Citation

Bham GH, Manepalli URR, Samaranayke VA. J. Transp. Saf. Secur. 2019; 11(3): 225-242.

Copyright

(Copyright © 2019, Southeastern Transportation Center, and Beijing Jiaotong University, Publisher Informa - Taylor and Francis Group)

DOI

10.1080/19439962.2017.1384417

PMID

unavailable

Abstract

Identification of hotspots or high-crash locations on highways is important to ensure public safety and minimize loss of life, risk of injury and/or trauma; reduce crash cost to society; and any interruption to traffic flow. Many performance measures are available for use in identification of hotspots. Current performance measures, however, suffer from lack of accounting for traffic volume, crash injury severity, justification of weights used for ranking of sites, and issues with statistical distribution of crash data. This article proposes a composite rank measure based on principal component analysis to overcome limitations of existing measures for network screening. The results of a proposed measure called Composite Principal Rank Measure (CPRM) is demonstrated with interstate, U.S. and state highway data and compared against the commonly used sum-of-rank (SOR) measure. CPRM and SOR are evaluated based on several empirical and simulated data tests. CPRM was found to be robust and outperformed the SOR measure and is recommended for identification of hotspots. To ensure extensive evaluation of the measures, the entire highway routes are examined in this article.

Language: en

Keywords

accident prone locations; blackspots; crash data; crash frequency; empirical Bayes; high accident locations; hotspot identification (HSID) measures; hotspots; injury severity; principal component analysis