This study investigates individuals’ preference towards autonomous vehicle (AV) as a privately-owned vehicle and as a vehicle in shared services. One of the key features of this study is to examine the effects of historical changes and experiences over the life course towards the adoption of AVs. A joint bivariate ordered probit model is developed using data from a retrospective survey conducted in the Okanagan region of British Columbia, Canada. The model captures unobserved error correlation between the response variables: preference towards AV ownership and using AV as a vehicle in shared services. The model confirms the existence of a significant error correlation between preferences for AV ownership and sharing. The model results suggest that AV adoption is significantly affected by individuals’ historical experiences over the life course, such as exposure to technology, household characteristics, vehicle ownership, and accessibility and built-environment attributes of the residential location. For example, individuals with historical exposure to technologies including average smartphone ownership/person in the household over the life course and availability of advanced vehicle technology such as lane assist, parking assist, and autonomous emergency stop in the current and previously owned vehicles are more likely to adopt AVs. Individuals historically living in suburban areas have a higher likelihood of owning AVs whereas historically urban dwellers living in higher mixed land use neighbourhoods over the lifetime are more likely to prefer AV sharing. Ownership of technology devices such as smartphones and google home may lead to owning as well as sharing AVs. In addition, the model retains the effects of attitudinal factors, and socio-demographics. For instance, pro-environment individuals are likely to own and share AVs. Pro-shared-mobility individuals are more likely to share AVs whereas they are less interested in owning.

本研究调查了个人对自动驾驶汽车 (AV) 作为私有车辆和共享服务车辆的偏好。本研究的主要特点之一是检查历史变化和经历对采用自动驾驶汽车的生命历程的影响。使用在加拿大不列颠哥伦比亚省 Okanagan 地区进行的回顾性调查的数据开发了联合双变量有序概率模型。该模型捕获了响应变量之间未观察到的错误相关性：对 AV 所有权的偏好和将 AV 用作共享服务中的工具。该模型证实了对 AV 所有权和共享的偏好之间存在显着的错误相关性。模型结果表明，个人在生命历程中的历史经历对 AV 的采用有显着影响，例如接触技术、家庭特征、车辆所有权以及居住地点的可达性和建筑环境属性。例如，历史上接触过技术的个人，包括一生中智能手机的平均拥有量/家庭中的人，以及现有和以前拥有的车辆中车道辅助、停车辅助和自动紧急停止等先进车辆技术的可用性采用 AV。历史上居住在郊区的个人拥有 AV 的可能性更高，而历史上居住在更高混合土地使用社区的城市居民更喜欢共享 AV。拥有智能手机和 google home 等技术设备可能会导致拥有和共享 AV。此外，该模型保留了态度因素和社会人口统计学的影响。例如，环保人士可能拥有和共享 AV。支持共享出行的人更有可能共享自动驾驶汽车，而他们对拥有的兴趣不大。