Using Speed as a Safety Indicator for Vehicle Turning Movements at Urban Intersections

Author(s): Bocktor, Ledezma-Navarro, Miranda-Moreno

Slidedeck Presentation:

Slidedeck link

Abstract:

Background:

Vehicle turning maneuvers (right- and left- turns) in urban intersections pose a threat to the safety of vulnerable road users (pedestrians and cyclists) as it makes up a substantial percentage of crashes that occur. Therefore, extensive research was conducted to study the effect of various geometric design treatments to improve road safety (impacts on speeds, conflicts, crashes, etc.); for example: exclusive right turn lanes and curb extensions. The aforementioned treatments include an adjustment to the turning (curb) radius, nonetheless, few studies have examined the direct effect of the tuning radius on safety, particularly in urban stop-control intersections.

Aims:

With the help of modern data collection techniques, this study aims to fill the gap presented in the literature on the investigation of the safety of turning movements at intersections using the operating vehicle speed measures derived from video trajectory data.

Methods:

This study presents two approaches of analysis: a macroscopic and a microscopic. For the macroscopic analysis, video data was collected and processed using commercial software to obtain speed measures for 35 non-signalized intersections with turning radius less than 10.0 meters in Montreal, Quebec. Utilizing computer vision and deep learning techniques, the software generates speeds from trajectories of all users. Using these speeds, a random effect linear regression model was used to examine the relation to the turning radius along with other geometric attributes. On the other hand, the microscopic approach examines the average speed change along the turning curve of two intersections in Toronto to produce speed profiles. The microscopic study, also, evaluates a treatment of curb radius reduction which is a rather new approach taken by the city of Toronto to reduce turning speeds. A linear regression model was run to evaluate the change in speeds before and after the geometric change.

Results:

The results of the macroscopic approach show a significant statistical association between the radius and speeds. An increase of 1 meter in the turning radius resulted in an increase of 0.775 kph for the 85th speed for all turns. Moreover, the microscopic results show a decreased speed in the turning profiles after the radius reduction was implemented. Conjointly, the linear regression analysis shows a decrease of 2.037 kph in the speeds of the right turning vehicles at the changed curb (also of statistical significance).

Discussion:

As anticipated, the results show a correlation between increased radius and increased speed. This was also confirmed by the case study presented by the intersections in the city of Toronto. The reduced curb radius design provides safety for all road users as it slows down cars and decreases the distance of exposure for pedestrians.

Conclusions:

In conclusion, the study shows a significant effect of the turning radius on speed reduction. Future analysis would examine the speed every meter along the curb to capture all changes to help understand the behaviour given different treatments.