Analysis of Drivers’ Behaviour at Different Road Safety Treatments in the City of Ottawa

Abstract
Investigation of the effectiveness of different road safety treatments is essential to maximize the returns on road safety initiatives and expenditures. An investigation based on collision data alone would not provide the complete picture of the effectiveness of such treatments as collisions are known to be random and rare events. Therefore, such an investigation should also involve an evaluation of drivers’ behaviour at the different treatments. This paper presents a comprehensive analysis of drivers’ behaviour at road safety treatments that target improving safety at intersections and midblock or road segments. Namely, the paper examines roundabouts and red-light cameras (RLCs) as intersection safety treatments and Dynamic Speed Display Signs (DSDSs), speed cameras, and road centerline signs as segment safety treatments. Drivers’ behaviour at each treatment type was examined using video data collected in the City of Ottawa. The video data collection covered 15 intersections, including 5 roundabouts and 10 signalized intersections (five with RLCs and five without RLCs) and 15 road segments including 5 segments for each treatment type, which are speed cameras, DSDS, and road centreline signs.
The safety impacts of segment treatments on driver behaviour were evaluated using a change of vehicle speed as a surrogate safety measure. On the other hand, the surrogate safety measures examined at intersection treatments were deceleration rate and Time-To-Collision (TTC) at RLCs and were Post-Encroachment Time (PET), TTC, and speed at roundabouts. The analysis indicated that RLC sites experienced more severe rear-end conflicts and harder deceleration compared to non-RLC signalized intersections. The PET and TTC analysis results for roundabouts showed a relationship between the ratio of number of conflicts to traffic exposure, termed as risk ratios, and collision frequency. The segment treatment analysis showed a change in speed from upstream to downstream speed cameras and DSDSs or before to after the installation of road centerline signs. A speed reduction of 12.18%, 9.39% and 8.9% in mean speed was observed at speed cameras, DSDSs and road centerline signs, respectively. The speed cameras also had a 40.40% compliance rate increase which is the highest increase within the segment treatments.