A new algorithm deployed earlier this year to on body scanners at airport security checkpoints is improving security and lowering false alarms, a Transportation Security Administration (TSA) official said on Thursday.
The low false alarm rate algorithm in the Advanced Imaging Technology (AIT) systems provides better security detection for on-person screening and lowers the false alarm rate by 50 percent, Austin Gould, assistant administrator for requirements and capabilities analysis, told the House Homeland Security Subcommittee on Oversight, Investigations, and Accountability.
The algorithm was produced by Leidos [LDOS], which also manufactures the ProVision AIT systems used to scan individuals at airport security checkpoints nationwide. The AIT systems feature automatic target detection algorithms to detect anomalies on individuals and alert Transportation Security Officers to the potential need for additional screening measures such as a pat-down on a certain area of a person’s body.
The new algorithm “means fewer pat-downs,” Gould said. “That means fewer people are getting touched. It helps us with keeping passengers happy as they go through the screening experience. And it helps us with keeping things moving at the checkpoint.”
The Leidos algorithm was installed in all of the company’s operational AIT systems as of June 30, a TSA spokesperson said. Non-operational AIT systems are waiting on the availability of computer boards before being upgraded with the low false alarm algorithm, the spokesperson said.
TSA is also exploring additional improvements to the AIT systems to further reduce false alarms and continue to improve detection performance, Gould said.
High-Definition (HD) AIT machines were recently demonstrated at the Department of Homeland Security Science and Technology Directorate’s Transportation Security Laboratory and these will under go prototype field evaluations in 2024, Gould said.
The HD-AIT technology is being provided by a relatively new security detection company, Liberty Defense, which is developing upgrade kits for the existing AIT machines. TSA has awarded Liberty funding for the prototype HD-AIT kits that include wideband HD imaging and artificial intelligence technology.
Gould said the HD-AIT upgrades will significantly reduce pat-downs and “allow us to rule out nuisance false alarms associated with clothing and body styles.”
Rep. Shri Thanedar (D-Mich.), the ranking member on the subcommittee, cited an earlier Government Accountability Office report that disclosed that the AIT systems have generated alarms that “disproportionately” alert on transgender passengers, people with religious headwear, persons with disabilities, and those with “coarse hair, including African American women.”
The reason for certain “demographics of people” and people wearing certain types of clothing being referred for secondary screening more frequently than others is not profiling but a technology issue, Gould said.
Rep. Nick LaLota (R-N.Y.) said he will soon introduce the “Better Screening Act,” which is aimed at reducing pat-downs and reduce the need for travelers to remove various items from their carry-on bags at checkpoints by improving technology deployments that also enhance security.
To further improve technology at the checkpoint, Gould highlighted that existing authorities require the AIT systems to use an avatar that essentially looks like a stick figure to show where an anomaly may be present on a person. TSA wants new legislative authority to “use a visual image of the person as they’re presented in their clothing, preserving all privacy associated with that,” he said. This would provide “a much more accurate location of where that threat object might be, providing for a much more targeted pat-down, which I think would also provide a better passenger experience.”
During the hearing, which examined technology in aviation security, Mario Wilson, TSA’s assistant administrator for acquisition program management, reiterated previous agency comments that at current funding levels deployments of new computed tomography (CT) systems to scan carry-on bags will not complete deployments until 2042. He also said that biometric-enabled credential authentication technology systems will not finish rolling out until 2049 given planned funding.
The checkpoint CT systems are TSA’s top technology priority but anemic funding requests the last two years are slowing deployments. David Pekoske, TSA’s administrator, has said he could complete the rollouts in five years if given between $300 million and $350 million annually for checkpoint CT systems.
Currently, Analogic, Integrated Defense & Security Solutions—now part of Tek84—and Smiths Detection are under contract to provide checkpoint CT systems and related baggage handling capabilities.
In fiscal year 2022, Congress funded TSA’s approximately $105 million request for checkpoint CT. For FY ’24, the agency is seeking $70 million. House appropriators have recommended $105 million but Senate appropriators have proposed zeroing the request, which is still well short of Pekoske’s hope for at least $300 million.
The low funding combined with unpredictability that flows from continuing resolutions and the threat of potential government shutdowns are hard on the security detection industrial base.
“Prime contractors and small businesses are challenged with developing accurate lead times on material,” Kevin Schmidt, vice president of government relations for Smiths Detection, told Defense Daily in September. “In addition, small contract awards and inconsistent orders lead to subcontractors downgrading our prioritization.”
An executive with another checkpoint CT provider also told Defense Daily in September that $105 million is the “bare minimum” of support for industry annually, adding that the funding also goes toward things like program management and installation costs and not the actual scanners.
All three companies have invested tens of millions of their own dollars into developing their respective checkpoint CT systems and breaking or slowing production makes it hard to “support the engineering team that needs to continually spiral performance development and impacts the ability to hang on to the skilled workforce necessary to put these systems in the field,” the executive said.
International demand for checkpoint CT systems is helping the industrial base but TSA is a “major driver of technology” and investment and “the lack of funding for this technology does put a significant impact on…industry’s ability to support the government’s initiatives,” he said.