IDENTIFYING AT HIGHWAY-RAIL GRADE CROSSING HOTSPOTS IN CANADA

This research presents a risk-based Hotspots identification model at highway-rail grade crossings in Canada. Two sets of models were developed to predict collision frequency and consequence at individual crossings. A two–dimensional graphic approach was adopted to combine these two models together to predict the risk at each crossing. Hotspots based on collision history tended to be wide spread in Canada, while the Hotspot list based on model prediction tended to cluster in the Prairie Region of the country for frequency and the eastern provinces of Ontario and Quebec for consequence. Most Hotspots identified based on collision frequency are located in urban areas with high AADT, while the Hotspots by collision consequence are mostly located in rural areas with high train speed.