Dynamic method to identify and visualize accident contributing factors at highway-rail grade crossings

Highway-Rail grade crossings are important locations on a highway network and safety at these locations is a critical issue for both highway and railroad network. Even though safety at grade crossings have improved over the years, over 200 people still lose their lives every year at grade crossings across the United States. Also, the injury and fatality rates are significantly higher for grade crossing crashes than other types of traffic crashes. Therefore the need for improving the safety at grade crossings is relevant today and more so especially since the demand in both highway and railroad systems continue to increase. Safety improvements at grade crossing locations can be suggested by establishing the contributing factors for accidents at each individual grade crossing and spearheading the safety improvement recommendation based on those contributing factors. Manually identifying the most significant contributing factors from the accident database at each grade crossing is not feasible due to various reasons. This thesis presents a new, easy method to extract the most frequent nested accident trends and helps an analyst visualize these trends using a tree based structure. This procedure is called the M+C method. The work presented in this thesis builds on top of the previous research conducted in this area by considering a more comprehensive set of accident attributes and by introducing a data{driven method of determining the order of the contributing factors to accidents at each location. The algorithm presented in this paper is also implemented in a computer program using the C++ language to automate the procedure thus reducing human e ort and error. Various examples illustrating the utility of the procedure to extract accident trends and contributing factors at a grade crossing. The use of this procedure to simultaneously analyze multiple crossings including all crossings along a corridor, all crossings within a county, all crossings with single accident are demonstrated in this thesis. This procedure is also used to identify new attributes to be used in accident prediction models. This computerized procedure, combined with a user{friendly interface could be a very useful tool that practitioners can use for quick and easy analysis of accidents at grade crossings.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Civil Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2017

Author dc:creator

• Mathew, Jacob

Contributors dc:contributor

• Benekohal, Rahim F.

• grade crossings
• accident visualization

Statement dc:rights

• Copyright 2016 Jacob Mathew

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/95397

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/95397