Since 2001, the security industry has poured immeasurable resources into aviation security. These resources have largely focused on checkpoint screening technology – more specifically, these devices offer the ability to accurately and reliably screen people and baggage for known threats to aircraft and related equipment.
Advanced x-ray, millimeter wave, and trace detection equipment by companies like Leidos, L3, Smiths Detection, and others are, of course, ubiquitous to frequent travelers.
More recently, the stubborn persistence of active shooter events in the United States has necessitated screening checkpoints at softer targets such as sports venues, performing arts centers and schools. The natural response has been to deploy much of the same technology and procedures developed for the aviation industry – i.e., divestiture of personal belongings, an x-ray and magnetometer routine, and secondary screening with handheld metal detector.
The problem here is two-fold. First, airport-level screening detracts from the user experience. Venues focused on providing an inviting and warm atmosphere are now blighted by security bottlenecks. Second, there is little consistency in soft target screening because there is no DHS standard or enforcement. While a threat and risk assessment might help one end-user determine how to gauge the sensitivity of screening equipment, contraband that might get flagged at one screening array might breeze through at a checkpoint next door.
Screening Technology for Softer Targets
Evolv Technology is one company that has focused on advancing technology for these softer target facilities. For the non-aviation world, the company has developed a unique solution that might just help end-users rethink the checkpoint screening value proposition. Already deployed in signature venues such as Jazz at Lincoln Center, the Pittsburgh Symphony Orchestra and Gillette Stadium, the company focuses on providing reliable contraband screening while maintaining a comfortable user experience.
“We have focused so much time and resource on aviation security…we are trying to put the same focus on non-aviation physical security assets,” explains Mike Ellenbogen, Founder and Head of Advanced Technology for Evolv. “We have taken a team of people that really understand sensors and machine learning and applied those resources to non-aviation security applications. Our focus has been on creating technology that brings together best-of-breed sensors and technology to distinguish real threats from harmless objects like keys and cell phones – all at the speed of life.”
Evolv has two flagship products, the Edge and Express. The Edge is based on millimeter wave technology and looks much like theft detection paddles at a library or retail store. Edge captures a single millimeter wave image in one-hundreth of a second. As a pedestrian walks through, the device captures enough images to eliminate motion blur and produce clear and reliable sensor imagery to the central processor. The processors interpreting the sensor data have been trained to ignore harmless objects – so keys, cell phones, and jewelry stay put. Purses and jackets can be carried through. While only one person can walk through at a time, the company says up to 800 people per hour can be processed through the Edge, as compared to approximately 200 through an airport security lane.
Evolv’s newer product, the Express, operates using advanced magnetic field sensing. In this scenario, multiple people can walk through at the same time – such as a mother holding hands with a child. A free flow of people through the sensor permits up to 3,600 people to be screened per hour. A red light indicates someone has violated the sensor, and security guards are prompted with an image of the violator overlaid with a box showing the exact location of the perceived threat on that person, enabling rapid secondary screening.
Machine Learning Impacts Another Technology
The secret sauce of both screening systems is machine learning – built into the processing of the sensor data and performed by Evolv’s Cortex software.
“While machine learning often involves both passive and supervised methods, Evolv uses a process of supervised machine learning only,” Ellenbogen explains. “We are constantly collecting real data from real people on firearms and other threats, and we combine that with data on harmless objects. We use this data to continually build the computer algorithms needed to accurately process sensor data and focus on the threats of concern.”
As Evolv advances and refines the algorithms over time, they can be centrally updated. All units come with a built in 4G LTE modem for remote updates and diagnostics. “Traditional backscatter x-ray and millimeter wave machines, while excellent at imaging, are based on a stop/start sequence that reviews one person at a time,” Ellenbogen says. “They are great for looking at threats to an aircraft; however, there are unique concerns about threats to an airplane vs. the concerns of a school or performing arts center.”
Editor's Note: Learn more about Evolv at http://evolvtechnology.com.
Brian Coulombe is Principal and Director of Operations at DVS, a division of Ross & Baruzzini. Contact him at [email protected], through Linked in at www.linkedin.com/in/brian-coulombe, or on Twitter @DVS_RB.
