Michigan prepares to analyze data collected through IoT devices
As Michigan pulls in an ever-expanding stream of data through connected devices, Deputy Chief Security Officer Paul Groll is hoping to make that information useful with a focus on analytics.
At the National Association of State Technology Directors’ annual conference, Groll detailed his state’s efforts to put sensors in the field and embrace the Internet of Things. While he noted that several of Michigan’s agencies are increasingly gathering new types of data from these connected devices, he’s aiming to move the state to the next level by examining what that information actually means.
“You need actionable intelligence to come out of this, otherwise you’re wasting your time,” Groll said. “It’s easy to bring in a lot of information and put it in a data lake and hold it because you don’t want to lose it, but it only brings you so much value until you put it through analytics of some kind. It needs to change behavior to be worth it.”
Groll points to his state’s Department of Transportation as the main agency in Michigan trying to fully take advantage of the IoT. Indeed, he notes that the department already has “thousands of devices” collecting data on the state’s transportation infrastructure, and adds that they’re “looking to have over a million sensors throughout the state” by 2022.
Additionally, the department is working with the University of Michigan’s connected vehicle research center in Ann Arbor, and he expects “there’s going to be a lot of data coming out of those” projects.
But now that they’ve started the data collection process in earnest, Groll is excited about the chance to start putting that information into action. One idea the department is “kicking around” is building a smartphone app that can work with a device’s accelerometer and GPS to help map the quality of the state’s roads, Groll said.
“We have some of the worst roads in the country,” Groll said. “This could help create a map to get some idea of what needs attention first.”
[Read more: More IoT devices mean states must get smarter about management — panel]
With all the sensors the department’s attached to the state’s bridges, he also sees immense analysis potential in that area.
“Some bridges in the state are hundreds of years old, so we need to understand what’s going on with them,” Groll said. “We have sensors out in the thousands. We need to bring that back through big data channels, store it locally through some sort of big data repository and start to do analytics.”
He suggested that “predictive analytics” through “machine learning” algorithms could be one avenue for the state to pursue to tackle that problem efficiently. John Andersen, IBM’s chief architect for national programs, pointed out that his company has been working on just that sort of technology to provide government agencies a way to keep up with the new influx of data from the IoT.
“The onslaught of data will only be increased with the IoT, yet there’s still just 24 hours in a day,” Andersen said. “Cognitive computing could be the answer. You can ask something like Watson a question and get a response that’s helpful.”
He noted that IBM’s started working with federal agencies, feeding new regulations into their system to understand if they conflict with existing rules and analyze their economic impact, and he sees no reason why states and localities couldn’t embrace that same sort of technology.
“These devices produce a lot of information, but there are still golden nuggets in there,” Andersen said. “It’s just harder to find them.”