Cycling for transport is a sustainable alternative to using motorised vehicles for daily trips and is a key form of micromobility. Travel time is a critical factor influencing cycling route choice behaviour and uptake. Thus, it is important to understand the factors affecting cycling travel time and speed and their impact on cycling behaviour. In this study, an agent-based transport simulation model with heterogeneous cycling speeds was developed and used for Melbourne to study the impact of a hypothetical traffic signal optimisation intervention along six key cycling corridors.
Linear regression and random forest models were used to identify factors affecting cycling speed, which informed the parameters of the agent-based model. Simulation outputs showed, on average, an increase of 4.1 % in the number of cyclists on the corridors, as existing cyclists chose to use these corridors, and an average reduction in cyclists’ moving travel time of 6.2 % for those using the intervention corridors (excluding time spent waiting at traffic signals). The findings provide insights into the effects of road attributes on cycling speed and behaviour, as well as the effectiveness of interventions aimed at reducing cycling delays. These insights are valuable for developing solutions to optimise urban infrastructure for micromobility, enhancing the efficiency and appeal of cycling as a viable transport option.