Doctoral Dissertations

Orcid ID

Date of Award


Degree Type


Degree Name

Doctor of Philosophy


Civil Engineering

Major Professor

Asad J Khattak

Committee Members

Asad J. Khattak, Candace Brakewood, Anahita Khojandi, Russell Zaretzki


Connected and automated vehicle technologies have the potential to significantly improve transportation system performance. In particular, advanced driver-assistance systems (ADAS), such as adaptive cruise control (ACC), cooperative adaptive cruise control (CACC), pedestrian crash prevention (PCP) system, and advanced automated collision notification (ACN) system may lead to substantial improvements in performance by decreasing driver inputs and taking over control of the vehicle. The main questions that might arise are 1) how we can quantify the potential environmental and safety impacts of these technologies and 2) what the potential for these technologies is to address some of the important transportation-related problems. Due to the limitation in the empirical CAV data, knowledge about their impacts is very limited. This dissertation attempts to fill a portion of the existing gap. The main focus of this study is on CAVs with a low level of automation. The key hypothesis is that the technologies utilized in the CAVs with a low level of automation (i.e., SAE automation levels 1 and 2) available in the market can benefit the environment, traffic safety, and the safety of vulnerable road users. The main contributions of this dissertation are 1) addressing research question one by analyzing new test-bed microscopic level data generated by CAVs and 2) addressing the second research question by identifying some important transportation-related problems that potentially can be addressed by CAV technologies. Overall, this dissertation estimates CAV technologies’ impacts based on the Haddon Matrix concept in terms of pre-crash, during-crash, and post-crash perspective. Methodologically speaking, advanced machine learning, data science, and frequentist techniques including truncated regression model, two-stages residual inclusion treatment to address endogeneity, geographically and temporally weighted regression model to address heterogeneity, explainable machine learning methods, XGBoost, and SHAP technique to interpret the result of XGBoost model are utilized to conduct the analysis. The results presented in this dissertation are derived from analyzing CAV testbed data and real-world crash data. Finally, the implications of the findings and future research areas are discussed in each chapter to a comprehensive understanding of the potential impacts of CAVs.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."