Microscopic analysis for accident data at railroad grade crossings

Analysis of accidents at railroad grade crossings has traditionally been focused on developing models to determine overall trends from a large number of locations over time. While this approach results in models that can provide significant information at the macro level, it may not capture factors that can be only observable from a microscopic perspective. This paper uses a simple micro approach to enhance the analysis of accidents at railroad grade crossings. Trends identified in the microscopic analysis are based on detailed accident information such as location and direction of vehicles and trains, crossing layout and its surroundings, and driver demographics, among others. A micro analysis provides a different perspective on the selection of items to be improved and on the identification of crossings for funding allocation. Examples of micro analysis on two crossings in Illinois highlighted trends on the direction and location of accidents, potentially related to crossing angle and an interconnected intersection; and trends on the driver population and nearby housing facilities. These results could not be obtained using standard macro models alone.