In order to estimate how some countermeasure or treatment might affect the safety of an entity it is necessary that we have thecapacity to: • Estimate what the expected number of consequences (fatality, injury and PDO collisions) would have been in the 'after' period had the treatment not been implemented, and • Compare this estimate with what the expected level of safety would be in the 'after' period if the treatment had been implemented. For example, in the case of unsignalized intersections, we would like to be able to measure how the safety of a particular unsignalized intersection would be changed, and by how much, if it were signalized. The capacity to measure this safety differential would provide the information needed in the decision-making process for determining whether the installation of a traffic signal is warranted (from a safety perspective) at a particular unsignalized intersection. This report outlines a methodology for measuring the safety of unsignalized and signalized intersections and comparing their relative levels of safety, and proposes an objective, quantitative process for determining whether or not a traffic signal is warranted at an unsignalized intersection. A detailed description is provided of a state-of-the-art scientifically based methodology for measuring the differential in safety levels expected between signalization and non-signalization for a particular unsignalized intersection over a 20-year evaluation period. The methodology is based on an extension of the Empirical Bayes Approach coupled with proven statistical modeling procedures for estimating accurate and unbiased safety impacts of signalizing an unsignalized intersection. A comprehensive process including all phases and tasks for the design and development of a modeling system is proposed. This system, once implemented, would provide the necessary information for making rationale decisions on whether a traffic signal is warranted at an unsignalized intersection or not. We therefore recommend that the above methodology and process be considered for establishing a policy and procedures for traffic signal safety warrants. This would ensure that a system for using traffic signal safety warrants would be based on the 'need' according to the actual safety performance of the particular unsignalized intersection being considered for signalization. (Author/publisher)
Samenvatting