Modeling motorists' speed and its impact on highway work zone safety for programming police enforcement

Erdong Chen, Purdue University

Abstract

Highway work zone safety has been a major concern for transportation agencies, especially at a time when a great number of maintenance projects are needed for the aging U.S. highway system. State transportation agencies have made various efforts recently to improve safety in their work zones, including implementation of police enforcement to deter speeding, aggressive driving, and other dangerous driving behaviors, with the ultimate goal of reducing crashes and improving traffic operations within their work zones. Prior to the work of this dissertation, there was a considerable gap between the existing body of literature and a thorough understanding of how police enforcement affects motorists' speed behaviors and how these behaviors affects highway work zone safety. In order to achieve the best safety benefits from such enforcement activities in work zones, research was needed to fill this gap. Thus, the research efforts of this dissertation had the following objectives: 1) to achieve a better understanding of highway work zone safety issues; 2) to understand how various police enforcement strategies affect driving behavior in work zones; and 3) to develop the Police Enforcement Programming Tool to assist state DOTs in optimizing their work zone enforcement activities. To achieve these objectives, the authors identified several tasks: (1) Estimate crash models, to gain better understanding about work zone crashes; (2) Conduct work zone police enforcement experiment, analyze driving behavior, and evaluate the effectiveness of various enforcement strategies; (3) Linking change in driving behavior and traffic safety, and develop an integrated tool for programming work zone police enforcement. All these objectives were satisfactorily achieved and the impact of this study is expected to extend to both academia and practice. During the work zone crash modeling, an unprecedented dataset was assembled, enabling very comprehensive analyses of work zone crashes and thereby adding new insights to the body of literature. The Police Enforcement Experiment of this dissertation included a wide variety of factors in a systematically designed setting and is the best-designed such study to date. Also, statistically significant conclusions were obtained from the subsequent advanced modeling analyses, which significantly improved the understanding of motorists' behaviors in work zones in response to police enforcement. All the research findings were incorporated in the Crash Modification Factors and the implementation-ready Police Enforcement Programming Tool, which has been delivered to the Indiana Department of Transportation and is expected to make a substantial impact for practice.

Degree

Ph.D.

Advisors

Tarko, Purdue University.

Subject Area

Civil engineering|Transportation

COinS