CBP considers open source license plate readers to save cash

Cameron Cross/Shutterstock.com

Agency plans to replace existing brand name tech with parts that can be operated and obtained without the contractor.

This story has been updated to include comment from ICE.

In a cost-cutting move, the Homeland Security Department wants to replace more than 500 brand-name systems that identify vehicle license plates at border stations with generic technology.

The plan is to phase in a more comprehensive open architecture model that is not dependent on any single company’s machinery or code when existing equipment breaks down at any of the 86 checkpoints nationwide, department officials said.

DHS Customs and Border Protection currently uses proprietary systems from Unisys and Perceptics for all inbound and outbound lanes at land crossings, and proprietary systems from ELSAG at Border Patrol checkpoints, CBP spokeswoman Joanne Ferreira said.

This week, CBP began tapping the market for ideas on workable open source versions of the plate readers. The agency is willing to use some proprietary technology, provided the licensing arrangement lets the government maintain and operate the technology on its own.

The current systems’ “proprietary elements should be eliminated or otherwise confined to well-defined elements [in a way that allows] others to operate and maintain deployed LPR components without the need for [manufacturer support],” stated an industry survey released Dec. 17.

In October 2010, CBP awarded Unisys a deal worth up to $350 million for five years to supply plate readers that will register cars traversing the Canadian and Mexican borders.  

The future system would consist of components patterned after freely available standards so that CBP can buy interchangeable parts from a range of vendors, agency officials said.

Today’s readers “will in time reach the end of their useful life and will require a technology refresh,” therefore, CBP is researching options “for a cost-effective replacement or update of the current LPR systems, or select portions of those systems,” the industry questionnaire stated.

Officials added, “one key objective is to reduce the total system life-cycle cost by moving to a nonproprietary, open architecture solution.” 

Whichever setup the agency chooses, it must be capable of functioning around-the-clock from various viewpoints, such as roadside or overhead, according to the notice. The system should be able to snap images showing a passing vehicle’s plate, as well as detect the plate’s country of origin and characters.

Agency officials also are seeking input on potential stumbling blocks and farsighted possibilities. For instance, CBP is interested in the feasibility of deploying a product that combines onsite license plate data with “with machine learning/pattern recognition to continuously improve read performance and accuracy.”

Separately, DHS Immigration and Customs Enforcement this summer had planned to buy similar readers from Vigilant Video for trailing illegal immigrants facing deportation, through a non-competitive acquisition. But ICE pulled the contract after another company came forward claiming to offer the same services.

On Friday, agency spokeswoman Danielle Bennett said, “ICE is continuing to evaluate its needs regarding this database tool and is currently planning to seek procurement at some point in the future.”

(Image via Cameron Cross/Shutterstock.com)