The agency is looking at ways to speed up passenger screening while improving security from "curb to gate."
Almost everybody loves to vacation and some even enjoy the flight to their getaway, but you’d be hard-pressed to find anyone with a soft spot for airport security.
Passenger screening policies are intended to keep the skies safe, but that's not what people think when they’re held up in line after an officer mistook a nail clipper for a switchblade. But John Fortune and his team at the Homeland Security Department are on a mission to cut those false alarms.
Fortune manages the Apex Screening at Speed program, an effort by the DHS Science and Technology Directorate to modernize how the Transportation Security Administration vets travelers and their carry-on bags before boarding a plane. Through machine-learning technology and some major hardware upgrades, the project aims to make screening faster and more accurate, and maybe let everyone, not just TSA Pre-Check passengers, wear shoes and jackets through security.
“Our vision for the program is to … improve aviation security from curb to gate,” Fortune said in a conversation with Nextgov. “We’re dealing with an adaptive, emerging adversary, and the threat space is not constant,” so it’s essential to equip TSA with an equally “dynamic, adaptive aviation screening system that can deal with whatever adversaries throw us.”
While the program could ultimately develop screening technologies that span the entire airport, the security checkpoint is its bread and butter. At the center of the checkpoint are two technologies—x-ray scanners to check carry-on bags and millimeter wave AIT machines to scan passengers—and Fortune’s team found better options for both.
The Apex team wants to replace outdated x-ray machines with CT scanners, which produce higher-quality, three-dimensional images that give TSA officers a much better idea of not only what items they’re looking at, but also what materials those items are made of.
CT scanners also open the door for machine-learning tools in the screening process, Fortune said. Each scan provides a vast amount of data his team can use to train algorithms that could help officers discern the difference between a bomb and, say, an antique clock.
“Right now TSA agents are solely responsible for identifying threats in the bags,” he said. “The new algorithms are going to give them a lot more information, be a lot more helpful in identifying things, [and] we hope it will reduce the workload.”
Fortune is optimistic that about the future of the tech, as three vendors have already passed TSA’s initial testing. Agency leaders have shown “a lot of interest” in adopting the new tech, he said, and they’ve taken steps to speed up the rollout.
CT scanners come at a “competitive” price compared to their x-ray counterparts, and the White House 2019 budget proposal included specific funds to buy the machines, Fortune said. Barring any major setbacks, he estimates travelers could see the scanners popping up in airports within the next 18 months.
Screening at Speed would also give the hardware on the passenger end a similar makeover. Fortune’s team is exploring options for “wide-band systems,” which scan passengers using more of the electromagnetic spectrum. The result is a much sharper image than the current full-body systems used in airports today can produce, reducing the need for passengers to shed layers of clothing before they’re scanned.
Though wide-band systems are still in the early development stages, Fortune said his team is already making progress on machine-learning algorithms to speed up the screening process.
This high-quality imaging may harken back to the controversial body-scanners that produced uncomfortably intimate images of passengers that human TSA agents viewed.
TSA has since changed its policy and the detailed scans are now overlaid with a generic representation of a human body so officers can’t see your near-nude physique. Though he noted the new technology produces even higher quality images, Fortune said the rules for how scans are reviewed won’t change. In other words, your body will still be hidden from agents—or others travelers who might be able to see screens.
Fortune’s team is also exploring a number of ways to improve airport security beyond the TSA checkpoint.
One promising technology is the Airport Risk Assessment Model, which uses predictive modeling to determine how TSA and law enforcement should allocate resources across the airport. The Port of Seattle initially developed the tool to monitor threats on the city’s ferry system and has since partnered with Homeland Security on a pilot project at Seattle-Tacoma International Airport.
Fortune’s team is also investigating how to use distributed sensors and biometric technology to assess the risk individual people present before they even get to the checkpoint. Though he said it’s a long way away, Fortune envisions a system that tracks travelers from the minute they enter the airport, perhaps running cursory scans of bags or checking for traces of different materials on their skin.
“Right now, when you get to a checkpoint, everyone has the same assumed risk score—but it doesn’t have to be that way,” Fortune said. If you create a system where “risk is evaluated continuously on a passenger-by-passenger basis … it doesn’t make sense to screen everybody the same way.”
He clarified the system would measure risk based solely on data collected at the airport, and while TSA will ultimately decide how the technology is used, racial and demographic profiling will not be part of the process.
Such tools wouldn’t be available for some time, and even if they become available down the line, technology can only go so far, Fortune said. In his eyes, many of the gains going forward will come from restructuring the way people physically move through airports.
“I’m afraid in the past we’ve been a little bit narrow in focusing on equipment,” he said. “At some point, you have to think of how you string all the pieces together to assess risk as a whole and look at the whole of aviation security.”