ARE WE READY TO EMBRACE CONNECTED AND SELF-DRIVING VEHICLES? A CASE STUDY OF TEXANS

Connected and autonomous vehicles (CAVs) represent a significant technological advancement with the potential to drastically reduce crashes caused by driver error. However, the increased convenience may induce higher travel demand, potentially negating safety gains and necessitating smart congestion-pricing strategies. Shared autonomous vehicles (SAVs) are also gaining attention as a new transportation mode, potentially impacting land use patterns and raising policy questions regarding land prices and SAV costs. Public opinion is crucial in determining the success and market penetration of CAVs. While previous studies have explored public opinions, most have focused on pairwise correlation analysis. This study builds upon existing research by using a larger, Texas-wide survey (1,088 respondents) to estimate econometric models that uncover multivariate relationships between public opinion on CAVs and various demographic, built-environment, and travel-related factors. This aims to understand the main determinants of public acceptance and to inform policy decisions regarding infrastructure, legal and safety issues, and other aspects of the connected and autonomous transportation system.

The literature review summarizes existing studies on public perception of CAVs. Several academic and professional studies, private firm surveys, and online polls consistently reveal public caution regarding driverless vehicles. Safety, affordability, and information security are frequently cited as major concerns. Studies varied in their methodologies and sample sizes, but several common themes emerged, including safety as a primary concern, varying preferences for automation levels across demographics (younger people and men more open to automation), and significant concerns about cost. The authors highlight that this study extends previous work by utilizing a larger sample size, incorporating a wider range of explanatory variables (e.g., crash history and opinions on safety regulations), and employing advanced econometric techniques beyond pairwise correlation analysis.

A Texas-wide survey with 93 questions across seven sections was administered through Survey Sampling International's continuous panel in June-July 2015. After cleaning and sanity checks, 1,088 responses were included in the analysis. Person- and household-level weights were calculated to correct for sample biases. Geographic locations of respondents were geocoded using Google Maps API and census data to understand the relationship between built-environment factors and opinions on CAV technologies. Interval regression (IR) models were used to estimate willingness to pay (WTP) for connectivity and various automation levels, accounting for interval and censored response data. Ordered probit (OP) models were employed to analyze interest in connectivity, adoption timing of AVs, SAV adoption rates under different pricing scenarios, home location shift decisions, and opinions on congestion pricing policies. Model selection involved starting with a larger set of predictors and iteratively removing those with the lowest statistical significance until all remaining covariates had p-values below 0.32. Practical significance was assessed by evaluating the change in response values due to a one-standard-deviation increase in each covariate. Practically significant predictors were those that resulted in at least a 0.2 standard deviation change in WTP in the IR models and at least a 40% shift in choice probabilities in the OP models.

The key findings are presented across several tables summarizing WTP for different automation levels and connectivity, opinions on SAV adoption, home location shifts, opinions on AVs and CVs, and support for tolling policies. Texans showed a willingness to pay an average of $127 for connectivity, $2,910 for Level 2 automation, $4,607 for Level 3, and $7,589 for Level 4. A substantial percentage (29.3% - 54.4%) were unwilling to pay more than $1,500 for these technologies. Regarding SAV adoption, only 7.3% of Texans planned to rely entirely on an SAV fleet even at $1 per mile. Most (81.5%) did not plan to change their home locations due to CAVs. Affordability and equipment failure were the greatest concerns, while learning to use AVs and privacy breaches were the least concerning. Texans were most likely to support adaptive traffic signal timing and least likely to support real-time parking price adjustments. The study also examined the factors influencing the adoption rates of SAVs at different pricing points and the decisions about home location shifts in the future. The econometric modeling results revealed several practically significant relationships between explanatory variables (such as driving experience, age, income, and location) and key responses (such as WTP and adoption rates).

The findings address the research question by quantifying Texans' attitudes towards CAV technologies and identifying the factors influencing their adoption. The significant relationships between demographics, built-environment attributes, and WTP/adoption highlight the importance of considering these factors in policymaking. For instance, the lower WTP for automation among older people suggests a need for targeted awareness campaigns to address concerns and promote understanding. The study's econometric approach offers valuable insights beyond simple correlations, providing a nuanced understanding of the complex interplay of factors affecting CAV adoption. The results are relevant to transportation planners and policymakers, providing data to forecast long-term adoption, identify high- and low-adoption regions, and develop tailored strategies.

This study provides valuable insights into Texans' opinions on CAVs and SAVs, quantifying WTP for different automation levels and identifying key factors influencing adoption. The findings highlight the importance of considering demographic and geographic factors in policy development and suggest potential areas for targeted interventions. Future research could focus on longitudinal studies to track changing public opinion and on more detailed analyses of the potential impacts of CAVs on land use patterns and traffic management strategies. Further studies in different geographical locations would also broaden the generalizability of the findings.

The study's cross-sectional nature limits the ability to definitively establish causality between observed relationships. The sample, while large, may not perfectly represent the entire Texan population, despite weighting efforts. The survey relied on self-reported data, which might be subject to biases. Changes in technology and public perception over time could also affect the generalizability of the findings. While the study provided a comprehensive analysis with numerous variables, other factors not included here may also play a significant role in shaping public attitudes and adoption rates of CAVs.