The Transportation Security Administration (TSA) recently said it has begun testing new software on its L-3 Communications’ [LLL]-built whole body imaging systems that can automatically detect potential threats hidden beneath a person’s clothing.
The pilot tests of the new software on L-3’s ProVision Advanced Imaging Technology began yesterday at Las Vegas McCarran International Airport. In the coming days, TSA will begin testing the anomaly detection software on ProVision machines currently installed at Hartsfield Jackson Atlanta International Airport and Ronald Reagan Washington National Airport, TSA said.
“Testing this new software will help us confirm test results that indicate it can provide the same high level of security as current advanced imaging technology units while further enhancing privacy protections currently in place,” TSA Administrator John Pistole said in a statement.
TSA also buys AIT machines from Rapiscan Systems, a division of OSI Systems [OSIS].
An agency spokeswoman told sister publication Defense Daily that the software for the L-3 units, which are based on millimeter wave imaging technology, have already met TSA’s detection standards in lab testing.
“When software being developed for backscatter units meets TSA’s detection standards, TSA will test it in the airport environment, too,” the spokeswoman said in an e-mail response to questions. “TSA anticipates testing this software for backscatter units at airports in the coming months.”
Rapiscan’s AIT technology uses backscatter X-ray imaging. It’s this technology that has provoked some concerns about the health effects of the radiation emitted by AIT machines. However, the U.S. government maintains that the amount of radiation given off by these machines is relatively insignificant. The millimeter wave technology is non- ionizing.
Testing of the new detection software began last fall at the TSA Systems Integration Facility. The agency said the software was developed by working with the Department of Homeland Security Science and Technology Directorate and private industry.
The anomaly detection software eliminates the controversial graphic images of individuals that are currently reviewed by a TSA screener in a remote viewing booth so that he or she can’t see the actual person being screened by the AIT machine. With the new software, potential threats are automatically indicated at a location on a generic outline of a person.
The new software is expected to further address privacy concerns individuals have with the AIT technology, namely the graphic depiction of their bodies and the screen used by the remote security officer.
If no potential threat items are detected, then an “OK” will appear on the screener’s monitor with no outline of the individual.
If an anomaly is detected, a generic outline of the person will appear on the monitor attached to the AIT unit and highlight any areas on the person that require further screening, the agency said. As with current procedures, TSA officers “will work with the passenger to resolve any alarms, which will include a targeted pat-down of areas that are highlighted on the screen,” TSA said.
The pilot tests are scheduled to last several months and will use one AIT machine at each airport.
One area of concern about the anomaly detection algorithms that has been mentioned by industry officials and aviation security experts is that they frequently don’t detect potential threats, meaning that without a security officer reviewing the raw image, security gaps will be opened.
TSA, in its e-mail response to questions about false alarm rates, said that the lab tests of the new software met qualification standards. The purpose of the new operational pilot tests is to “determine whether it is a viable option for deployment. As with the current version of AIT, false alarms remain a possibility. For security reasons, TSA cannot disclose specific detection capabilities.”
TSA purchased and deployed nearly 500 AIT machines last year with most of the business going to Rapiscan. TSA’s FY ’11 budget contains funding to purchase 500 more of the body imagers, but industry officials have said they don’t expect the agency to buy more systems until the auto-detect algorithms are certified.
L-3’s AIT machines deployed to Amsterdam’s Schiphol Airport have been operating with the auto-detect algorithms for more than a year.