EPA’s data chief embraces ‘lean’ philosophy

Robin Thottungal, who came to EPA in September, has embarked on a series of mini data projects for the agency's offices.
EPA Chief Data Scientist Robin Thottungal speaks at the ATARC Federal Big Data Summit on Dec. 8, 2015. (ATARC)

The Environmental Protection Agency’s new chief data scientist is using a “lean” startup approach — a philosophy that emphasizes testing different ideas for a project and gathering feedback before creating a minimum viable product — to find better ways to use the reams of regulation, compliance and enforcement data the agency collects.

As a key part of that strategy, Robin Thottungal, who joined the EPA late last year, is working on a handful of small data projects with the agency’s offices to test each organization’s readiness to use data, while examining its existing technical infrastructure and staff.

He hopes that, through his work, EPA regulators and other officials will get better at finding patterns in the information they collect and more quickly take action.

“We have pockets of people who are already doing things to make it proactive,” he told FedScoop. “We want to make sure this is part of the DNA of EPA.”


For one project, Thottungal is working with the EPA’s Office of Land and Emergency Management to build a platform to help regulators better monitor sensors at sites on the federal Superfund list, which comprises some of the most contaminated areas of the country.

“Their challenge is they are monitoring these sites using real-time sensors,” he said. “Now, what we are thinking about is: What would it look like in the 21st century when they have more real-time sensors and they can create large data sets? What kind of technology change do we need to bring in to support that kind of a system?”

The team has already developed the technical infrastructure for the project, but they’re working out some final details, Thottungal said. They hope to launch a pilot by the end of the year.

As part of his strategy, Thottungal is also developing what he calls a “community of practice” to bring together people who are using data in the agency. So far, he’s brought in 200 people from EPA offices across the country to meet and exchange information on best practices.

“We use this as a platform for other parts of the agency to talk about the space in which they were more proactive using analytical methods,” he said.


Thottungal’s work is part of an overarching effort within the EPA to rethink how the agency uses technology. Since Chief Information Officer Ann Dunkin has taken the helm of the EPA’s Office of Environmental Information last year, she’s led what she calls a “digital transformation” — focusing on implementing the Federal Information Technology Acquisition Reform Act and helping stand up a more modern way for her agency to communicate with state regulators. She and her team also have been working to redesign the IT office while moving toward using open source technology and innovative development strategies.

In the process, she brought in the agency’s Chief Technology Officer Greg Godbout, the former executive director for the General Services Administration’s tiger team 18F, who similarly has been advocating for agencies to embrace the lean movement.

Dunkin told FedScoop Thottungal is trying to find a way to help the agency better use its “incredible wealth of data” to better accomplish its missions.

“We have all this regulatory data, we have all this sensor data, and his job is to help us use that data to be more effective — better regulators, better protect the environment, reduce burden,” Dunkin said. “So, he’s working across the agency to build a strategy that really brings all our data together.”

Agencies across government have been bringing in experts to find ways to better use their data. Thottungal came to EPA in September. And in November, the Commerce Department announced it hired its first chief data scientist. The White House also has its own chief data scientist as well — DJ Patil, who was brought on last year. Thottungal said he’s met with Patil and discussed sustaining data science in the government, building a talented workforce and improving IT acquisition strategies.


Overall, Thottungal said he hopes Patil thinks of EPA as a testbed for data science across the government.

“If you want to think about the relationship I’m trying to be build with DJ, it’s ‘How can the lessons that I’m learning here at EPA as a chief data scientist can be used by him to create these that would affect other agencies?’” he said.

Contact the reporter on this story via email, or follow her on Twitter @whitneywyckoff. Sign up for all the federal IT news you need in your inbox every morning at 6:00 here:

Latest Podcasts