MODELING THE IMPACT OF CONNECTED AND AUTOMATED VEHICLES ON DRIVING BEHAVIORS AND SAFETY: A DRIVING SIMULATOR STUDY

Abstract

Connected vehicles (CVs), equipped with advanced sensors, can communicate safety messages to drivers. Automated vehicles (AVs), designed with the ability to automate safety-critical control functions, will redefine the traditional role of drivers. This dissertation aims to investigate the impact of connected and automated vehicles (CAVs) on driving behaviors and safety outcomes using data from driving simulator experiments. More specifically, the research objectives include: 1. Modeling the impacts of CVs on driving aggressiveness and situational awareness in highway crash scenarios. 2. Modeling the impacts of CV technologies on driving behaviors and safety outcomes in highway crash scenarios under diverse weather conditions, including clear and foggy weather. 3. Understanding the factors influencing driver safety performance within CAV technology during safety-critical events that necessitate driver takeover.

Structural equation modeling (SEM) was utilized to examine the interrelationships among the use of CV warnings, psychological factors including aggressiveness and situational awareness, driving behavior, weather conditions, safety outcome, and other variables. Random effects logit models were developed to understand the contributing factor to CAV drivers’ takeover performance in safety-critical events. Results showed that the proposed CV warnings significantly reduced aggressiveness and increased situational awareness, contributing to improved safety especially on a horizontal curve. Foggy weather had an overall negative impact on safety on a horizontal curve, despite that it increased drivers’ situational awareness. Additionally, CV warnings could notably improve the drivers’ takeover performance in safety-critical events. The insights gained from this dissertation are crucial in shaping the development of advanced driving assistance systems and automated driving systems. that seamlessly integrate psychological factors. They underscore the importance of customization based on weather conditions and location-specific factors. Moreover, this study provides valuable insights into improving human-machine interactions of CAV systems in safety-critical events, ultimately contributing to safer roads.