Daniel Neill had an unexpected person reach out to him. A commander in the Chicago Police Department, Jonathan Lewin, had come across research recently presented by Neill, director of the Event and Pattern Detection and Pattern Laboratory, and his colleague, Wil Gorr, professor of public policy and information systems. The Heinz College professors used historical data and predictive analysis to show that smaller crimes, like fights or vandalism, can lead to more violent crimes. Lewin thought the model could reduce crime in Chicago.
That led Neill to develop CrimeScan, a software program built on their research and on 911 call data that could offer predictions on what’s going to happen on a particular day on a particular city block. Chicago police began using CrimeScan, and it has been effective in helping police predict crime hot spots and target patrols accordingly. At the same time, Gorr completed research showing that crime prediction and police interventions can be effective in preventing serious violent crimes even in residential areas where fear of crime is the greatest but otherwise police have not had good approaches for crime reduction.
Given the success, Chicago wanted more predictive analysis—from traffic problems to sewage issues. After receiving a $1 million grant from the Bloomberg Foundation to build a citywide predictive analytics platform, Chicago city officials reached out to Neill once more.
“We were thinking across every city department and thinking what they might benefit to predict,” says Neill.
Neill created CityScan, a software package that uses 311 call data to predict city problems. For example, he found that sanitation complaints, like a water main break or uncollected garbage, could be a predictor of rats. The Department of Streets and Sanitation liked the hypothesis and are currently using CityScan.
Meanwhile, Neill and Gorr are working with Chicago’s Department of Innovation and Technology to discuss new collaborations, including transportation and public health. Neill is planning to develop free and open-source software, so that other cities can implement CityScan’s predictive analytics as well.
