GSA seeks insights from combining data sets

The General Services Administration believes that by combining disparate databases, each currently stove-piped within its own line of business, it can discover some nascent trends hidden in the agency's wealth of data.

The General Services Administration believes that by combining disparate databases, each currently stove-piped within its own line of business, it can discover some nascent trends hidden in the agency’s wealth of data.

Chief Information Officer David Shive has plans to launch an enterprise-wide data management program in coming months as part of Administrator Denise Turner Roth’s strategy to make GSA more data-driven. The initiative, Shive told FedScoop in a recent interview, would give the agency a broader perspective of the less obvious trends affecting its business from end-to-end.

“What we’re finding, and what the commercial world has already found out, is when you look at those data sets as an aggregate, there are things you can know,” he told FedScoop. “Looking at building data and acquisition data, there’s probably some value in knowing the stuff that’s bought to sustain a building, and looking at those two data sets together and pulling in other data sets…when you start to pull those data sets into larger pools, and then get really, truly smart people — data scientists — to look at that, there’s things that you can know that would shape and inform the business direction of an organization.”

In the age of the Internet of Things, when new devices are made to act as sensors constantly collecting and sending data, GSA is poised to benefit from the corresponding tsunami of data, especially in its Public Building Service


“The Internet of Things — it’s a catch phrase,” Shive said. “In the public building world, it’s our reality. And we absolutely are starting to pull data from all of these sources and weed those in.”

As part of GSA’s Smart Buildings initiative, the agency is focused foremost on adding technology to the buildings it owns to make them “smarter” and more energy efficient. And that’s not just sensors that turn lights on when people enter a room. GSA has installed sensors all around its headquarters in downtown Washington, D.C., that measure things like temperature, air quality, brightness and movement of employees to develop insights on energy use, respond automatically and give building managers information to make long-term decisions.

“We measure the amount of power we’re using in various rooms, various hallways, in our data centers, in computer rooms … We also measure when people are sitting at their desks or not,” Shive said. “What we’re finding is those sensors are able to grab a lot more data than when we originally deployed them, and being able to capture that data allows you to know certain things, like how people move about a building … how they respond to things like sun pouring in a window or particularly bad air quality that nobody can see or sense or feel, but the sensors can see it.”

“What we’ve started to do as part of our Smart Buildings initiative is pull all of that data in and let smart people take a look at that,” the CIO said. “Because you can look at individual data elements and know certain things, but when you look at the data as an aggregate, smart people who know about buildings can start to sense things — patterns.”

That kind of data-rich environment, Shive said, also exists in parts of GSA’s business, like mission-assurance and acquisition, and combining those elements opens a new world of potential insights.


“We’re investing a lot of time and effort into that very thing this year,” Shive said.

While he acknowledged that GSA may be a little late to the enterprise data management world when compared to its industry counterparts, Shive said his office is bullish on developing a team and strategy to help facilitate GSA’s mission to be more data-driven. Earlier this year, he hired the agency’s first chief data officer, Kris Rowley, and he plans to hire a team of data scientists to lead the agency’s new enterprisewide data efforts.

Latest Podcasts