Date of Award:
Master of Science (MS)
Civil and Environmental Engineering
Recent trends in crashes indicate a dramatic increase in both the number and share of pedestrian and bicyclist injuries and fatalities nationally and in many states. Crash frequency modeling was undertaken to identify crash prone characteristics of segments and non-signalized intersections and explore possible non-linear associations of explanatory variables with crashes. Crowdsourced “Strava” app data was used for bicycle volume, and pedestrian counts estimated from nearby signalized intersections were used as pedestrian volume. Multiple negative binomial models investigated crashes at different spatial scales to account for different levels of data availability and completeness. The models showed high traffic volume, steeper vertical grades on roads, frequent bus and rail stations, greater driveway density, more legs at intersections, streets with high large truck presence, greater residential and employment density, as a larger share of low-income households and non-white race/ethnicity groups are indicators of locations with more pedestrian and bicycle crashes. Crash severity model results showed that crashes occurring at mid-blocks and near vertical grades were more severe compared to crashes at intersections. High daily temperature, driving under influence, and distracted driving also increases injury severity in crashes. This study suggests potential countermeasures, policy implications, and the scope of future research for improving pedestrian and bicycle safety at segments and at non-signalized intersections.
Rahman, Md Rafiur, "Network-Wide Pedestrian and Bicycle Crash Analysis with Statistical and Machine Learning Models in Utah" (2022). All Graduate Theses and Dissertations. 8423.
Copyright for this work is retained by the student. If you have any questions regarding the inclusion of this work in the Digital Commons, please email us at .