Analyzing Bicycle Riding Characteristics Based on Naturalistic Urban Drone Observations: Free Riding, Following Behavior, and Overtaking Maneuvers

Despite the increasing share of nonmotorized road users, several aspects relating to modeling and planning insights have not yet been sufficiently addressed in research. The behavior of bicyclists in simulation models often differs from their real-world riding style. Therefore, to identify realistic parameters, a solid database is needed to be able to address these issues. Progressive data collection methods allow for tracking individual trajectories of nonmotorized road users and analyzing their behavior. This can then be used to improve existing modeling approaches and gain further insights. In this work, we present an analysis of cyclist riding characteristics in an urban area. A trajectory data set collected by a swarm of drones in the city of Munich was used to investigate (i) acceleration behavior after a traffic light stop, (ii) following and free-flow riding scenarios, and (iii) overtaking maneuvers. In comparison with existing studies, despite great heterogeneity among the cyclists, the overall dimensions of speed, headway, and spacing were found to be similar. In addition, it was possible to gain insights relating to the development of different parameters over time, which had been difficult to determine with previous data sources. This work therefore contributes to the common understanding of cyclists’ behavior, adding another element of bicycle characteristics to the fundamentals of modeling approaches.