01624344

Component

Star Rating protocol, applied by International Road Assessment Program (iRAP), awards 1 to 5 stars depending on the level of safety which is ‘built-in’ to the road. Although this procedure is known and used world -wide, there have been only a few validation studies, showing the relationship between Star Ratings and crashes. In addition, the past studies relied on crash rates or crash costs per kilometer, derived from Police-reported data only. This approach is not the best, since it incorrectly assumes a linear relationship between crash frequency and the measure of exposure, and does not control for the confounding effect of regression to the mean. The objective of the paper is to fill the gap and conduct validation study using state-of-the-art empirical Bayes approach, combining observed crash frequency with expected crash frequency based on a crash prediction model. For the purpose of this exploratory study, Star Ratings from the Hungarian rural road network were used. The proposed approach proved to be feasible, with results confirming the relationship between increasing Star Ratings and decreasing crash frequencies. However, some limitations are also reported, which should be addressed by further and improved validation studies.

This paper was sponsored by TRB committee ANB10 Standing Committee on Transportation Safety Management.

01618707

17-01991

English

11p

2017

Transportation Research Board 96th Annual Meeting

Digital/other

Figures; Maps; References (37) ; Tables

Crash rates; Highway safety; Mathematical prediction; Ratings

Hungary

Highways; Safety and Human Factors

Transportation Research Board Annual Meeting 2017 Paper #17-01991

PRP, TRIS, TRB, ATRI

Dec 8 2016 10:42AM