Monash University Accident Research Centre - Report #13 - 1990
Authors: B. Corben, C. Foong
Full report in .pdf format [1.9MB]
The Pro-active Traffic Engineering Safety Study was undertaken to explore the opportunities to employ a predictive approach to improving safety through modifications to the road and traffic environment. This approach would operate alongside and complement the existing Accident Black Spot Program.
Right-tum-against crashes at intersection signals, accounting for an estimated 700 casualty crashes in Victoria annually, was an aspect identified as potentially suited to a pro-active approach. The study involved the collection of selected information on the accident histories, and physical and traffic characteristics at 53 signalised intersections. Statistical analysis of the data was carried out to determine whether any of the physical and/or traffic variables were associated with higher accident occurrence and to develop one or more models to predict right-turn-against crash occurrence for a given intersection approach.
Right-tum-against accident occurrence was clearly associated with the volumes of conflicting vehicles. The analysis also led to the development of a seven-variable, regression model for predicting right-tum-against crash occurrence. The model explained 85% of the variance in accident occurrence, though validation is still required.
As with existing techniques for capacity analysis, the model has considerable potential to be used in the design of intersection signals to objectively determine the need for fully controlled right-tum phases, or other appropriate countermeasures, to reduce the likelihood of a serious accident record developing. Guidelines for signal designers and system operators are proposed.