Modeling Bicycle Facility Operation: a Cellular Automaton Approach

State-of-the-art methods for the safe and efficient design of bicycle facilities are based on difficult to collect data and potentially dubious assumptions regarding cyclist behavior. Simulation models could offer a way forward, but existing bicycling models in the academic literature have not been validated using actual data. This paper attempts to address both of these shortcomings simultaneously by conducting a field study to obtain real-world bicycle data and implementing a simulation using a multilane and inhomogeneous cellular automata model to reproduce the observations. The resulting model is found to emulate field conditions while possibly under-predicting bike path capacity. The analysis indicates that the model's potential as for planning could be high given additional work on the underlying model specification and the collection of additional data. Given current concerns surrounding regional air pollution, climate change and urban congestion, this research is timely. If we begin to see more substantial mode shifting to non-motorized modes, this and similar models could become standard tools in the city or regional planner's toolkit. After a discussion of the context in which this research is being conducted, we review the relevant literature on bicycle facility design and bicycle traffic operation, before summarizing the real-world data collection methods and results. The simulation model is then presented along with results and discussion comparing the modeled to observed data. We conclude with suggestions for future work in data collection and model development.