|This dissertation aims to identify the effectiveness of engineering traffic control devices for wrong-way driving (WWD) from a driver behavior perspective, which contains two tasks: i) WWD fatal crash analysis, and ii) driving simulator study design and analysis.
For the WWD fatal crash analysis, the study aims to conduct a comprehensive analysis of WWD fatal crashes on divided highways using 17 years (2004–2020) of WWD fatal crash data which were extracted from the National Highway Traffic Safety Administration (NHTSA) Fatality Analysis Reporting System (FARS) database. The FARS database has the distinct advantage of providing WWD fatal crashes nationwide with more than 140 variables that record crash, vehicle, driver, passenger, and pedestrian information, whereas previous studies were mostly based on the limited WWD crash data collected from several states. The analysis included: i) updating the trends for WWD fatal crashes up to 2020; ii) analyzing the distribution of WWD fatal crashes among states, genders, and ages; and iii) examining the WWD fatal crash risk factors using binary logistic regression and multiple correspondence analysis (MCA). The results revealed that i) an average of 302 WWD fatal crashes happened each year, and the number of WWD fatal crashes has not been declining over the years; ii) over 60% of WWD fatal crashes are alcohol involved; and iii) young driver, male driver, alcohol-impaired driver, and nighttime conditions are predominate in the WW fatal crashes. The research results help readers understand the national trend of WWD fatal crashes and the risk actors that cause WWD fatal crashes, which may support the decision-making for FHWA, state DOTs, and local governments in the future. Most importantly, these results, especially the contributing factors, will be considered as the base of the scenario development and lab testing design for the driving simulator study.
For the driving simulator study, the main purpose is to evaluate the effectiveness of the proposed traffic control devices (TCD(s)) to prevent WWD for highly intoxicated drivers by analyzing drivers’ behavioral data collected from the driving simulator and eye-tracking device. The driving simulator study provides a unique opportunity to study drivers’ behaviors when they face different kinds of wrong-way-related TCD(s), which can not be achieved through traditional field data collection and crash data analysis. 30 male participants with an average age of 25 were recruited for the driving simulator study. The non-alcohol session and real-alcohol session were given to the participants in counterbalanced order. For each session, participants were required to complete three scenarios that contains the proposed TCD(s) in counterbalanced order. The driving simulator data and eye-tracking video data were collected for analysis purposes, which encompasses more than 1,500 minutes of video and more than 5,506,580 data points at 0.16-s intervals, respectively. First, the general information such as drivers’ information and actual BrAC level was summarized for the study. Then the comparison of the forward driving scene between normal and alcohol-impaired conditions was analyzed based on visualization and Chi-square results. Finally, the effectiveness of TCD(s) was evaluated by three criteria: i) the number of WWD events; ii) fixation duration; and iii) brake response distance. The research results provided detailed information regarding how drivers react facing different TCD(s) based on those three criteria, which can be utilized by transportation agencies for TCD selection and to better design off-ramps and prevent WWD events.