Exploring Differences in Public Interests toward Automated Vehicles’ Ownership in Burdened and Non-Burdened Communities

Recent advancements in automated vehicle (AV) technologies have ignited considerable public interest, emphasizing the need to examine the potential disparities in the acceptance of these technologies across diverse communities. This study investigates the differences in public interest toward the adoption of automated vehicles between burdened (BCs) and non-burdened communities (non-BCs) using the 2017-2019 Puget Sound Household Travel Survey data. The study uniquely captures the burden indicators prevalent among census tracts in the Puget Sound region by integrating the newly released Transportation Communities data with the comprehensive Puget Sound household travel survey at the census tract level. Public interest toward the ownership of AVs is selected as the dependent variable with five ordered outcomes, namely “Not interested at all,” “Somewhat uninterested,” “Neutral,” “Somewhat interested,” and “Very interested”. The key explanatory variables include household sociodemographic and socioeconomic attributes, level of education, current mode of commute, and indicators of transportation insecurity. The study applies a rigorous and robust statistical model, i.e., a segmented ordered logit model, to model the impact of key enforcers and barriers to AV ownership in burdened and non-burdened communities. In addition to the inference-based statistical model, the study estimates a robust Artificial Intelligence technique, namely Random Forest, for enhanced prediction of public sentiments toward AV adoption, which is expected to assist in transportation planning. By addressing the potential disparities in the public acceptance of automated vehicles, this research can guide the development of targeted outreach programs and policy interventions tailored to bridge the gap between BCs and non-BCs.