The research investigation analyzed the effects of environment, geometric characteristics, and highway and railroad traffic patterns with respect to grade crossing accidents in rural areas, and developed warrants for protective devices at rural grade crossings. Previous research efforts have been concerned with long period accident experience or with before-and-after studies of various protective devices. In this research, 289 grade crossings that had experienced one or more accidents were compared to 241 randomly selected non-accident grade crossings. Regression analysis was used to develop a model for predicting relative hazard, the occurrence of absence of an accident, for the sample locations. Relative hazard was expressed as a function of average daily highway traffic, average daily train traffic, roadside distractions, pavement width and number of tracks. Modification of the model permitted the development of warrants for selecting the recommended type of protective device for crossings in rural areas. These warrants are predicted on protection levels that are currently employed at rural rail-highway grade crossings in indiana. (Author/publisher)
Abstract