint security and remains just one of many tools. He says the agency also wants to make more use of vapor-weight canines that can detect molecules of explosives in the air after a parcel or bag has gone by.

Risk-Based Passenger Screening

TSA continues to review its options for testing risk-based passenger screening concepts at checkpoints to enable expedited screening for known or trusted travelers, Pistole says. The agency plans to do “internal work” this summer, including training for its Behavioral Detection Officers, and educating and training its workforce for how a known-traveler model would work, and then do some pilot projects at some airports this fall that may include a dedicated lane for these travelers, he says. The program would be expanded in 2012, Pistole says. Pistole says the known-traveler effort could look different from one airport to the next to account for differences at each airport. “So we need to manage expectations with the traveling public,” Pistole says. Creating a known traveler program would allow TSA to “focus its limited resources on those we know the least about and may cause us the most problems,” and for known passengers get away from “some of the physical screening that we’ve come to be known for,” Pistole says. The types of information that TSA would seek from voluntary participants goes beyond the typical name, date of birth, and gender data that the agency obtains from all individuals prior to flying in order to do a watchlist check. Pistole says additional information could include travel history. Military service members, people with top secret security clearances and participants in Customs and Border Protection’s Global Entry program are all people that we know a lot about and would be easy to add into any known traveler program that TSA manages, Pistole says. The known traveler program would not be fee-based, he says. TSA is hoping to generate lessons for any known-traveler program through its effort to expedite checkpoint screening for airline pilots. This identity-based solution has already been pilot-tested at several airports and the agency, with the help of the airlines and pilot associations, is working toward a broader roll out. Pistole says the groups involved in the effort are continuing to work out the technology involved, which he says must be common from airport to airport. Once the known pilot effort is implemented and shows success for several months, then TSA will be ready to incorporate additional crew members into the effort, Pistole says. In the near-term the known crew member effort will not incorporate biometrics as part of the identity-based solution, Pistole says. “That might be an end state we build to but because of additional costs and time that is involved…I wanted to do something initially recognizing pilots as the most trusted people on the aircraft,” he says.

Air Cargo Update

Pistole says that the best way to secure the global air cargo supply chain will also likely entail a risk-based approach. Ever since the failed plot last October involving explosive devices TSA has worked with other countries, the air cargo industry and CBP and “what we have learned is to do a piece by piece screening of each item of cargo or say mail parcels, over say 500 grams or whatever it may be, would really shut down the global supply chain which we have no interest in doing,” he says. Instead, TSA wants to work with its partners to get advanced information about packages, particularly those originating overseas from high risk areas where screening isn’t as thorough and destined for the U.S. aboard passenger planes, Pistole says. TSA wants to know if the shipper is known, that is, does the shipper have a relationship with the air carrier or forwarder, and is the shipment known, he says. “What we have done with industry and what they have developed on their own is develop some rule based protocols to say, ‘does this make sense, that somebody is paying $500 to ship a computer printer and some books and clothes from Yemen to Chicago?” Pistole says. “So it’s that combination of getting advanced information, similar to advanced passenger information, passenger name records; that construct for cargo. I think we can achieve a high percentage of cargo from the [coming into] the U.S. but getting to 100% with any confidence would require substantial additional resources for us not only to trust but verify what has happened on the ground from all those last points of departure around the world.”