University research team releases app to help Flint residents assess water risk levels
University of Michigan researchers have released an app that helps Flint residents assess lead contamination levels by utilizing functions related to water testing and providing infrastructure-related information. The team also worked with Flint Mayor Karen Weaver on data analytics for lead pipe removal.
The app, released in early December, gives residents information about distribution centers for water and water filters, locations where lead has been found in drinking water, areas where infrastructure is being replaced within Flint, the likelihood that water in a home or particular location is contaminated and instructions for water testing.
Jacob Abernethy, professor of electrical engineering and computer science, was one of the faculty members approached by Google to create an app that would assist those affected by the Flint water crisis, which led to the creation of MyWater-Flint app.
“What came about was an idea for an app that would be useful for (Flint’s) citizens to access a central hub of information about the water crisis.” Abernethy said. “(Google) had people who did software development, but they didn’t have people who did sort of data science, machine learning analytics stuff. With the help of a grad student a year and a half ago, I helped initiate this thing through Michigan called the Michigan Data Science Team.”
Abernethy worked with a group of students who organized and analyzed the data on Flint’s water crisis. They also worked with other professors and had support from Google throughout the process.
“We’ve made a collaboration with professors at UM-Flint, and we basically focus on all these data and analytics stuff and they focus on the software development side of things,” he said. “We’ve had weekly meetings with folks at Google who’ve helped with (user interface) and other issues and give advice on development.”
Cyrus Anderson, Michigan Data Science Team member and University alum, said they worked on identifying where the highest lead levels were while developing the app and identifying lead risk levels.
“It wasn’t clear what was causing the lead problems in Flint, so what the team first did was presenting where the high levels (of lead) were likely to be,” Anderson said. “From there, we built models predicting these high levels in order to find the risk factors associated with different aspects of the housing. Part of our project has also been figuring out where exactly the lead pipes are likely to be and predicting whether or not a house will have lead pipes or not.”
Abernethy pointed out that lead poisoning is not an issue that affects just Flint; many states do not recognize universal blood testing regulations, resulting in states like Pennsylvania having childhood lead-poisoning rates nearly double Flint’s.
“They found that Flint’s water problems are not actually unique to Flint — there are a large number of census tracts in the U.S. that have blood lead levels above the monitoring level,” Abernethy said. “Flint may only be the tip of the iceberg. The belief is that the idea of this information can be applied to different municipalities around the country — we definitely think this is very possible.”
The lack of information about the location of lead service lines and discord surrounding the need for partial or full pipelines has hindered the FAST Start program, Mayor Weaver’s pipe removal initiative. Abernethy and his team worked on this initiative as well, using their data science experience to help overcome these issues.
“The government of Michigan put forward approximately $25 million to do lead pipe removal in Flint, and one of the big recovery efforts was to get rid of all these service lines,” he said. “It’s quite an expensive process, trying to determine where lead is, and certain parts of the city had a lot more lead than other parts of the city.”
Abernethy's team used the data provided by FAST Start in order to provide useful information to those in the program. The group analyzed a variety of factors and have provided access to the data through the app.
“And, of course, there are a lot of working variables here, like whether or not the property was occupied or the likelihood of the homeowner replacing the service lines on their own and not informing the government,” he said. “Some of the data that’s been collected from this FAST Start program, we are actually making available in the app.”
This implementation can be viewed on the MyWater-Flint homepage. Users can view what infrastructure is being replaced and what is currently being worked on. This can help Flint residents identify which areas have high lead risk levels.
LSA sophomore Aliah Richter said she believed it was important to utilize the prevalence of social media and developing technology in situations like the Flint water crisis.
“I think the entire concept is really innovative,” Richter said. “We’re so connected by technology today that pretty much the one thing that could connect all of us is phones and computers. So having an app like that and being able to interact with other people in your area that are also being affected by the situation — I think it’s really important for residents to be able to identify areas that are more likely to have high levels of lead, so I really like the idea of this app.”