What can a bike lane do? Performance metrics for proposed bicycle infrastructure
PI: Joan Walker, UC Berkeley; Co-PI: Offer Grembek, UC Berkeley
Abstract: Encouraging bicycle use and making cycling safer is increasingly a priority. However, forecasting the potential effects of infrastructure modifications on bicycle travel is currently a weak area, making investment decisions difficult. Performance metrics that have been developed attempt to relate roadway configurations to ambiguously defined “bicycle quality,” which typically has little empirical basis. As an alternative, we propose a framework to develop specific metrics related to the environmental, economic, and health impacts for proposed bicycle infrastructure projects.
We will use a combination of two key sources for bicycle ridership behavior: disaggregate trips from travel diary data and aggregate statistics from observed bicycle counts. The disaggregate data will be used to develop a bicycle-sensitive mode choice model that explicitly reflects bicycle infrastructure characteristics. This is used to assess mode shift and environmental and economic benefits. To assess safety outcomes, forecasts of link-based bicycle volumes are critical to estimate crashes from safety performance functions. To estimate these volumes, we will develop a model of observed counts as a function of demand estimates from the mode choice model and the local transport/land use characteristics. The expected outcome is a practical, yet rigorously developed procedure for generating performance metrics for bicycle infrastructure investments.
Quarterly Progress Reports
Click on the links below to download any of the quarterly progress reports already finished.