Project Details

​Cycling is an environmentally friendly mode of transportation that can potentially mitigate traffic congestion and improve air pollution.  City planners and decision-makers have begun to incorporate the needs of cyclists when building infrastructures. However, factors like annual average daily bicycle volume (AADB) and cycling injury risk should be taken into consideration during the planning. As technology develops, cost-effective crowdsourced bicycle data and traditional bicycle count data from permanent continuous count stations are useful tools in investigating the bicycle volume and determining injury risk to cyclists. This research project used bicycle count data from permanent continuous counters and data from a smartphone application (i.e., Strava). Upon installation, a permanent continuous bicycle or pedestrian counter undergoes validation to determine a correction factor. This factor accounts for systematic equipment error including false positive bicycle counts that are attributed to motor vehicles and natural occlusions of a pedestrian sensor when a pair of pedestrians walk side-by-side, among other possible errors. The research team analyzed the validation and correction factor calculation methodology used in the North Carolina Non-Motorized Volume Data Program (NC NMVDP) by examining the impacts of rounding on corrected count data, appropriate temporal aggregations for applying linear correction factors, the minimum number of non-zero observations required to properly calibrate an Eco-Counter system, differences between the magnitude of correction factors as calculated using historic programmatic processes and those calculated with linear regression methods, and methods for accounting for accuracy, tolerance, and uncertainty in count data recorded by an Eco-Counter system. The research team also developed a bicycle volume prediction model based on both the corrected bicycle count data recorded by the permanent continuous count stations in the NC NMVDP and crowdsourced bicycle data from the smartphone applications. The research team also generated a ridership map according to the prediction results. In addition, the cyclist crash data were integrated with the crowdsourced bicycle data using ArcGIS software. Locations with high crash risks were identified and injury risk models were developed. Important factors that affect crash risks are selected and discussed.​