A Data Mining Approach for Identifying Key Factors Affecting the Severity of Highway-Railway Grade Crossing Collisions Involving Vulnerable Road Users

The major objective of this study is to identify the main factors associated with the injury severity of vulnerable road user (VRU) involving in collisions at highway railroad grade crossings (HRGC) using data mining techniques. To address the problem of pedestrian safety at highway railroad crossings, it is essential to understand the underlying relationship between pedestrian injury severity outcomes and various factors. Previous studies have investigated the injury severity problem of VRUs. However, most of these efforts had focused on collisions at road intersections with few on grade crossings. Furthermore, these studies applied mostly traditional statistical models which have the limitation in analyzing the interaction of multiple factors. This paper applies two data mining techniques called Association Rules and Classification-Regression Tree (CART) algorithms on the U.S. Federal Railroad Administration (FRA) HRGC collision database for the period of 2007 - 2013 to identify VRU injury severity factors at HRGCs. The results show that train speed, collision time, age and gender, weather condition, and illumination are the key factors influencing the injury severity. The findings of this research could be applied for identifying high-risk hotspots and developing cost-effective countermeasures targeting VRUs at HRGC.


  • English

Media Info

  • Media Type: Digital/other
  • Monograph Title: TRB 94th Annual Meeting Compendium of Papers

Subject/Index Terms

Filing Info

  • Accession Number: 01557459
  • Record Type: Publication
  • Report/Paper Numbers: 15-4543
  • Files: TRIS, TRB, ATRI
  • Created Date: Dec 30 2014 1:30PM