Graduation Year

2024

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Engineering

Major Professor

Pei-Sung Lin, Ph.D.

Co-Major Professor

Yu Zhang, Ph.D.

Committee Member

Zhenyu Wang, Ph.D.

Committee Member

Fred Mannering, Ph.D.

Committee Member

Seckin Ozkul, Ph.D.

Keywords

Rural and Urban Crashes, Mixed Logit Model, Out-of-Sample, Heterogeneity in Mean and Variance, COVID-19

Abstract

A work zone can be described as a roadway designated for construction, maintenance, or other activities that temporarily disrupt the normal traffic flow. Due to this disrupts in traffic conditions, work zones may bring sudden shifts in driving experience, depending on the sort of work zone present, creating unpredictable maneuvers, and putting both motorists and workers at risk. To understand these issues, this study uses work zone accident data obtained from Florida and uses several methodological techniques, including the mixed logit model with variance and mean heterogeneity, temporal instability test, log-likelihood test, out-of-sample analysis, and ANOVA test.

Firstly, this study looks into drivers’ pre-crash actions, which are leading causes of single-vehicle work zone crashes. The model looked into five major pre-crash improper actions in single-vehicle work zone crashes and the outcomes determined a variety of variables impacting drivers' pre-crash inappropriate actions, including driver and weather conditions, work-zone type, and temporal and lighting conditions. Highlighting the importance of engineering or educational remedies to increase driver alertness and lower improper actions, the investigation underlines that driving psychology notably influences safety-related driving habits in work-zone areas.

Second, the analysis examines the driver injury severities across rural and urban work zones of Florida, over the period of four years. The difference between rural and urban work zones was demonstrated using a series of log-likelihood tests. Models from different years of rural and urban datasets highlighted the significant impact of factors such as speed limits, work zone type, and traffic volume on work zone injury severities, with variations observed between rural and urban areas. Regardless of their spatial variations common safety problems and countermeasures, including driver education, improved signage, and enhanced law enforcement, were identified for both rural and urban work zone areas.Next, this research assesses the impact of the COVID-19 pandemic on single-vehicle work zone crashes by separating the dataset into periods before, during, and after the COVID-19 pandemic. Through random parameters, multinomial logit model factors such as lane departure, vehicle manufacturing year, and driving speed significantly influence crash injury severities. Shifts in injury predictions before, during, and after the pandemic were investigated through predictive analysis. The findings of this study provide valuable insights for decision-makers and practitioners in enhancing work zone safety during unprecedented challenges.

Lastly, this study explores the driving pattern and speed of drivers across different types of work zones (lane closure, lane shift, and Work on shoulder/median). With various test scenarios and utilizing speed sensors, the study found that the presence of Active Work Zone Awareness Devices (AWAD) and their combination with Law Enforcement or iCone devices showed a decline in average spot speeds in both upstream and activity area of the work zone thereby promoting safe driving and reducing risky driving behaviors. The dissertation concludes with a summary of the findings.

Download

Share

COinS