Is This Machine the Future of Airport Security?
The Qylatron, used daily at San Francisco’s Levi’s Stadium, promises better, faster security screening
We’re all familiar with the security procedures at sports arenas, amusement parks and other large venues. You wait in line, then walk through a metal detector and hand over your bag to a security guard, who quickly sweeps through its contents with a flashlight before handing it back to you.
“What if there was a better way?” says Lisa Dolev.
An Israeli air force veteran and security consultant who specializes in suicide bombings, Dolev wanted to create a better security scanner for years. But watching footage of the Madrid terrorist attacks in 2004 pushed her into action. That evening, she sat down and drew a sketch of a new kind of security machine.
That sketch became the Qylatron Entry Experience Solution, a technology which Dolev says can provide faster, easier and more accurate security screening. The Qylatron is currently being used on a limited daily basis at Levi’s Stadium in San Francisco. The U.S. Department of Homeland Security's Transportation Security Laboratory is also considering the system for a variety of security checkpoints, including at airports.
The Qylatron looks like a futuristic bee hive, with multiple hexagonal boxes stacked on top of one another. A patron puts his or her ticket in the machine’s ticket slot, which opens a door to one of the pods. The person then places a bag inside, and the door locks. Inside the machine, various sensors scan the bag for weapons and other banned items. If the bag is determined to be safe, the door unlocks. If not, an alarm goes off to begin a security procedure.
Dolev is cagey about the details of how the scanning works. “We’re in security,” she says. But she can say that the machines use a combination of multi-view x-rays, chemical sensors and artificial intelligence.
The artificial intelligence component is, perhaps, the most unique aspect of the Qylatron. The machine’s algorithm allows it to “learn” about different objects, making it better at distinguishing threats from ordinary items. For example, a Qylatron in a rainy city might quickly learn the shape of an umbrella.
The Qylatron can also communicate with “peer” machines around the globe to enhance its learning. Peers might include airport security systems or subway security scanners—basically, any machine that has a similar purpose and is working to detect similar threats.
The Qylatron’s intelligence allows it to be personalized based on its venue. A Qylatron at an amusement park might learn how to detect picnic foods and allow them to pass unimpeded, while a machine at an alcohol-free concert venue could quickly learn the signature of vodka hidden in a bottle of Diet Coke.
By scanning patrons’ tickets, the machines can also personalize their approach based on known identities. A VIP might get a special welcome message on the machine’s external screen. A chef who has come to cook at a venue might be allowed to bring in knives, whereas the same knives in another guest's bag would cause a security alert.
“Each machine needs to have a different algorithm,” Dolev says. “It becomes specialized and learns for the venue.”
The system can process 600 people an hour—five at a time—and needs only four human operators.
Creating the Qylatron took seven years of research and “an army” of engineers—chemical, mechanical, electrical and industrial—as well as designers, marketers and more. The Qylatron is the signature product of Qylur Intelligent Systems, Dolev’s San Francisco-based security company.
The name Qylatron may sound like something from a Star Trek episode. But it was actually inspired by nature. “Qylur” is a sonic echo of the star-nosed mole (Condylura cristata). Star-nosed moles, though blind, can make quick decisions thanks to the thousands of sensory receptors in their snouts.
Dolev says Qylur technology may have applications other than security, such as agriculture or medical diagnostics. The key is the way the technology uses sensory data in combination with artificial intelligence to make quick decisions about how to proceed with a specific task.
“That’s what I mean by intelligent machines,” Dolev says. “Machines that make critical decisions and changing decisions.”