Impacts of Autonomous Vehicles on Mode Choice Behavior in the Context of Short- and Medium- Distance Intercity Travel

This paper studies the impacts of autonomous vehicles on mode choice behavior in the context of shortand medium-distance intercity travel. Based on the theory of planned behavior, a structure equation model isdeveloped, through which latent psychological variables of individuals towards autonomous vehicles are developed, including perceived behavioral control, subjective norms, attitudes, and behavioral intentions. These latent psychological variables are then integrated into a random parameter Logit model to develop a hybrid choice model. The City of Wuhan is used as a case to carry out an empirical study, and the study results show that: in the utility function, the coefficients of three variables, including in-vehicle time, access and exit and waiting time, and travel cost, are not fixed but follow a normal distribution with a mean of -0.014, -0.008, and -0.010 and with the standard deviations of 0.014, 0.021, and 0.017, respectively. When the perceived behavior control and attitude of individuals towards autonomous vehicles increased by 1 unit, the probability of using autonomous vehicles to travel increased by 64.3% and 77.9%, respectively. For every 1% decrease in the travel cost and in-vehicle time of autonomous vehicles, the probability of choosing autonomous vehicles and intercity shuttles increases by 0.403% and 0.467%, respectively. This paper studies the impacts of autonomous vehicles on mode choice behavior in the context of shortand medium-distance intercity travel. Based on the theory of planned behavior, a structure equation model isdeveloped, through which latent psychological variables of individuals towards autonomous vehicles are developed, including perceived behavioral control, subjective norms, attitudes, and behavioral intentions. These latent psychological variables are then integrated into a random parameter Logit model to develop a hybrid choice model. The City of Wuhan is used as a case to carry out an empirical study, and the study results show that: in the utility function, the coefficients of three variables, including in-vehicle time, access and exit and waiting time, and travel cost, are not fixed but follow a normal distribution with a mean of -0.014, -0.008, and -0.010 and with the standard deviations of 0.014, 0.021, and 0.017, respectively. When the perceived behavior control and attitude of individuals towards autonomous vehicles increased by 1 unit, the probability of using autonomous vehicles to travel increased by 64.3% and 77.9%, respectively. For every 1% decrease in the travel cost and in-vehicle time of autonomous vehicles, the probability of choosing autonomous vehicles and intercity shuttles increases by 0.403% and 0.467%, respectively. Study results show that travelers have heterogeneous preferences toward the attributes of the transport service offered by autonomous vehicles, such as in-vehicle time, access/egress and waiting time, and travel costs. It is also found that perceived behavioral control and behavioral attitudes have significantly positive impacts on traveler's choice on autonomous vehicles. Therefore, reducing travel costs and travel time of autonomous vehicles can increase the attractiveness of autonomous vehicles.