01592867

Component

In this research the authors investigate monitoring crash counts on roadways as a tool to improve roadway safety. The authors propose a continuous spatiotemporal monitoring methodology to simultaneously monitor multiple locations and determine if significant changes occur in crash rates from baseline rates. The authors evaluate multivariate Normal, Poisson, and Exponential cumulative sum (CUSUM) schemes for monitoring of traffic crash data to determine if statistically significant changes occur. Simulation results show that cumulative sum procedures effectively detect crash frequency shifts at multiple locations. Recommendations are made for tuning the control chart parameters, the choice between monitoring crash counts and time between crashes, and considering multiple intersections to increase subgroup sizes. The authors illustrate the application of the proposed approach using a case study involving intersection-related crash data from the Florida Department of Transportation (FDOT) for several intersections in Tallahassee, Florida. Using the proposed methodology, transportation officials can grasp the dynamic changes in the crash occurrences more quickly, which can help them formulate more effective safety improvement decisions.

This paper was sponsored by TRB committee ABJ80 Standing Committee on Statistical Methods.

01584066

16-5699

English

19p

2016

Transportation Research Board 95th Annual Meeting

Web

Figures; Maps; References (22) ; Tables

Crash rates; Highway safety; Intersections; Mathematical models; Monitoring; Spatial analysis

Tallahassee (Florida)

Highways; Safety and Human Factors

Transportation Research Board Annual Meeting 2016 Paper #16-5699

TRIS, TRB, ATRI

Jan 12 2016 6:30PM