This proposal has three main components aimed at enhancing safety of railroad grade crossings and their surroundings through better understanding of 1) accident occurrence, contributor factors, and accident prediction, 2) real-time prediction and monitoring of potential conflicts at grade crossings, and 3) location and operation of emergency response resources. In Component 1, accident frequencies at crossings will be predicted not only from a macro perspective, but also including results from a new microscopic analysis to enhance predictions at single crossings, and to conduct corridor and regional studies. The micro analysis will include a dynamic tree structure for discovery and a temporal probabilistic analysis of accidents. New accident prediction models will be developed using this multi-scale analysis. Component 2 will study train delays to accurately estimate train arrival times at grade crossings, leading to effective management of emergency response resources and supporting driver alerts at un-signalized grade crossings in personal navigation devices. The forecast of arrival times will involve train positioning information for delay predictors, and developing historic and real-time statistical algorithms. Component 3 will develop tools to determine optimal deployment and capacity of emergency response facilities, considering uncertainty from incidents, network capacity reductions, and grade crossing closures during and after the incidents. This component will also develop models to optimize coordination and decision making between jurisdictions with different objectives. Knowledge and tools developed with these three components will enhance safety at grade crossings through short and long term implications in planning and operation activities conducted by public agencies and the railroad industry.
Start date: 08/2014
Project Status: Active
Research Area: Transportation Safety and Traffic Flow