Exploring citizens' preference for on-demand valet parking service under uncertain parking process: a Bayesian estimation approach

In major global cities, parking problems have intensified owing to increased private vehicle usage and decreased availability of parking spaces. To address this issue, researchers have explored effective parking management policies based on the parking behaviour of citizens. This study investigates citizens' preferences for an emerging solution called on-demand valet parking (OVP) through a stated preference (SP) survey conducted in the Seoul Metropolitan Area, South Korea. The survey included time-related uncertainties that occur during self-parking and OVP, such as searching time for a parking space, waiting time for the valet to arrive, and time spent on handing over the vehicle to the valet. Using mixed logit models with Bayesian estimation, this study analyzes individual-level parameters to understand heterogeneity in preferences. The results indicate that citizens prioritize parking costs and time, and that uncertainty in the parking process significantly influences their preferences. Moreover, while the average elasticity suggests that certain attributes are relatively elastic, individual-level analysis revealed significant heterogeneity in responsiveness. Policymaking and service designs based on population-level parameters may overlook this heterogeneity. Cluster analysis further elucidates this heterogeneity by identifying four distinct types of decision-makers based on their sensitivity to various attributes, such as walking time and uncertainty. These insights can guide more personalized and effective service enhancements, ensuring that the specific needs of different driver groups are met. These findings underscore the importance of considering individual preferences and uncertainties in parking behaviour for effective transportation policy and service provision. Further research can explore decision-rule heterogeneity to better understand citizen behaviour in the context of parking choices.