Spatial Disparities in Safety Benefits from Automated Vehicle Adoption in Alabama and Kentucky

Abstract: This study explores the spatial disparities in safety benefits from Automated Vehicle (AV) adoption across rural and urban areas in Alabama and Kentucky. While AV technologies have the potential to reduce crashes caused by human error, the uneven distribution of interest in AV adoption raises concerns about how these benefits will be realized across different regions. The research aims to quantify potential crash reductions associated with AVs, focusing on rear-end, sideswipe, and rollover crashes. By integrating AV ownership predictions and crash data, this study offers insights into how AVs may affect road safety in diverse geographic contexts.
The study utilizes crash records from Alabama and Kentucky for 2019, combined with a model predicting AV adoption interest at the Census Block Group (CBG) level. An ordered probit model is employed to estimate AV interest based on sociodemographic factors such as age, education, income, race, and vehicle ownership. The predicted AV interest is used to assess crash reduction potential, not only within each CBG but also in neighboring areas through an inverse distance weighting (IDW) procedure. This approach accounts for spillover effects, capturing how AV ownership in one area influences safety outcomes in surrounding regions.
The analysis reveals significant spatial disparities in crash reduction potential between rural and urban areas. Urban areas, with high crash frequencies but low AV adoption interest, exhibit the greatest opportunity for crash reduction. For instance, Alabama’s urban regions recorded 13,465 rear-end crashes, with a potential reduction of 3,987 if AVs were fully adopted. Similarly, Kentucky’s urban areas reported 12,865 rear-end crashes, with an estimated reduction of 3,618. In rural areas, though crash frequencies are lower, the potential for reducing severe crashes like rollovers is notable. For example, Alabama’s rural regions with high crash rates and AV interest showed a potential reduction of 287 rollover crashes. These findings emphasize the importance of matching AV deployment strategies with regional crash profiles to maximize safety benefits.
The results highlight the need for equitable deployment of AV technologies to ensure that both rural and urban communities benefit from improved road safety. While urban areas require interventions targeting rear-end and sideswipe crashes, rural regions would benefit from technologies designed to prevent rollovers. Policymakers should prioritize the introduction of AV features that align with the unique safety needs of each area, ensuring that the benefits of AV adoption are widely distributed.
The use of IDW to capture spillover effects highlights the importance of viewing AV benefits in a regional context, as travel across neighboring areas affects overall safety outcomes. Methods and data limitations are described in order to inform future work. Overall, this research provides valuable insights for policymakers and planners, supporting informed decision-making in AV deployment strategies to enhance traffic safety across diverse areas.