How CBP saved thousands of staff hours with RPA

Late last year, Customs and Border Protection used 50 bots to move 30 terabytes of email to a new email system in one of the largest robotic process automation (RPA) projects at the agency, according to the provider of the technology.

In the latter half of 2018, IBM provided CBP's Office of Information Technology with a series of RPA bots to move the equivalent of 350 million archived emails from one system to another, IBM Homeland Security Client Lead Jonathan Riksen said.

The data had been slated to be moved by a team of CBP employee volunteers, Riksen said. RPA reduced the processing time for a terabyte of data from two months to one day.

CBP went through DHS’ First Source contract vehicle for the IBM RPA solution.

Because the President’s Management Agenda looks to shift human workers from low-value to high-value work, agencies are increasingly leveraging RPA, according to Riksen. Although the agency has been investigating how to use robots for mission work, he said, the big email transfer job helped open CBP managers’ eyes to everyday, back-office applications for the technology.

Other technology vendors are seeing similar effects, as agencies look to simplify tedious processes and free up humans for more-substantive, mission-oriented work.

Cloud computing, said Dave Levy, federal government vice president at Amazon Web Services, also is a “huge enabler” for the federal workforce. It has allowed federal IT workers to move away from having to maintain physical servers in data centers to more important mission-oriented tasks.

“The federal workforce is hungry for new sets of technologies,” such as RPA, artificial intelligence and machine learning, he said in an interview with FCW.

In a presentation at the AWS federal cloud summit in Washington, company consultants described a project to help a federal agency transfer its internal “media wiki,” which housed its administrative processes, from agency servers to the cloud.

The agency, which AWS consultants didn’t name, chose to strip the application down, “refactoring” the source code to make it cloud native instead of just re-platforming as-is.

The process used for the app transfer, said Cole Hubbard, a consultant at AWS, has become a model to move hundreds of other apps to the cloud at that agency.