This paper presents a methodology to evaluate the effectiveness of a bicycle treatment (bike boxes) at intersections using a before-after surrogate safety analysis based on longitudinal video-data analysis. As a surrogate safety measure, cyclists’ red-light violations are quantified for two periods before and two periods after the installation of a bicycle box at a signalized intersection in Montreal. For this purpose several hours of video were collected before and after the installation of the treatment. Based on the video data, red-light violations and potentially associated factors were collected for each cyclist that crossed the intersection, such as sex, age, group size, use of helmet, whether a cyclist stopped before crossing, vehicle-cyclist gap, etc. Violations with a short vehicle-cyclist gap were classified as dangerous (i.e., those situations in which cyclists pass the intersection during the red phase with a small vehicle gap). For the data analysis, a multinomial logit regression technique was used to identify the factors that increase or decrease the probability of cyclist violations as well as their changes over time. Both raw estimates and model estimates show that the presence of a bicycle box has a significant impact on the total number of cyclists’ violations; however, the impact on the number of dangerous violations is not clear. More video data from other intersections before and after the treatment implementation is required to validate these preliminary conclusions. Moreover, the video-data generation and surrogate approach proposed here can be applied to the evaluation of other bicycle treatments. (Author/publisher)
Samenvatting