Road rules, infrastructure and risk − how law-abiding are e-scooter riders?

Use of both privately owned and shared e-scooters is continuing to proliferate in cities along with concerns about rider and pedestrian safety as well as compliance with road rules. The goal of this research was to understand e-scooter road rules compliance and contributing factors focusing on helmet use and speed limits. This paper documents a study of micromobility and active transport users across two years with observations captured by traffic cameras at eight sites in Brisbane, Australia. Machine learning was used for initial processing of more than 600,000 observations of e-scooter users, cyclists and pedestrians. In addition to transport mode (pedestrian, cyclist, e-scooter), observations included e-scooter ownership (shared vs . private), speed, helmet use, infrastructure type (footpath, traffic lane, bike lane, separated cycle track) as well as urban form and density (i.e., central business district, urban, suburban). Logistic regression is used to provide a robust model of the factors associated with e-scooter users road rules compliance. Speed compliance is highest on separated cycle tracks and is positively correlated with urban density and greater head protection. E-scooter riders wearing helmets rode faster than those not wearing helmets, and those with full-face helmets rode the fastest. Separated cycle tracks are associated with higher rates of helmet compliance than other infrastructure types for shared e-scooter riders. Private e-scooter riders were most likely to wear a helmet when riding on roads or bike lanes, potentially compensating for greater speeds on those infrastructure types. Overall, these findings reveal separated infrastructure encouraged road rules compliance and provide evidence of e-scooter users demonstrating risk compensation decisions around helmet use, speed and riding locations. The results here promote the importance of building separated cycling infrastructure in order to encourage road rules compliance.