Date of Graduation
Master of Science in Civil Engineering (MSCE)
Second Committee Member
Artificial Intelligence, Electric Vehicles, Low-carbon Transportation, Machine Learning, Travel Behavior
The transportation sector stands as a significant contributor to greenhouse gas emissions in the United States, with its environmental impact steadily escalating over the past few decades. This has prompted government agencies to facilitate the adoption and usage of low-carbon transportation (LCT) options as alternatives to fossil-fuel-powered transportation. LCTs include modes of transportation that minimize the overall carbon footprint of the transportation sector by relying on energy sources that are environmentally sustainable. These sustainable transportation options have also garnered significant interest in the transportation research community. For government agencies and researchers alike, a comprehensive understanding of the adoption and usage of LCT is necessary to fully realize their carbon mitigation potential. The transportation sector is made up of several sub-sectors (e.g., light duty vehicle sector, heavy-duty vehicle sector, off-road equipment sector), which differ with regard to their compatibility and acceptance of LCT. Likewise, the research gaps for these sub-sectors also vary. This heterogenous nature of the transportation sector necessitates a separate consideration of the factors associated with LCT adoption and usage in the different sub-sectors. Consequently, this thesis explores these factors separately for the the light-duty vehicle (LDV) sector, the heavy-duty vehicle (HDV) sector and off-road equipment (ORE) sector. In the LDV sector, electric vehicles (EVs) have been some of the most widely accepted form of LCT. However, the literature has primarily explored their usage with respect to vehicle miles travelled. This thesis features a unique exploration of EV usage through the lens of vehicle choice in the LDV sector. To this end, the thesis employs a two-step machine learning framework (clustering and decision trees) on the National Household Travel Survey 2017 datasets. For HDVs and OREs, on the other hand, the influence of behavioral factors (awareness and impression) on LCT adoption has not been widely studied. To address this research gap, this thesis conducts a series of semi-structured interviews and analyzes them using a qualitative content analysis. The results of the analysis were refined by generative artificial intelligence (AI). Moreover, the analysis also explores the varying importance of factors (behavioral and other) of LCT adoption within different types (e.g., long-haul vs short-haul, large fleet vs small fleet) of organizations. The findings from the exploration of the LDV sector can be valuable to policymakers attempting to boost EV usage within the sector. The exploration of usage patterns can also provide auto manufacturers a unique perspective on the shortcomings of EVs that need to be addressed. For the HDV and ORE sectors, the findings can guide government agencies to develop tailored campaigns that leverage the influence of behavioral factors on LCT adoption. Moreover, the insights uncovered from the utilization of generative AI can serve as a baseline for researchers seeking to conduct AI-assisted qualitative studies in transportation.
Chowdhury, V. (2023). Decoding Usage and Adoption Behavior of the Low-Carbon Transportation Market: An AI-driven Exploration. Graduate Theses and Dissertations Retrieved from https://scholarworks.uark.edu/etd/5159