DHS ‘very pleased’ with Biometric Technology Rally, results will drive future R&D
The Department of Homeland Security’s (DHS) Science and Technology Directorate (S&T) conducted its first Biometric Technology Rally in March at S&T’s Maryland Test Facility (MdTF) to test face and face/iris recognition systems. The MdTF designed a standard security checkpoint to test the capability of biometric identity systems to acquire and match images from a diverse volunteer population in a realistic time constraint.
The Biometric Technology Rally was built off of other recent S&T partnerships with TSA and Customs and Border Protection (CBP) to improve the speed and reliability of biometric systems.
DHS said, “Balancing speed and security at checkpoints, like airports, is essential to ensuring safe, reliable travel. Many of these checkpoints are increasingly using biometric technology to improve speed and reliability. While recent improvements in biometrics have lowered failure to match rates, many systems fail to quickly acquire biometric information in the first place.”
CBP Deputy Executive Assistant Commissioner John Wagner said, “At Customs and Border Protection, our biggest challenge over the past several years has been collecting biometrics for people departing the United States where, unlike the arrivals process, we operate without a dedicated, secure environment.”
Consequently Wagner said, “We are looking to work with partners and stakeholders to develop a process that doesn’t shut down travel, but enhances security and efficiency.”
Meanwhile, S&T also conducted a Stakeholder Demonstration Day in which 100 representatives from more than 20 government components and private organizations visited the MdTF to learn more about the industry participants and their systems. Organizations represented included DHS components, the Department of Defense, Department of State, National Institute of Standards and Technology, General Services Administration, United Kingdom, and various airports and airlines.
Melissa Conley, Senior Advisor for the Transportation Security Administration’s Office of Requirements and Capabilities Analysis, said, “Over the 13 years that I’ve been at TSA, this is one of the first technology rallies that has offered such a diverse suite of capabilities for us to look at in a short period of time. It was very informative.”
Results from the rally will be used to inform DHS procurement activities such as operational tests, pilots, and system acquisitions. This event was the first in a series of S&T Biometric Technology Rallies that will focus on “different biometric modalities and areas of interest.”
“The results of this rally will help inform the focus of future MdTF testing events, and the team will walk away with a number of ideas about how to make the next rally even more effective,” said Vemury.
“Today, TSA got to see a bunch of different solutions, a bunch of different use cases, and that’s really important market research for us as we move forward in implementing biometric technology,” Conley said. “S&T is doing all of the leg work for us to make sure these algorithms and these technologies and these companies meet certain thresholds so we can streamline our procurement process.”
DHS said the effectiveness of the test “measured performance of the biometric system, including failures to acquire, process, and match images.”
S&T challenged industry biometrics systems to identify 99 percent of all volunteers using the system in under 20 seconds.
Industry was challenged to provide systems with at least a 95 percent user satisfaction rate. The volunteers used a kiosk to rate their happiness after using each system. S&T challenged industry participants to meet specific objective performance goals along three interlinked categories of evaluation metrics: effectiveness, efficiency, and satisfaction.
In addition to these technical metrics, S&T also wanted to develop efficient, easy to use systems that are “enjoyable to use as measured by satisfaction feedback.”
“Efficiency measured the average time volunteers spent using each system,” DHS explained, saying, “S&T challenged industry to design intuitive systems that volunteers could understand and use in just five seconds.”
During the course of the 10-day event, 364 volunteer test subjects “representing a diverse range of ages and demographics performed a total of 4,368 individual system assessments,” DHS said, noting, “Innovative systems from 11 industry participants representing 10 countries from around the world were selected to participate in the rally.”
“Recent improvements in biometric matching technology have greatly reduced biometric matching error rates; however many systems still fail to quickly and reliably acquire biometric information in the first place,” said Arun Vemury, DHS S&T program manager for the agency’s Borders and Maritime Security Division led by S&T’s Biometric Technology Engine (BT-E), which was developed in 2017 “as a method of rapidly screening emerging biometric technologies to facilitate operational insertion,” DHS explained, adding that, “The BT-E acts as a knowledge base for data related to biometric technologies examined by DHS S&T with the goal of informing future large-scale acquisitions and sharing this data within DHS, across other agencies, and the HSE.”
“This event aims to further reduce the failure rate by challenging industry to develop more intuitive, faster, and more reliable face and iris collection systems,” Vemury said.
“The results of this rally will help inform the focus of future MdTF testing events, and the team will walk away with a number of ideas about how to make the next rally even more effective,” Vemury said.
Conley agreed, saying,“TSA got to see a bunch of different solutions, a bunch of different use cases, and that’s really important market research for us as we move forward in implementing biometric technology. S&T is doing all of the leg work for us to make sure these algorithms and these technologies and these companies meet certain thresholds so we can streamline our procurement process.”
The testing involved 64 participants; 4,368 system assessments; 7,500 face images collected; more than 2,000 iris identifications; more than 15,000 face identifications; and more than 20,000,000 comparisons.
We were “very pleased by the outcome of this rally,” Vemury said. “Not only will the results be beneficial for understanding to capabilities of cutting edge systems and driving future research and development, but industry participants gained a better understanding of the strengths and possible areas for improvement of their biometric systems, having now used them in a systematic, rigorous, and realistic, high-throughput environment.”