Estimating interaction delay in bicycle traffic from point measurements

Planning for bicycle traffic is getting increasingly popular in cities, often motivated by that it is beneficial for the environment and the public health, and that it is a space efficient mode of transport. However, research on bicycle traffic has traditionally focused mainly on safety and mode choice, and less on traffic engineering topics such as investigating travel times, delays, and capacities. This study contributes to filling this gap by presenting a general method to estimate a volume -- delay function for bicycle traffic based on point measurements of passage time, speed, and lateral position, and, using estimates of the value of time in bicycle traffic, estimating the marginal cost of the interaction delay. The proposed method is based on an established method to estimate the distribution of desired speeds, and extends this to estimate the delay of the observed cyclists. The method is demonstrated using a data set from a bridge in Stockholm, Sweden, showing that there is significant interaction delay and associated cost even for relatively modest volumes, implying that interaction delay and its cost should be considered when planning for bicycle traffic.