Modeling the impact of the curb radius on operating speeds and other surrogate safety measures using video and GPS trajectory data

Motor-vehicle turning maneuvers at intersections are often involved in crashes and serious injuries, in particular collisions involving vulnerable road users. In Canada alone, 30% of fatalities and 40% of serious injuries took place in intersections. Thus, to improve intersection safety, geometric design related treatments have been implemented and studied in past literature. These include curb radius reduction and curb extensions which aim to reduce turning speeds and crossing distances. The radius effects and treatments have been investigated for large intersections and road curves in rural areas using observed crash and point-speed measures, with limited studies examining small urban intersections. With the help of modern data collection techniques (video footage and GPS data), this study aims to fill the gaps presented in the literature on the investigation of the safety of curb radii and other geometric elements in local urban intersections using surrogate safety measures derived from vehicle trajectories and statistical regression models. More specifically, this research aims at: (1) develop a methodology to evaluate the impact of the curb radius on the speed of vehicle turning maneuvers in local intersections with small radius using Montreal video trajectory data and a cross-sectional approach. (2) To evaluate the safety effectiveness of curb radius reduction as a traffic calming treatment using a naïve before-after study and video data collected from two intersections in Toronto, Canada. In addition to speeds, post-encroachment time is used as a surrogate indicator, (3) To expand on the use of GPS smartphone data for establishing a methodology for evaluating the safety of turning movements to overcome the shortcomings of video cameras. For this purpose, GPS-based surrogate safety measures (85th percentile, median speeds, and deceleration) were extracted and modeled using mixed-effect regression models for a large data set from Quebec City.Among other results, a statistically positive and significant association between the radius and speed measures was observed for the video trajectory data. For instance, in the cross-sectional analysis, an increase of 1-meter in the curb radius results in an increase of 0.775 kph for the 85th speed and 1.208 kph for the median speed for all turns. For the before-after study, an average decrease was witnessed for all speed measurements for both intersections studied after the curb reduction treatment implementation. The countermeasure contributed to 1.6 kph and 0.783 kph decrease for 85th percentile and median speed, respectively, for both intersections when examined using the mixed-effects linear regression models. Lastly, using GPS smartphone data, the findings also support the relation between exhibited speeds and the measured corner curb radius. Using regression models of speeds, for every 1-meter increase in the measured curb radius, a statistically significant increase of 0.4 kph in both 85th percentile and median speeds was observed. Other intersections attributes, like the signalization, had an influence on the speed measures. As for the deceleration models, an increase of 1-meter in the radius resulted in a 0.01 m/s2 increase in the severity, thus, decreasing the safety of the intersection. Overall, the effectiveness of curb radius and reduction treatment was confirmed in the three different case studies. Despite the modest impacts in most of the cases, the importance of radius reductions is implied. Also, future work is required to address the limitations presented by the filtering techniques for the GPS data. Moreover, cross-calibration of the models can be used to evaluate the means of data collection by collecting GPS data for the video trajectory models and vise-versa

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.