Browsing by Subject "crash prediction"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Hennepin County Pedestrian Crash Study
    (2023) Ackerman, Ryan; Johnson, Isak; Murphy, Daniel; Trejo, Tristan
    Our study analyzed historical pedestrian crashes throughout Hennepin County and ranked crash locations based on crash occurrence over a ten-year period (2012-2021). For analysis purposes, crashes were split into two categories: intersections and midblocks. Crashes primarily occurred in urban areas, and collisions resulting in fatal injuries were rare. We created a tiered ranking system to group together locations with similar levels of crash occurrence to guide potential county improvement projects. Using ArcGIS Pro, we developed crash point maps to spatially represent crash locations and severity in each Hennepin County Commissioner District. We then created Safety Performance Functions (SPFs) by conducting a statistical analysis of crash data using a Negative Binomial Regression model. The variables we chose for statistical analysis were identified in previous studies as statistically significant variables that influenced pedestrian crashes. We used our SPFs to predict future crash locations and crash severity at intersections and midblocks over the next ten years. Our SPFs predicted fewer crashes at intersections and midblocks over the next ten years than the actual number of crashes over the tenyear study period. This can be partially attributed to our model, which was relatively weak, but can also be attributed to a lack of data. In particular, pedestrian count data would likely have increased the accuracy of our model, but this is not easily accessible. Our study opens the door to future research by transportation planning professionals who can make proactive, informed decisions about reducing pedestrian crash risk throughout Hennepin County based on our research.