Incorporating crash severity to improve highway safety project prioritization

R. Tanzen, R. Souleyrette, T. Wang, W. Staats

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Released in 2010, the Highway Safety Manual (HSM) provides guidelines for evaluating highway safety improvements and prioritizing potential projects. Adopting the HSM methodology, several states in the US use Excess Expected Crashes (EEC), a parameter dependent on Safety Performance Functions to rank safety projects. However, this method is limited by several methodological disadvantages (e.g., the severity of the observed crashes and the magnitude of the projected crashes by the Empirical Bayes (EB) method are not considered). This paper aims to improve highway safety project ranking and describes a new safety scoring method developed for the Kentucky Transportation Cabinet (KYTC). It is now being used in KYTC’s Strategic Highway Investment Formula for Tomorrow (SHIFT) project prioritization process. The method considers crash severity and incorporates EB estimates and the EEC metric in a multifactor score. Additionally, it introduces a “goal-driven” EEC, which represents the potential for reaching targets specified in the state’s Strategic Highway Safety Plan. To demonstrate the methodology, the method is tested on KYTC’s list of potential projects for the 2020 SHIFT cycle.

Original languageEnglish
Pages (from-to)155-168
Number of pages14
JournalAdvances in Transportation Studies
Issue numberSpecial issue
StatePublished - 2022

Bibliographical note

Publisher Copyright:
© 2022, Aracne Editrice. All rights reserved.


  • Empirical Bayes Estimate
  • excess expected crashes
  • Safety Performance Functions
  • safety project prioritization

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Automotive Engineering
  • Safety, Risk, Reliability and Quality
  • Transportation


Dive into the research topics of 'Incorporating crash severity to improve highway safety project prioritization'. Together they form a unique fingerprint.

Cite this