Indiana Police Use Data, Mapping To Predict Crashes


The Indiana Police Department has released a new mapping tool to the public that will allow them to find out when crashes are most likely to occur throughout the city. The map was released to general availability this week and relies on historical crash data from the the Indiana Department of Transportation (INDOT).

According to INDOT, a reduction of 1 percent of all crashes could result in savings of up to 35 million dollars per year for Indiana travelers, while protecting the lives. Indiana PD engaged the state’s Management and Performance Hub, a data shop inside government, to build the website which predicts the possibility of crashes at a given location within a three-hour window.

The map updates based on the selections, using color shading to indicate the probability of a crash occurring on that date and time in each 1km by 1km grids throughout the state. Blue indicates a low probability, yellow indicates a moderate probability and red indicates a high probability. Locations of relevant historical crashes can be seen as red and gray dots as the user zooms in on an area. Red dots represent fatal and EMS response crashes while gray dots indicate property damage crashes.

Indiana’s tool is the first tool that renders historical probabilities in this way. The Ohio State Patrol has used its own tool that works on statistical modeling software to give police a sense of where high risk areas have been in terms of semi-regular accidents. Esri geospatial mapping is providing the visualization technology for Indiana’s Crash Prediction Map.

Police departments throughout the US are using predictive analytics and statistical modeling in a variety of ways in order to predict patterns. The most benign use cases are those like this that predict crashes. In other instances, police have also used analytics data to understand personal traveling patterns, enhance surveillance and predict crime.

Those interested in trying out the map can use it here.