The relationship between geometric inconsistencies and accident experience on two-lane rural highways is discussed. Inconsistent geometric features pose operational and safety problems because of disparities between the expected and actual driver workload and attention level requirements of the features. U.S. design policy, which is based on the design-speed concept, does not provide a systematic basis for preventing geometric inconsistencies. Several European countries and Australia, however, have recognised this limitation and have supplemented the design-speed concept with consideration of operating speeds along alignments. The two basic approaches for identifying and quantifying geometric inconsistencies are discussed: one is based on operating speeds and the other, on driver workload. Procedures based on operating speeds are used in Europe and Australia, and two such procedures have been developed in the United States. These procedures use the change in operating speeds to quantify the severity of geometric inconsistencies. Their application is limited to evaluation of horizontal alignment, however. Driver workload has the advantage of being applicable to most types of inconsistencies but the disadvantage of not being directly measurable. The results of a preliminary study of the statistical relationship between geometric inconsistencies (quantified in terms of driver workload values using a procedure developed by Messer et al.) suggest that driver workload values are good predictors of accident experience on two-lane rural highways.
Samenvatting