Departure time choice and bottleneck congestion with automated vehicles

Abstract

It is widely expected that automated vehicles (AVs) will revolutionise travel experience by better facilitating various on-board activities. While these activities could make travel more pleasant, as is often supposed, they could also affect daily schedules, the related travel choices, and finally, the aggregate travel patterns – possible influences that are still insufficiently studied. For example, a morning commuter deciding to perform some home or work activities during travel, instead of at home or work, could also reconsider his departure time to work. More such travellers together could reshape traffic congestion. This paper models exactly this scenario. It formulates new scheduling preferences, which account for home and/or work activities during morning commute, and uses these (1) to analyse the optimal departure times when there is no congestion, and (2) to obtain the equilibrium congestion patterns in a bottleneck setting. If there is no congestion, it is predicted that AV users would depart earlier (later), if the on-board environment supports their home (work) activities. If there is congestion, AV users that perform home (work) activities during travel skew the congestion to earlier (later) times, and AV users that perform both activities increase both early and late congestion. Engaging in any activity during travel worsens congestion, at least when assuming that AVs do not increase bottleneck capacity. If future AVs would be specialised to support only home, only work, or both home and work activities, and would do so to a similar extent, then ‘Work AVs’ would increase the congestion the least.