Advanced Technologies in Border Protection

This post is also available in: עברית (Hebrew)

The U.S. Customs and Border Protection (CBP) will integrate into its operations three innovative solutions that address pressing challenges faced by the Department of Homeland Security. The technologies were developed by three startups with the DHS Science and Technology Directorate (S&T) Silicon Valley Innovation Program (SVIP).
A technology developed by Tamr was designed for the enhancement of the Global Traveler Assessment System (GTAS), a non-proprietary computer application available to partner countries that provide the capability to screen foreign travelers.
CBP developed GTAS in accordance with UN Resolution 2178 to combat foreign terrorist fighters by using industry-regulated traveler information. Tamr’s software allows for improved entity resolution — the analysis of multiple datasets to determine matches between entities, datasets and possible relationships — within GTAS. This technology is now fully incorporated into the system, according to dhs.gov.
Another startup, Echodyne, created the Metamaterial Electronically Scanning Array (MESA) radar system. This system uses metamaterials — engineered, artificial materials with properties not found in nature — to build a new architecture for an all-electronic scanning radar system. The use of metamaterials means MESA has significantly lower cost, size, weight and power-usage than other radar systems.
The compact, lightweight radar units have the potential for multiple applications in a variety of border security scenarios.
In addition to the testing of the technology’s capabilities in improving situational awareness, the solution is currently being used as the primary detection and cueing component on autonomous surveillance towers currently deployed in the San Diego Sector. These towers are being piloted with the potential of incorporation into border surveillance programs.
The third company, DataRobot, applied automated machine learning (AML) to GTAS to expedite the development of predictive models.
Currently, the time required to develop predictive models places those models at risk of being outdated before they are completed. By applying AML to this development process, DataRobot is able to produce models faster and more accurately. AML is also easier to use than traditional machine learning — it can automatically complete complex tasks while simplifying the user experience.
The technology is now being used to help CBP conduct the counter-narcotics mission, identify ways to improve the facilitation of lawful trade and travel, and develop and test synthetics datasets.