Automated vehicles (AVs) have great potential to revolutionize the transportation sector and landscapes of future cities. The impacts of AVs on urban space, however, are far from clear. Mobility-on-Demand (MOD) services, on the other hand, are readily available in many places. This study seeks to explore (1) how Automated Mobility-on-Demand (AMOD) might affect urban residents' levels of accessibility and their residential relocation decisions; and (2) how these impacts might vary across space and socioeconomic groups. We use an agent-based microsimulation platform to assess two future AMOD scenarios in Singapore relative to a baseline. Results suggest that the addition of AMOD could enhance the overall accessibility of the population, but not if private transport modes, including private cars, taxis, and human-driven on-demand services, are prohibited. On the other hand, if private modes are eliminated, AMOD could alleviate inequality in accessibility as it appears to benefit the disadvantaged socioeconomic groups to a larger extent. We also find that AMOD deployment would not induce outward migration, nor would it increase home-work location imbalance. This study demonstrates how large-scale microsimulation can be leveraged to assess AMOD scenarios. The findings have some implications for preparing for the inevitable and potentially disruptive emergence of AVs.

自动驾驶汽车 (AV) 具有极大的潜力，可以彻底改变未来城市的交通运输部门和景观。然而，自动驾驶汽车对城市空间的影响尚不清楚。另一方面，按需移动 (MOD) 服务在许多地方都很容易获得。本研究旨在探讨 (1) 自动按需出行 (AMOD) 如何影响城市居民的可达性水平及其住宅搬迁决策；(2) 这些影响如何因空间和社会经济群体而异。我们使用基于代理的微观模拟平台来评估新加坡相对于基线的两个未来 AMOD 场景。结果表明，增加 AMOD 可以提高人口的整体可达性，但如果私人交通方式，包括私家车、出租车和人力驱动的按需服务，则不会 被禁止。另一方面，如果消除私人模式，AMOD 可以减轻可及性方面的不平等，因为它似乎在更大程度上使弱势社会经济群体受益。我们还发现 AMOD 部署不会导致向外迁移，也不会增加家庭与工作位置的不平衡。这项研究展示了如何利用大规模微观模拟来评估 AMOD 场景。这些发现对于为不可避免且可能具有破坏性的自动驾驶汽车的出现做好准备具有一定的意义。这项研究展示了如何利用大规模微观模拟来评估 AMOD 场景。这些发现对于为不可避免且可能具有破坏性的自动驾驶汽车的出现做好准备具有一定的意义。这项研究展示了如何利用大规模微观模拟来评估 AMOD 场景。这些发现对于为不可避免且可能具有破坏性的自动驾驶汽车的出现做好准备具有一定的意义。