This report explains the procedure involved in developing Safety Performance Functions (SPF) for the road segments and intersections in the state of Illinois, under the jurisdiction of the Illinois Department of Transportation. SPFs predict the relationship among traffic, geometric conditions of the road and crash density, crash severity, and crash type. The SPFs are developed in such a way that they can be easily incorporated into the SafetyAnalyst tool that is being developed by FHWA to analyze and improve the safety of road elements. The SPFs are used to calculate a given site’s Potential for Safety Improvement (PSI) and thus help in identifying the locations that have the highest potential for improvement. A literature review was conducted as a part of the study to identify the methodology that would be needed. The literature review included studies on identifying the statistical techniques best suited for the requirement and identifying road element variables that have to be considered while developing SPFs. This report also includes the background and rationale behind the use of techniques such as Empirical Bayesian method, Sliding Window technique, and other such procedures that have been used in the analysis. As part of the study, road segments and intersections were classified into peer groups such that members of a peer group would have homogenous characteristics. Network Screening was conducted for all state-maintained (marked and unmarked) routes to identify high-crash locations, which directly supports the development of the 2008 Illinois Five Percent Report to FHWA. This project also develops a VBA (Visual Basic for Applications) software tool that can be directly used by Illinois Department of Transportation (IDOT) officials to update SPFs and PSI screening in the future. This tool incorporates the statistical and computation models in an easy-to-use Excel spreadsheet environment. The software also automates the decision support process for identifying high crash sites in the Illinois roadway network. The system requirements and the procedure involved in using the software are explained in this report.
[1]
Bhagwant Persaud,et al.
STATISTICAL METHODS IN HIGHWAY SAFETY ANALYSIS
,
2001
.
[2]
John N. Ivan,et al.
Hierarchical Bayesian Estimation of Safety Performance Functions for Two-Lane Highways Using Markov Chain Monte Carlo Modeling
,
2005
.
[3]
Ezra Hauer,et al.
Screening the Road Network for Sites with Promise
,
2002
.
[4]
Bruce N. Janson,et al.
Diagnostic Methodology for the Detection of Safety Problems at Intersections
,
2002
.
[5]
Bradley P. Carlin,et al.
BAYES AND EMPIRICAL BAYES METHODS FOR DATA ANALYSIS
,
1996,
Stat. Comput..
[6]
Ezra Hauer,et al.
Observational Before-After Studies in Road Safety
,
1997
.