The problem of red-light-running is widespread and growing; its cost to society is significant. However, the literature is void of quantitative guidelines that can be used to identify and treat problem locations. Moreover, there has been concern voiced over the validity of various methods used to identify problem locations, especially when automated enforcement is being considered. This paper documents the development of a procedure for identifying and ranking intersection approaches with the potential for improvement in the area of red-light-related crashes. One component of this procedure is a safety prediction model. A sensitivity analysis of this model indicates that red-light-related crashes decrease with an increase in yellow interval duration and a reduction in speed limit.
Abstract