Analyzing the Intention of AI on-demand Service Using Mode Choice Models

Exploitation of RP and SP survey data in Kozoji new town considering latent variables and heterogeneity

Abstract

AI on-demand services are efficient for low-density demand, but they often suffer from inadequate service levels and inappropriate fare systems. In this study, we develop a mode choice model based on a questionnaire survey conducted in Kozoji new town to analyze users’ intentions, which can be valuable for implementing the new service. The model integrates a stated preference (SP) model with information such as attitudes toward the new service, personal mobility restrictions, and users’ usual modes of transportation. This approach contributes to identification of the attributes that make people more willing to use the service, their attitudes toward it, and any mobility restrictions they may have. Furthermore, it enables a detailed analysis of demand sensitivity.