Urban air mobility: A comprehensive review and comparative analysis with autonomous and electric ground transportation for informing future research

https://doi.org/10.1016/j.trc.2021.103377Get rights and content 获取权利和内容
Under a Creative Commons license
open access 开放获取

Highlights 亮点

  • We compile a database of about 800 UAM, EV, and AV articles.
    我们编制了一个包含约 800 篇有关 UAM、EV 和 AV 的文章的数据库。

  • We use insights from the EV and AV literature to inform future UAM research directions.

  • UAM research has primarily focused on aircraft technologies and operations.
    UAM 研究主要集中在飞机技术和运营方面。

  • EV and AV research has a greater emphasis on technology adoption and integration with existing infrastructure.

  • To date, the majority of UAM research has been conducted by U.S. researchers.
    迄今为止,大部分 UAM 研究都是由美国研究人员进行的。

Abstract 摘要

Urban air mobility (UAM), if successful, will disrupt urban transportation. UAM is not the first disruptive technology in transportation, with recent examples including electric ground vehicles (EVs), autonomous ground vehicles (AVs), and sharing services. In this paper, we conduct a meta-analysis of about 800 articles in the UAM, EV, and AV areas that have been published from January 2015 to June 2020, and compare and contrast research thrusts in order to inform future UAM research. Alongside this effort, we conduct an in-depth review of articles related to demand modeling, operations, and integration with existing infrastructure. We use insights from the meta-analysis and comprehensive review to inform future UAM research directions. Some of the potential research directions we identify include: (1) developing more refined demand models that incorporate the timing of when individuals will adopt UAM; (2) developing high-fidelity simulation models for UAM operations that capture interactions among vertiport locations, vertiport topology, demand, pricing, dispatching, and airspace restrictions; (3) explicitly considering one-way demand and parking constraints in demand and operational models; and (4) developing more realistic time-of-day energy profiles for UAM vehicles in order to assess whether the current electrical grid can support UAM operations.
城市空中移动(UAM)如果成功,将打破城市交通。UAM 并不是交通领域的第一个颠覆性技术,最近的例子包括电动地面车辆(EVs)、自动驾驶地面车辆(AVs)和共享服务。在本文中,我们对从 2015 年 1 月至 2020 年 6 月发表的约 800 篇关于 UAM、EV 和 AV 领域的文章进行了元分析,并进行了比较和对比研究动向,以指导未来的 UAM 研究。在这个努力的同时,我们对与需求建模、运营和与现有基础设施的整合相关的文章进行了深入审查。我们利用元分析和全面审查的见解来指导未来的 UAM 研究方向。 我们确定的一些潜在研究方向包括:(1)开发更精细的需求模型,将个体采用城市空中交通(UAM)的时间考虑在内;(2)开发高保真度的 UAM 运营模拟模型,捕捉垂直起降机场位置、拓扑结构、需求、定价、调度和空域限制之间的相互作用;(3)在需求和运营模型中明确考虑单向需求和停车约束;以及(4)开发更真实的 UAM 车辆每天不同时间的能源消耗模型,以评估当前电网是否能支持 UAM 运营。

Keywords 关键词

Urban air mobility
Air taxi
Electric vehicle
Autonomous vehicle


1. Introduction 介绍

In recent years, there has been exponential growth in the number of publications related to aerial on-demand mobility1. A search of conference papers and journal publications in the American Institute of Aeronautics and Astronautics (AIAA) database shows that from 2015 to 2019, the number of annual publications in this area grew from 4 to 94. Interest in this area, commonly referred to as urban air mobility (UAM) or advanced air mobility (AAM)2, is driven in part by advancements in battery, distributed electric propulsion, and autonomy technologies that are leading to the development of a new class of aircraft, commonly referred to as electric vertical takeoff and landing (eVTOL) aircraft. These new eVTOL air taxis are expected to be safer, quieter, and less expensive to operate and maintain than existing vertical takeoff and landing aircraft, i.e., helicopters. Given current battery limitations, much of the research to date has focused on intracity or urban travel; however, extensions to regional and intercity missions are envisioned in the coming decades.
近年来,与空中按需出行相关的出版物数量呈指数增长。在美国航空航天学会(AIAA)数据库中搜索会议论文和期刊出版物,可以发现从 2015 年到 2019 年,这一领域的年度出版物数量从 4 篇增长到 94 篇。对这一领域的兴趣主要源于电池、分布式电动推进和自主技术的进步,这些技术正在推动一类新型飞机的发展,通常被称为电动垂直起降(eVTOL)飞机。预计这些新型 eVTOL 空中出租车将比现有的垂直起降飞机(即直升机)更安全、更安静、更经济。鉴于目前电池的限制,迄今为止的研究主要集中在城市内或城市间的旅行;然而,在未来几十年内,预计将扩展到区域和城际任务。

UAM represents a disruptive new technology, particularly if information-enabled platforms such as ride-hailing apps similar to those used by transportation network companies such as Lyft or Uber are used to connect operators with demand in real time. Never before has the potential for large-scale aerial operations within our cities been so real, as evidenced by the fact that in 2019 there were over 1,000 test flights of full-size eVTOL aircraft, and as of March 2020 at least 12 eVTOL aircraft were in the process of obtaining certification from the U.S. Federal Aviation Administration (FAA) (Dietrich and Wulff, 2020). To date, much of the research in UAM has been driven by the aerospace field and has focused on aircraft technology and aircraft operations, including the interface of UAM in the national airspace system (NAS); however, to be successful, UAM will need to integrate with our existing city infrastructure in ways that are acceptable to local communities, while providing service levels that offer time savings over existing modes at a price point that individuals are willing to pay.
UAM 代表着一项颠覆性的新技术,特别是如果像 Lyft 或 Uber 这样的交通网络公司使用的信息化平台(如叫车应用程序)实时连接运营商和需求。过去从未有过如此真实的大规模城市空中运营的潜力,这一点可以从 2019 年全尺寸电动垂直起降飞行器进行了 1000 多次试飞,并且截至 2020 年 3 月,至少有 12 架电动垂直起降飞行器正在美国联邦航空管理局(FAA)获得认证的过程中(Dietrich 和 Wulff,2020)。迄今为止,UAM 领域的大部分研究都是由航空航天领域推动的,主要关注飞机技术和飞机运营,包括 UAM 与国家空域系统(NAS)的接口;然而,要取得成功,UAM 需要以符合当地社区接受的方式与我们现有的城市基础设施相整合,同时提供比现有交通方式更节省时间的服务水平,以个人愿意支付的价格。

UAM is not the first disruptive technology in transportation. Electric and/or autonomous ground vehicles (EVs and/or AVs) are new technologies that are disrupting travel and share many similar characteristics with UAM. For example, like UAM, EVs and AVs need to integrate with existing urban infrastructure; their operations are heavily dependent on battery charging and fast-charging capabilities; and their profitability is influenced by factors including community acceptance and consumer willingness to pay. Additionally, because UAM is typically envisioned to employ fleets of shared-use aircraft, there are many similarities to ride sharing concepts from ground transportation, particularly shared autonomous vehicles (SAVs) and shared autonomous electric vehicles (SAEVs). Owing to these similarities, insights gained from the EV, AV, and ride sharing research communities will be applicable to the UAM community and can help inform future UAM research directions. The broad interrelationships between autonomous, electric, and sharing technologies in the transportation literature—and UAM’s place in this landscape—are illustrated in Fig. 1.
UAM 不是交通领域的第一个颠覆性技术。电动和/或自动驾驶地面车辆(EV 和/或 AV)是正在颠覆出行方式的新技术,与 UAM 有许多相似之处。例如,与 UAM 一样,EV 和 AV 需要与现有城市基础设施整合;它们的运营严重依赖电池充电和快速充电能力;它们的盈利能力受到社区接受度和消费者支付意愿等因素的影响。此外,由于 UAM 通常被设想为使用共享飞行器的车队,与地面交通的共享概念,特别是共享自动驾驶车辆(SAV)和共享电动自动驾驶车辆(SAEV)有许多相似之处。由于这些相似之处,从 EV、AV 和共享出行研究社区获得的见解将适用于 UAM 社区,并有助于指导未来的 UAM 研究方向。自动驾驶、电动和共享技术在交通文献中的广泛相互关系以及 UAM 在其中的位置如图 1 所示。

  1. Download : Download high-res image (56KB)
  2. Download : Download full-size image

Fig. 1. Interrelationship of autonomous, electric, and sharing technologies in the transportation literature.
图 1. 在交通文献中,自主、电动和共享技术之间的相互关系。

The objective of this paper is to provide a comprehensive review of UAM publications that have been published from 2015 to 2020 and to conduct a comparative analysis with publications during this same time period on EVs and AVs in order to help inform future UAM research. First, we compile a database of about 800 publications in the UAM, EV, and AV areas and classify their primary area(s) of research. Next, we conduct a meta-analysis comparing the overall research thrusts of the two communities. Finally, we conduct a more detailed analysis comparing the research approaches and results related to demand modeling, operations, and integration with existing infrastructure across the UAM and EV/AV areas. We use the comparative analysis to identify important factors that should be considered in the design and operation of UAM systems and areas of research that will potentially be important for the UAM community to investigate. To the best of our knowledge, our paper represents the first comprehensive review of UAM-related topics that conducts a comparative analysis of the ground vehicle and aircraft literatures for the purposes of identifying research opportunities and needs within UAM.
本文的目标是对 2015 年至 2020 年间发表的城市空中交通(UAM)相关出版物进行全面回顾,并与同一时期的电动车(EV)和自动驾驶车辆(AV)相关出版物进行比较分析,以帮助指导未来的 UAM 研究。首先,我们编制了一个包含约 800 篇 UAM、EV 和 AV 领域出版物的数据库,并对它们的主要研究领域进行分类。接下来,我们进行了一项元分析,比较了这两个领域的整体研究重点。最后,我们进行了更详细的分析,比较了 UAM 和 EV/AV 领域在需求建模、运营和与现有基础设施整合方面的研究方法和结果。我们利用比较分析来确定在设计和运营 UAM 系统时应考虑的重要因素,以及 UAM 社区可能需要研究的领域。据我们所知,我们的论文是首次对 UAM 相关主题进行全面回顾,并进行地面车辆和飞行器文献的比较分析,以确定 UAM 领域的研究机会和需求。

Multiple reviews have previously been conducted related to autonomous vehicles (Fagnant and Kockelman, 2015), shared autonomous vehicles (Narayanan et al., 2020), charging infrastructure and operational strategies for electric ground vehicles (Funke et al., 2019, Hardman et al., 2018, Shen et al., 2019), adoption of electric ground vehicles (Rezvani et al., 2015, Hardman, 2019), carsharing (Meisel and Merfeld 2018; Illgen and Höck, 2018) and ride-hailing (Tirachini, 2019). Our paper differs from these prior reviews in that we focus on identifying papers from the EV and AV areas that contain ideas, modeling assumptions, methods, or results that are applicable and can help inform UAM research. Our paper complements other reviews of UAM research, most notably that of Straubinger et al. (2020), that classifies UAM research areas into eight broad areas: air vehicles, regulation, infrastructure, operations, market actors, integration, acceptance, and modeling. It is our hope that our paper will become a resource document for those currently pursuing UAM research and will spur new interdisciplinary UAM research.
此前已经进行了多次关于自动驾驶车辆(Fagnant 和 Kockelman,2015)、共享自动驾驶车辆(Narayanan 等,2020)、电动地面车辆的充电基础设施和运营策略(Funke 等,2019;Hardman 等,2018;Shen 等,2019)、电动地面车辆采用(Rezvani 等,2015;Hardman,2019)、汽车共享(Meisel 和 Merfeld,2018;Illgen 和 Höck,2018)和打车服务(Tirachini,2019)的综述。我们的论文与这些先前的综述不同之处在于,我们专注于识别与电动车和自动驾驶领域相关的论文,这些论文包含了适用于城市空中交通研究的思想、建模假设、方法或结果。我们的论文是对城市空中交通研究的其他综述的补充,尤其是 Straubinger 等人(2020)的综述,该综述将城市空中交通研究领域划分为八个广泛的领域:空中车辆、监管、基础设施、运营、市场参与者、整合、接受度和建模。我们希望我们的论文能成为当前从事城市空中交通研究的人们的资源文件,并推动新的跨学科城市空中交通研究的开展。

The balance of this paper contains seven sections. Section 2 provides a brief history of UAM and an overview of different eVTOL aircraft designs. Section 3 documents the methodology we used to conduct our review and the results from the meta-analysis. The comparative analysis of UAM, EV, and AV research over a five-year period and directions for UAM research related to demand modeling, integration with existing infrastructure, and operations are discussed in 4 Demand modeling, 5 Integration with existing modes and infrastructure, 6 Operations, respectively. The paper concludes with a summary of main conclusions and limitations of the analysis.
本文的内容分为七个部分。第二部分简要介绍了城市空中交通(UAM)的历史,并概述了不同的电动垂直起降飞行器设计。第三部分记录了我们进行综述的方法和元分析的结果。第四部分讨论了在五年期间对 UAM、电动汽车(EV)和自动驾驶(AV)研究的比较分析,以及与需求建模、与现有基础设施的整合和运营相关的 UAM 研究方向。第五部分是需求建模,第六部分是与现有交通方式和基础设施的整合,第七部分是运营。文章最后总结了主要结论和分析的局限性。

2. History of UAM and eVTOL aircraft designs
2. UAM 和 eVTOL 飞行器设计的历史

This section provides an overview of current and prior UAM services and the different classes of eVTOL aircraft that are under development. The discussion explains the perceived market potential for UAM and points to the different business and operational strategies that aircraft manufacturers and UAM operators are pursuing.
本节概述了当前和以前的城市空中交通(UAM)服务以及正在开发中的不同类别的电动垂直起降飞行器(eVTOL)。讨论解释了 UAM 的市场潜力,并指出飞行器制造商和 UAM 运营商正在追求的不同商业和运营策略。

2.1. History of UAM service and value estimates for the emerging eVTOL UAM market
2.1. UAM 服务的历史和新兴 eVTOL UAM 市场的价值估计

The concept of urban air mobility is not new, with examples of UAM services using helicopters dating to the 1940s. From 1947 to 1971, Los Angeles Airways used helicopters to transport people and mail in the Los Angeles area, including between Disneyland and the Los Angeles International Airport (LAX). Los Angeles Airways experienced two accidents caused by mechanical failure in 1968 and subsequently ceased operations (Harrison, 2017, as referenced in Thipphavong et al., 2018). From 1953 to 1979, New York Airways used helicopters to fly passengers between Manhattan locations and the three major airports in New York City (Newark Liberty International Airport [EWR], LaGuardia Airport [LGA], and John F. Kennedy International Airport [JFK]). This service similarly ceased due to several accidents caused by mechanical failure (Witken, 1979, as referenced in Thipphavong et al., 2018, Mayor and Anderson, 2019). The cost of a passenger ticket on a New York Airways shuttle was between $5 and $9, or about $47 to $86 in 2019 dollars (Mayor and Anderson, 2019). These early UAM operations successfully operated for more than two decades, ultimately ceasing operations due to safety concerns. These historic examples provide evidence of the potential value to consumers for similar (albeit safer) UAM services today.
城市空中移动的概念并不新鲜,早在 1940 年代就有使用直升机的 UAM 服务的例子。从 1947 年到 1971 年,洛杉矶航空公司在洛杉矶地区使用直升机运送人员和邮件,包括迪士尼乐园和洛杉矶国际机场(LAX)之间的运输。洛杉矶航空公司在 1968 年发生了两起由机械故障引起的事故,随后停止运营(Harrison, 2017,参考文献为 Thipphavong 等,2018)。从 1953 年到 1979 年,纽约航空公司使用直升机在曼哈顿各地和纽约市的三个主要机场(纽瓦克自由国际机场[EWR],拉瓜迪亚机场[LGA]和约翰·肯尼迪国际机场[JFK])之间运送乘客。由于多起机械故障引起的事故,该服务也停止运营(Witken, 1979,参考文献为 Thipphavong 等,2018,Mayor 和 Anderson,2019)。纽约航空公司的乘客票价在 5 美元到 9 美元之间,相当于 2019 年的 47 美元到 86 美元(Mayor 和 Anderson,2019)。这些早期的 UAM 运营成功运营了二十多年,最终因安全问题而停止运营。 这些历史例子为今天类似(尽管更安全)的城市空中交通服务对消费者的潜在价值提供了证据。

Several helicopter operators are providing on-demand urban passenger air service. BLADE operates3 between various locations in Manhattan and one of the three main airports in New York City (JFK, LGA, and EWR). Flights are bookable within 30 min of departure and the one-way cost is $195; additional charges starting at $85 apply for baggage above 25 lb (that is transported via a ground service), last-minute bookings, and cancellations received within three hours of departure (BLADE, 2020). In 2019, Uber partnered with HeliFlite4 to offer flights from Manhattan to JFK airport between the hours of 1 PM and 6 PM, Monday through Friday. Uber Copter can be booked on demand or up to five days before the flight (Matthews, 2019). The helicopter option appears on the Uber app and the one-way cost ranges from $200 to $225 per person and includes one piece of luggage up to 50 lb (Ballentine, 2019, Uber, 2020). In 2016, Airbus started Voom, an on-demand booking platform that connected travelers to helicopter service providers in São Paulo, Brazil, and later expanded service to Mexico City and the San Francisco Bay Area before permanently ceasing operations due to COVID-19 in April of 2020 (Airbus, 2020b). Flights in Mexico City were bookable within 60 min of departure of the flight or could be reserved up to seven days in advance (Airbus, 2018).
几家直升机运营商提供按需的城市客运航空服务。BLADE 在曼哈顿的各个地点和纽约市的三个主要机场(JFK、LGA 和 EWR)之间运营航班。航班可在起飞前 30 分钟内预订,单程费用为 195 美元;超过 25 磅的行李(通过地面服务运输)将收取额外费用,临时预订和起飞前三小时内取消预订也将收取额外费用(BLADE,2020 年)。2019 年,优步与 HeliFlite 合作,提供从曼哈顿到 JFK 机场的航班,时间为周一至周五的下午 1 点至下午 6 点。优步直升机可以按需预订,或在飞行前最多提前五天预订(Matthews,2019 年)。直升机选项出现在优步应用程序上,单程费用为每人 200 至 225 美元,包括一件重量不超过 50 磅的行李(Ballentine,2019 年,Uber,2020 年)。 2016 年,空中客车公司推出了 Voom,一个按需预订平台,将旅客与圣保罗、巴西的直升机服务提供商连接起来,并在之后扩展到墨西哥城和旧金山湾区,但由于 COVID-19 的影响,该平台于 2020 年 4 月永久停止运营(Airbus,2020b)。在墨西哥城,航班可以在起飞前 60 分钟内预订,或者提前最多七天预订(Airbus,2018)。

These modern-day on-demand helicopter services have been important to UAM researchers, as they provided information about customer preferences (booking patterns, willingness to pay, most popular routes) and “operational challenges related to a lack of infrastructure, public acceptance, [and] on-demand versus scheduled routes” (Airbus, 2020b). They help set the context for how the UAM community is envisioning the possibilities for stimulating demand through lower per-passenger mile (pax-mile) operating costs with new eVTOL aircraft.

BLADE and Uber Copter charge about $30 per pax-mile in Manhattan, whereas Voom charged about $10 per pax-mile (Booz Allen Hamilton, 2018, Reiche et al., 2018). Uber Elevate estimates the cost of a passenger helicopter service at about $8.93 per pax-mile (Holden, 2018), and McKinsey and Company estimates the cost between $6 and $8 per seat-mile5 (Johnson, Riedel, and Sahdev, 2020). Uber Elevate has reported that they anticipate at the launch of their on-demand air taxi that service costs will be $5.73 per pax-mile but will decrease in the near term to $1.84 by increasing utilization through ride-hailing (Holden, 2018). In later conferences, Uber Elevate noted that at $2.00 per pax-mile, the flight operating cost would be $662/hr as compared to $1,253/hr that is more common among helicopters operating today (Uber Elevate, 2019). On an hourly basis, longer-term, Uber Elevate anticipates that advancements in manufacturing and autonomy will decrease both fixed and variable costs, resulting in $0.44 per pax-mile cost; in comparison, in 2017, the American Automobile Association (AAA) estimated the full cost of auto ownership in the U.S. to be between $0.46 and $0.61 per mile (Holden, 2018, AAA, 2017). The McKinsey report estimates near-term costs between $2.50 and $4.50 per seat-mile and long-term costs between $0.50 and $2.50 (Johnson, Riedel, and Sahdev, 2020). The UAM cost estimates provided by Uber Elevate and McKinsey and Company are optimistic compared to other reports, such as one conducted for the National Aeronautics and Space Administration (NASA) that forecasts the costs of a five-seat eVTOL at $6.25/pax-mile in the near term but “in the long term, operational efficiency, autonomy, technology improvements may decrease costs by 60%” (i.e., $3.75/pax-mile) (Booz Allen Hamilton, 2018, Reiche et al., 2018). Although the cost estimates of providing UAM service vary, near-term and long-term eVTOL operations will likely operate at lower costs compared to current helicopter service, resulting in more demand for UAM service. In comparison, costs for shared autonomous vehicles in the 2030s to 2040s have been estimated by multiple studies to range from $0.29 to $0.49 per-seat-mile (Johnson, 2015, Albright et al., 2015, Chen et al., 2016, Corwin et al., 2016), with lower cost estimates corresponding to purpose-built SAVs. These costs range from five times to ten times lower than the UAM cost estimates noted above in the mid-term to near-parity with UAM in the long-term; however, direct comparisons are challenging at present because of the uncertainty in the technologies and development timelines for both SAVs and UAM.
在曼哈顿,BLADE 和 Uber Copter 每人每英里收费约 30 美元,而 Voom 每人每英里收费约 10 美元(Booz Allen Hamilton,2018 年,Reiche 等,2018 年)。Uber Elevate 估计乘客直升机服务的成本约为每人每英里 8.93 美元(Holden,2018 年),而麦肯锡公司估计每个座位每英里的成本在 6 至 8 美元之间(Johnson,Riedel 和 Sahdev,2020 年)。Uber Elevate 报告称,他们预计在他们的按需空中出租车推出时,服务成本将为每人每英里 5.73 美元,但通过增加乘车率,这一成本将在短期内降至 1.84 美元(Holden,2018 年)。在后来的会议上,Uber Elevate 指出,以每人每英里 2.00 美元的价格,飞行运营成本将为每小时 662 美元,而现今直升机的运营成本通常为每小时 1253 美元(Uber Elevate,2019 年)。从长期来看,Uber Elevate 预计制造业和自主技术的进步将降低固定和可变成本,导致每人每英里成本为 0.44 美元;相比之下,2017 年,美国汽车协会(AAA)估计美国汽车所有权的全部成本在 0.46 至 0.61 美元每英里(Holden,2018 年,AAA,2017 年)。麦肯锡报告估计,短期成本在每个座位英里 2.50 美元至 4.50 美元之间,长期成本在每个座位英里 0.50 美元至 2.50 美元之间(Johnson,Riedel 和 Sahdev,2020 年)。Uber Elevate 和麦肯锡公司提供的城市空中出行成本估计相对于其他报告来说是乐观的,例如美国国家航空航天局(NASA)进行的一项研究预测,短期内五座电动垂直起降飞行器的成本为每人英里 6.25 美元,但“长期来看,运营效率、自主性和技术改进可能会降低成本 60%”(即每人英里 3.75 美元)(Booz Allen Hamilton,2018 年,Reiche 等,2018 年)。尽管提供城市空中出行服务的成本估计各不相同,但短期和长期的电动垂直起降飞行器运营成本可能会比目前的直升机服务更低,从而导致对城市空中出行服务的需求增加。相比之下,2030 年至 2040 年共享自动驾驶车辆的成本估计在每个座位英里 0.29 美元至 0.49 美元之间(Johnson,2015 年,Albright 等,2015 年,Chen 等,2016 年,Corwin 等,2016 年),较低的成本估计对应于专门设计的自动驾驶车辆。 这些成本比上述中期的 UAM 成本估计低五倍到十倍,长期来看与 UAM 成本基本持平;然而,由于 SAV 和 UAM 的技术和发展时间表的不确定性,目前很难进行直接比较。

Despite these differences, what is notable is the extent that research in the eVTOL area has grown in the last five years, and how quickly some manufacturers are moving toward certifying their aircraft. Part of the interest in designing eVTOL aircraft is due to the value many believe is present for passenger UAM markets. Table 1 summarizes these valuation estimates for different geographies. For U.S. markets, these estimates range from $500B to $1B in the near term to $1B–$3.6B in 2025 to $17.7B in 2040. Globally, these estimates range from $0.3B to $3B in the short-term (2018 to 2025 time frame) to $32B by 2035. Many of the valuations of the UAM markets distinguish between intracity markets and intercity or regional markets, reflecting that as battery technologies advance, eVTOL aircraft will be able to fly longer missions. Looking ahead, Roland Berger and Porsche forecast larger UAM valuations for intracity taxis and airport shuttles than for regional intercity flights (Roland Berger, 2018, Porsche Consulting, 2018, as quoted in Volocopter, 2018).
尽管存在这些差异,值得注意的是,在过去五年中,电动垂直起降飞行器领域的研究程度以及一些制造商向认证其飞机的方向迅速发展。设计电动垂直起降飞行器的兴趣部分是因为许多人认为乘客城市空中出行市场具有价值。表 1 总结了不同地理区域的估值。对于美国市场,这些估值在短期内为 500 亿美元至 10 亿美元,到 2025 年为 10 亿美元至 36 亿美元,到 2040 年为 177 亿美元。全球范围内,这些估值在短期内为 30 亿美元至 300 亿美元(2018 年至 2025 年时间范围),到 2035 年为 320 亿美元。许多对城市空中出行市场的估值区分了城市内市场和城际或区域市场,反映出随着电池技术的进步,电动垂直起降飞行器将能够执行更长的任务。展望未来,罗兰贝格和保时捷预测城市内出租车和机场班车的城市空中出行市场估值将超过区域城际航班(Roland Berger,2018 年,Porsche Consulting,2018 年,引自 Volocopter,2018 年)。

Table 1. Valuation estimates for passenger UAM markets.
表 1. 乘客城市空中交通市场的估值。

Report 报告Market 市场Geography 地理学Valuation 估值
Booz Allen Hamilton1
Airport shuttle and 机场班车和
intracity air taxi 市内空中出租车
U.S. 美国$500B unconstrained market
5000 亿美元无限制市场

0.5% captured near term at $2.5B
0.5%在短期内占据了 25 亿美元。
Deloitte2 德勤Intracity and regional markets
U.S. 美国$1B in 2025 for intracity
2025 年城市内部的 10 亿美元

$13.8B in 2040 for intracity
2040 年城市内部的投资为 138 亿美元。

$2.6B in 2025 for regional
2025 年地区的 26 亿美元

$3.9B in 2040 for regional
2040 年地区的 39 亿美元
Frost and Sullivan3
Passenger service 乘客服务Global 全球$0.3 M in 2018 to $3B in 2023
2018 年为 0.3 亿美元,2023 年为 30 亿美元。
KPMG4 毕马威Intracity and regional service
Global 全球12 M enplanements per year by 2040
到 2040 年每年有 12 百万人次的登机次数

400 M enplanements by 2050
到 2050 年,预计将有 4 亿人次的乘机登机
Porsche Consulting5
Passenger service 乘客服务
(intracity and regional) (城市内和地区间)
Global 全球$1B by 2025 2025 年达到 10 亿美元
$21B intracity by 2035 到 2035 年,城市内部的投资将达到 210 亿美元
$11B regional by 2035 到 2035 年,地区总额将达到 110 亿美元

References: 1Booz Allen Hamilton (2018) and Reiche et al. (2018); 2Lineberger et al. (2019); 3as quoted in eHang, 2020b, eHang, 2020a; 4Mayor and Anderson (2019); 5Porsche Consulting (2018).
参考文献: 1 Booz Allen Hamilton(2018)和 Reiche 等人(2018); 2 Lineberger 等人(2019); 3 如 eHang(2020b)、eHang(2020a)所引用; 4 Mayor 和 Anderson(2019); 5 Porsche Consulting(2018)。

2.2. Overview of eVTOL aircraft designs
2.2. 电垂直起降飞行器设计概述

Worldwide, there are multiple efforts focused on designing eVTOL aircraft, and more than $2B has been invested in this industry (Sherman, 2020). Collectively, these designs represent fundamentally different design concepts. Multiple publications provide overviews of the different technical specifications and characteristics associated with eVTOL aircraft (e.g., see Roland Berger, 2018, Porsche Consulting, 2018). The Vertical Flight Society (VFS) provides one of the more thorough overviews of the different types of eVTOL aircraft and maintains a database of known eVTOL designs (Electric VTOL News™, 2020). According to VFS, as of March 5, 2020, there were a total of 260 aircraft6 that included 99 vectored thrust, 39 lift + cruise, 26 wingless multicopters, 46 hover bikes/flying devices7 and 20 eHelos and eGyros (Sherman, 2020). Across these designs, there are large variations in the number of seats, speed, and range.
全球范围内,有多个项目致力于设计电动垂直起降飞行器(eVTOL),已经投资了超过 20 亿美元在这个行业(Sherman,2020)。这些设计代表了根本不同的设计概念。多个出版物提供了关于 eVTOL 飞行器的不同技术规格和特性的概述(例如,参见 Roland Berger,2018 年,Porsche Consulting,2018 年)。垂直飞行学会(VFS)提供了对不同类型的 eVTOL 飞行器的更全面的概述,并维护着已知 eVTOL 设计的数据库(Electric VTOL News™,2020)。根据 VFS 的数据,截至 2020 年 3 月 5 日,共有 260 架飞行器,其中包括 99 架向量推力、39 架升力+巡航、26 架无翼多旋翼、46 架悬停摩托车/飞行装置和 20 架电动直升机和电动陀螺(Sherman,2020)。在这些设计中,座位数、速度和航程存在很大的差异。

Vectored thrust aircraft can use any of their thrusters8 for both lift and cruise; representative examples include the Lilium Jet (2 to 5 seats; 186 mph; 186-mile range), Airbus A3 Vahana (1 seat; 118 mph; 31-mile range), Bell Nexus 4EX (5 seats; 150 mph; 150-mile range), and Joby S4 (5 seats; 200 mph; 150-mile range) (Lilium, 2020, Hawkins, 2019, Airbus, 2020a, Bell Flight, 2020, Pope, 2019, Goldstein, 2019, Bogaisky, 2020). According to Sherman, vectored thrust designs—the most common among potential eVTOL designs—will likely be the most efficient eVTOL aircraft but also likely the most difficult to bring to market due to the complexity of designing the aircraft to safety transition between vertical flight and forward flight.
矢量推力飞机可以使用任何一个推力器来实现升力和巡航;代表性的例子包括 Lilium Jet(2 至 5 个座位;186 英里/小时;186 英里航程)、Airbus A Vahana(1 个座位;118 英里/小时;31 英里航程)、Bell Nexus 4EX(5 个座位;150 英里/小时;150 英里航程)和 Joby S4(5 个座位;200 英里/小时;150 英里航程)(Lilium,2020 年,Hawkins,2019 年,Airbus,2020a,Bell Flight,2020 年,Pope,2019 年,Goldstein,2019 年,Bogaisky,2020 年)。根据 Sherman 的说法,矢量推力设计是潜在 eVTOL 设计中最常见的,可能是最高效的 eVTOL 飞机,但由于设计飞机在垂直飞行和前向飞行之间的安全过渡的复杂性,也可能是最难进入市场的。

The lift + cruise is another popular aircraft category under development that has two sets of independent thrusters—one set that is used only for cruise and a second set that is used only for vertical lift. Lift during cruise flight is provided by one or several wings. Representative examples include the Aurora Flight Sciences Pegasus (2 seats; 112 mph; 50-mile range; Aurora Flight Sciences, 2020), EmbraerX Eve9 (5 seats; speed and range not public), and Wisk Cora (2 seats; 100 mph; 25-mile range) (Electric VTOL News™, n.d.a, Electric VTOL News™, n.d.b, EmbraerX, 2020, Wisk, 2020).
垂直起降+巡航是另一种正在开发中的流行飞行器类型,它有两套独立的推进器,一套仅用于巡航,另一套仅用于垂直起降。巡航飞行时,提供升力的是一个或多个机翼。代表性的例子包括 Aurora Flight Sciences Pegasus(2 个座位;112 英里/小时;50 英里航程;Aurora Flight Sciences,2020)、EmbraerX Eve(5 个座位;速度和航程未公开)和 Wisk Cora(2 个座位;100 英里/小时;25 英里航程)(Electric VTOL News™,无日期 a,Electric VTOL News™,无日期 b,EmbraerX,2020,Wisk,2020)。

Wingless multicopters are another common design that use their thrusters to produce lift not only for takeoff and vertical flight but for cruise, as well. Representative examples of these aircraft include the Volocopter VC200 (2 seats; 50–62 mph; 19-mile range), the eHang 216 (1 seat; 81 mph; 22-mile range), and the LIFT Aircraft Hexa (1 seat; 60 mph; 12–15 mile range) (Volocopter, 2018, Volocopter, 2020, eHang, 2020a, LIFT Aircraft, 2020). The LIFT Aircraft Hexa is an ultralight passenger air vehicle that seats one passenger who controls the aircraft. Ultralight aircraft will be restricted to recreational use and speeds of under 60 mph but will likely be some of the first eVTOL aircraft to enter the market, as they do not require aircraft and pilot certification under Federal Aviation Regulations Part 103 (FAA, 1982, as noted by Sherman, 2020).
无翼多旋翼机是另一种常见设计,它们利用推进器产生升力,不仅用于起飞和垂直飞行,还用于巡航。这类飞机的代表性例子包括 Volocopter VC200(2 座位;50-62 英里/小时;19 英里航程)、eHang 216(1 座位;81 英里/小时;22 英里航程)和 LIFT Aircraft Hexa(1 座位;60 英里/小时;12-15 英里航程)(Volocopter,2018 年;Volocopter,2020 年;eHang,2020a 年;LIFT Aircraft,2020 年)。LIFT Aircraft Hexa 是一种超轻型乘客飞行器,可容纳一名乘客控制飞行。超轻型飞行器将受到限制,仅限于娱乐用途和时速不超过 60 英里,但很可能成为首批进入市场的电动垂直起降飞行器,因为它们不需要根据联邦航空法规第 103 部分获得飞行器和飞行员认证(FAA,1982 年,如 Sherman,2020 年所述)。

Rotorcraft designs are another area being considered for UAM applications. These concepts include both electric helicopters and novel autogyros (i.e., helicopter-like aircraft in which the rotor rotates not by shaft power from the engine but by the force of air flowing through it; propulsion in forward flight is provided by a separate propeller). Representative examples include the Jaunt Air Mobility gyrocopter (5 seats; 175 mph; range unknown) and the Pal-V Pioneer flying car (2 seats; 99–112 mph; 250–300-mile range) (Jaunt Air Mobility, 2020, Blain, 2020, PAL-V, 2020). Although the popular press often refers to UAM/eVTOL aircraft as “flying cars,” these aircraft typically do not meet the historical definition of a “roadable aircraft” that can be both driven on the ground as a car and flown as an airplane. However, the Pal-V is a roadable aircraft. The Pal-V is not an electric aircraft (hence not an eVTOL) and is intended for personal use and not necessarily UAM; however, it is representative of the novel rotorcraft configurations being explored in the UAM field. Fig. 2 provides examples of each of the aircraft designs discussed above.
旋翼飞行器设计是另一个被考虑用于城市空中交通应用的领域。这些概念包括电动直升机和新颖的自转旋翼机(即,类似直升机的飞行器,其旋翼不是通过发动机的轴动力旋转,而是通过空气流经旋翼产生的力;前进飞行时由独立的螺旋桨提供推进力)。代表性的例子包括 Jaunt Air Mobility 的陀螺旋翼机(5 个座位;175 英里/小时;航程未知)和 Pal-V Pioneer 飞行汽车(2 个座位;99-112 英里/小时;250-300 英里航程)(Jaunt Air Mobility, 2020, Blain, 2020, PAL-V, 2020)。尽管大众媒体经常将城市空中交通/电动垂直起降飞行器称为“飞行汽车”,但这些飞行器通常不符合“可上路飞行器”的历史定义,即既可以作为汽车在地面上行驶,又可以作为飞机飞行。然而,Pal-V 是一种可上路飞行器。Pal-V 不是电动飞行器(因此不是电动垂直起降飞行器),旨在个人使用,而不一定是城市空中交通;然而,它代表了城市空中交通领域正在探索的新颖旋翼飞行器配置。图 2 提供了上述每种飞行器设计的示例。

  1. Download : Download high-res image (187KB)
  2. Download : Download full-size image

Fig. 2. Examples of different UAM aircraft designs.
图 2. 不同的城市空中交通飞行器设计示例。

Image credits: Joby Aviation, 2021, Lilium, 2021, Beta, 2021, Wisk Aero, 2021, Lift, 2021, Jaunt Air Mobility. 2021.
图片来源:Joby Aviation,2021 年,Lilium,2021 年,Beta,2021 年,Wisk Aero,2021 年,Lift,2021 年,Jaunt Air Mobility,2021 年。

What is clear from the discussion above is that there is currently a lack of convergence in designs and underlying business models envisioned by the eVTOL community. The non-convergence of design concepts reflects the novelty of these new battery and electric propulsion technologies and uncertainties regarding how these new technologies will impact aircraft performance. It also reflects a lack of consensus on which missions (or market segments) these aircraft can profitably serve, and whether the aircraft should be flown by the passenger, an onboard pilot, a remote pilot, or autonomously.

3. Meta-Analysis of research in UAM, EV, and AV
3. UAM、EV 和 AV 研究的元分析

3.1. Methodology and scope of review
3.1. 检查的方法和范围

To identify relevant publications in UAM, we conducted a keyword search of “urban air mobility,” “air taxi,” and “UAM” in the AIAA publication database. A similar search was conducted on Scopus using the same keywords but adding exclusion terms for “drone” and “UAV.” The searches were initially conducted in the spring of 2020 and were updated in mid-July 2020.
为了确定与城市空中移动相关的出版物,我们在 AIAA 出版物数据库中进行了关键词搜索,关键词包括“城市空中移动”、“空中出租车”和“UAM”。在 Scopus 上进行了类似的搜索,使用相同的关键词,但添加了排除词“无人机”和“无人机”。这些搜索最初是在 2020 年春季进行的,并在 2020 年 7 月中旬进行了更新。

The search results included journal and conference publications relevant to UAM that were published from January 1, 2015, to June 30, 2020, in which the aforementioned keywords appeared in the title or abstract. A total of 251 publications were identified from the AIAA database and an additional 61 from the Scopus search.
搜索结果包括了从 2015 年 1 月 1 日到 2020 年 6 月 30 日发表的与城市空中交通相关的期刊和会议论文,其中标题或摘要中出现了上述关键词。从 AIAA 数据库中找到了 251 篇论文,从 Scopus 搜索中找到了额外的 61 篇。

To identify relevant EV and AV articles, we reviewed the table of contents of key journals from the transportation field from January 2015 to June 202010 and identified articles that were relevant based on their titles and abstracts. We explicitly decided not to use a keyword search for this part of the analysis, so that we could go through the titles and identify publications that were relevant to UAM research, such as ride-hailing or carsharing, that may not directly fall into searches returned using EV and AV keywords. EV, AV, carsharing, and ride-hailing are synergistic areas within the ground transportation field, given interest in using future AVs as an electric fleet that operates as a carsharing or ride-hailing service. However, a simple search of “ridesharing” on Scopus of publications published since 2015 conducted in September 2020 returned over 2500 publications. Thus, we opted to use a more directed approach by carefully reviewing titles and abstracts from selected journals to identify papers in the ground transportation literature that showed potential for having ideas, concepts, methods, or results that could inform UAM research.
为了确定相关的电动车和自动驾驶文章,我们回顾了 2015 年 1 月至 2020 年 6 月期间交通领域的重要期刊的目录,并根据标题和摘要确定了相关的文章。我们明确决定不使用关键词搜索来进行分析的这一部分,这样我们可以浏览标题并确定与城市空中交通研究相关的出版物,例如打车或共享汽车,这些出版物可能不会直接通过使用电动车和自动驾驶关键词返回。电动车、自动驾驶、共享汽车和打车是地面交通领域内的协同领域,因为人们有兴趣将未来的自动驾驶车辆作为电动车队运营的共享汽车或打车服务。然而,仅在 2020 年 9 月在 Scopus 上进行的自 2015 年以来发表的出版物的“打车”简单搜索就返回了超过 2500 篇文章。因此,我们选择了更有针对性的方法,通过仔细审查选定期刊的标题和摘要,确定地面交通文献中可能具有为城市空中交通研究提供思路、概念、方法或结果的论文。

Given our overarching objective in comparing the EV/AV and UAM fields is to glean insights from the EV/AV areas that may be applicable to the UAM area, we excluded some papers in the EV/AV areas that were not directly applicable to the UAM field. For example, papers that discuss strategies for safely merging AV ground vehicles into traffic are not applicable to UAM given UAM has another dimension for conflict avoidance and different traffic management rules than ground transportation modes. Similarly, when doing a detailed analysis of a particular area (such as demand segmentation), we tagged all articles that fit into the category, but then focused our in-depth discussion on the subset of articles most relevant to UAM (e.g., we exclude a discussion of how EV vehicle characteristics like acceleration influence EV purchases).

We reviewed articles from the following journals—the number of articles in total and those we included in our analysis are shown in parentheses: Transportation Research Part A (1392 published; 125 inventoried); Transportation Research Part B (970 published; 62 inventoried); Transportation Research Part C (1971 published; 100 inventoried); Transportation Research Part D (1281 published; 124 inventoried), Transportation (545 published; 51 inventoried); and Transportation Science (444 published; 16 inventoried).
我们回顾了以下期刊的文章 - 总共发表的文章数和我们在分析中包含的文章数分别如下:《交通研究 A 部分》(发表 1392 篇;纳入分析的 125 篇);《交通研究 B 部分》(发表 970 篇;纳入分析的 62 篇);《交通研究 C 部分》(发表 1971 篇;纳入分析的 100 篇);《交通研究 D 部分》(发表 1281 篇;纳入分析的 124 篇);《交通》(发表 545 篇;纳入分析的 51 篇);以及《交通科学》(发表 444 篇;纳入分析的 16 篇)。

The final number of articles we identified includes 312 for UAM and 478 for EV/AV research. For each of the 790 articles, we identified research themes by associating up to six keywords based on a review of the abstracts (or where unclear, a review of the articles). For each publication, we recorded author and publication information. Information for each of these 790 articles, including DOI links, are included in an Excel sheet as a supplemental document to this paper. Co-author Garrow, an expert in travel behavior modeling from civil engineering, tagged the articles related to EVs and AVs, and co-author German, an expert in aircraft design from aerospace engineering, tagged the articles related to UAM. While the subject classifications are arguably subjective, they nonetheless enable us to identify high-level trends across the fields.
我们确定的文章总数包括 312 篇关于城市空中交通(UAM)的文章和 478 篇关于电动车辆/自动驾驶(EV/AV)研究的文章。对于这 790 篇文章中的每一篇,我们通过审查摘要(或在不清楚的情况下,审查全文)来确定研究主题,并根据最多六个关键词进行关联。对于每一篇出版物,我们记录了作者和出版信息。这 790 篇文章的信息,包括 DOI 链接,作为本文的补充文件,以 Excel 表格的形式提供。共同作者 Garrow 是土木工程中旅行行为建模方面的专家,他标记了与 EV 和 AV 相关的文章;共同作者 German 是航空航天工程中飞机设计方面的专家,他标记了与 UAM 相关的文章。虽然主题分类可能有主观性,但它们仍然使我们能够识别出各个领域的高级趋势。

3.2. UAM publications 3.2. UAM 出版物

Based on our review of UAM-related articles, we conducted a meta-analysis focused on two overarching themes: (1) categorization of the technical content of the articles, and (2) analysis of the affiliations of the authors. The former theme provides insights into the breadth and depth of the topics addressed in UAM research, and the latter provides insights into what nations, organizations, and individuals are actively focused on UAM research.
根据我们对与 UAM 相关的文章的审查,我们进行了一项以两个主题为重点的元分析:(1)对文章技术内容的分类,以及(2)对作者所属机构的分析。前一个主题提供了关于 UAM 研究所涉及的主题广度和深度的见解,而后一个主题则提供了关于哪些国家、组织和个人正在积极关注 UAM 研究的见解。

To categorize the content in the UAM-related articles, we first identified low-level topic categories that were present in multiple articles, and we created corresponding content tags. In defining these categories, we were guided in part by our knowledge of new technical topic areas related to eVTOL aircraft that are being actively addressed within the UAM community, e.g. “Distributed Electric Propulsion” and “Aero-Propulsive Interactions.” We then grouped related low-level tags hierarchically under higher-level categories associated with traditional research disciplines related to aircraft technology and operations, e.g. “Propulsion,” “Aerodynamics,” and “Simulation.” Finally, we grouped these higher-level categories into two overarching categories: “Aircraft Technology” and “Market and Operations.” The resulting categorization reflects our attempt to identify and group common themes in UAM research cogently; however, we do not claim that the categorization is mutually exclusive, collectively exhaustive, or unequivocal.
为了对与城市空中交通(UAM)相关的文章内容进行分类,我们首先确定了多篇文章中存在的低级主题类别,并创建了相应的内容标签。在定义这些类别时,我们部分地参考了我们对与电动垂直起降飞行器相关的新技术主题领域的了解,这些领域正在 UAM 社区中积极探讨,例如“分布式电动推进”和“空气推进相互作用”。然后,我们按照与飞机技术和运营相关的传统研究学科将相关的低级标签按层次分组,例如“推进系统”、“空气动力学”和“仿真”。最后,我们将这些高级类别分为两个总体类别:“飞机技术”和“市场与运营”。这种分类反映了我们试图明确地识别和分组 UAM 研究中的共同主题的努力;然而,我们并不声称这种分类是相互独立、完全穷尽或明确无误的。

The hierarchical categories are shown in Fig. 3. The numbers in parentheses indicate the number of articles with lower-level tags assigned to the corresponding category. The number of articles indicated for each higher-level parent category are summative of all children tags for the category. Note that any one article is likely to have been assigned more than one tag based on the breadth of topics covered in the article. The individual low-level content tags corresponding to the overall categories are not shown in Fig. 3 to limit the size of the figure; however, these tags are provided in the spreadsheet provided as supplemental material to this article.
图 3 显示了层次分类。括号中的数字表示分配给相应类别的较低级别标签的文章数量。每个较高级别父类别所示的文章数量是该类别的所有子标签的总和。请注意,根据文章涵盖的主题广度,任何一篇文章很可能被分配多个标签。图 3 中未显示与整体类别相对应的个别低级内容标签,以限制图的大小;然而,这些标签在提供给本文的补充材料电子表格中提供。

  1. Download : Download high-res image (166KB)
  2. Download : Download full-size image

Fig. 3. UAM themes from AIAA and Scopus search, January 2015 to June 2020.
图 3. 从 AIAA 和 Scopus 搜索中得出的 UAM 主题,时间范围为 2015 年 1 月至 2020 年 6 月。

The first observation from this analysis is that current articles on UAM have a nearly even split of content related to “Aircraft Technology” (295 papers) and “Market and Operations” (248 papers). This thematic balance likely reflects an understanding within the community of the “chicken-and-egg” issue associated with the emergence of UAM, i.e., aircraft must be technically capable of serving the missions required for profitable large-scale UAM operations, and a market must exist for the types of missions and operations that can be supported given the technological limitations of emerging aircraft. A concrete example of this interplay is related to eVTOL aircraft with battery electric propulsion. These aircraft have the capability of being much quieter and more economical than current generation helicopters, potentially allowing widespread operations in urban environments at low ticket prices. However, battery electric eVTOL aircraft have very limited range and speed capability because of the low specific energy of current and near-term batteries, potentially limiting the potential for the aircraft to serve an adequate network of origins and destinations and to offer adequate travel time savings compared to other modes when trip times are dominated by ingress and egress on short-ranged flights.
这项分析的第一个观察结果是,目前关于城市空中交通(UAM)的文章几乎平均分布在“飞行器技术”(295 篇论文)和“市场与运营”(248 篇论文)两个主题上。这种主题平衡可能反映了社区对于 UAM 出现的“先有鸡还是先有蛋”的问题的理解,即飞行器必须在技术上具备为盈利的大规模 UAM 运营提供所需任务的能力,同时市场必须存在于能够在新兴飞行器技术限制下支持的任务和运营类型。一个具体的例子是与电池电动垂直起降飞行器(eVTOL)相关的相互作用。这些飞行器具有比当前一代直升机更安静和更经济的能力,潜在地可以以低票价在城市环境中进行广泛运营。 然而,由于当前和近期电池的比能量较低,电池电动垂直起降飞行器的航程和速度能力非常有限,这可能限制了飞行器在为足够的起点和终点网络提供服务以及在行程时间主要由短程飞行的进出口所主导时,相比其他交通方式提供足够的旅行时间节约的潜力。

Within the “Aircraft Technology” category, the majority of papers had content related to “Propulsion” (82 papers) and “Aircraft Design and Performance” (92 papers). The “Propulsion” category includes papers with content related to new propulsion architectures relevant for UAM, including “Electric Propulsion” (battery powered propulsion; 29 papers), “Distributed Electric Propulsion” (multiple electric motors powering rotors or fans located in multiple locations on the aircraft; 9 papers), and “Hybrid Propulsion” (propulsion powered by both batteries and a combustion engine; 18 papers) as well as papers with a focus on associated propulsion component technologies such as “Battery” (12 papers) and “Fuel Cell” (4 papers). The majority of the papers in the “Aircraft Design and Performance” category is related to “Design Methods” (30 papers) or present a “Concept Study” of the design and performance of a novel UAM aircraft configuration (41 papers). The “Aircraft Technology” category has a significant number of papers related to “Aerodynamics” (43 papers), as well, with many papers focused specifically on the aerodynamics associated with eVTOL aircraft and other distributed propulsion configurations, i.e. “Propeller, Rotor, Ducted Fan” (16 papers) and “Aero-Propulsive Interaction” (7 papers).
在“飞机技术”类别中,大部分论文内容与“推进系统”(82 篇论文)和“飞机设计与性能”(92 篇论文)有关。 “推进系统”类别包括与城市空中交通相关的新型推进系统架构的论文,包括“电动推进”(电池驱动推进系统;29 篇论文),“分布式电动推进”(多个电动马达驱动位于飞机多个位置的旋翼或风扇;9 篇论文)和“混合推进”(电池和燃烧发动机共同驱动推进系统;18 篇论文),以及关注相关推进系统组件技术的论文,如“电池”(12 篇论文)和“燃料电池”(4 篇论文)。 “飞机设计与性能”类别中大部分论文与“设计方法”(30 篇论文)有关,或者是对新型城市空中交通飞机配置的设计和性能进行“概念研究”(41 篇论文)。 “飞行器技术”类别中有大量与“空气动力学”相关的论文(43 篇),同时也有许多论文专注于与电动垂直起降飞行器和其他分布式推进配置相关的空气动力学,即“螺旋桨、旋翼、风扇罩”(16 篇)和“空气动力相互作用”(7 篇)。

The large number of papers focused on novel propulsion—especially electric propulsion—and ways to design novel aircraft with new propulsion technology is not surprising. Indeed, electric propulsion is widely recognized as one of the underpinning technical enablers of aircraft capable of serving the UAM market. What is more surprising is that other key technical disciplines for enabling UAM aircraft such as “Autonomy” (11 papers), “Acoustics” (16 papers), and “Safety” (18 papers) have relatively few papers and are arguably underrepresented relative to their importance to the field. The relatively few papers focused on autonomy is likely a result of the very broad character of autonomy research, which focuses on technical fundamentals, as well as a myriad of application domains, including ground AVs. This breadth has likely resulted in few researchers focusing on autonomy research specifically for the emerging field of UAM. Additionally, aviation is a highly-regulated industry with inherent skepticism about the potential of autonomy for replacing pilots in the near future; this viewpoint has led to research in simplified vehicle operations (SVO) focused on enabling piloted aircraft with increasingly automated but not autonomous systems for flight control and navigation (Goodrich and Moore, 2015). The relatively few papers in the “Acoustics” category may result from the ramp-up of the research community to develop fundamentally new foundational computational tools and appropriate metrics for UAM aircraft noise, which differ substantively from traditional aviation noise metrics (Josephson, 2017). The few papers in the “Safety” category may be a result of the need to make initial research progress to address the novelty of UAM aircraft, which require envisioning entirely new paradigms for achieving safety. For example, the simple rotor systems in eVTOL aircraft do not typically offer the potential for “autorotation” for safe descents after an engine failure that is available to helicopters; instead eVTOL aircraft are designed with multiply-redundant powertrain components to prevent a complete propulsion failure in flight (Fredericks, 2016). The few papers in these categories may represent an opportunity for researchers to have impact by engaging in UAM research in these critically important research fields.
大量的论文关注新型推进技术,尤其是电动推进技术,以及设计新型飞机的方法,这并不令人意外。事实上,电动推进被广泛认为是能够为城市空中交通市场提供服务的飞机的基础技术之一。更令人惊讶的是,其他关键的技术学科,如“自主性”(11 篇论文)、“声学”(16 篇论文)和“安全性”(18 篇论文),相对较少,可以说在该领域的重要性相对低估。相对较少关注自主性的论文可能是自主性研究非常广泛的结果,它关注技术基础以及包括地面自动驾驶车辆在内的众多应用领域。这种广度可能导致很少的研究人员专门关注自主性研究,特别是针对新兴的城市空中交通领域。 此外,航空业是一个受到高度监管的行业,对于自主飞行取代飞行员的潜力存在固有的怀疑;这种观点导致了简化飞行器操作(SVO)的研究,重点是为了使有驾驶员的飞行器具备越来越自动化但不是完全自主的飞行控制和导航系统(Goodrich 和 Moore,2015)。在“声学”类别中相对较少的论文可能是由于研究界正在加紧开发基础性的全新计算工具和适用于城市空中交通(UAM)飞行器噪音的合适指标,这些指标与传统航空噪音指标有实质性的区别(Josephson,2017)。在“安全”类别中的少数论文可能是因为需要取得初步的研究进展来应对 UAM 飞行器的新颖性,这需要构想完全新的实现安全的范式。 例如,eVTOL 飞行器中的简单转子系统通常不具备直升机所具备的发动机故障后安全下降的“自转”能力;相反,eVTOL 飞行器采用了多重冗余的动力传动组件,以防止飞行中完全失去推进力(Fredericks,2016)。这些类别中的少数论文可能代表了研究人员在这些至关重要的研究领域参与城市空中交通研究的机会。

Within the “Market and Operations” category, the majority of papers had content related to “Air Traffic Management” (83 papers) and “Aviation Operations” (80 papers). The “Air Traffic Management” category includes papers focused on topics such as exploring paradigms for integrating large volumes of UAM air traffic within the existing NAS (Mueller et al., 2017, Thipphavong et al., 2018), constraints on UAM operations based on current operations at major airports (Vascik and Hansman, 2017, Vascik et al., 2018), and assessing and increasing airspace density and throughput for UAM operations (Goodrich and Barmore, 2018, Lowry, 2018). The “Aviation Operations” category includes papers with a focus on topics of economic and practical interest to UAM air carriers, including flight planning (Stouffer and Kostiuk, 2020) flight scheduling and dispatch (Roy et al., 2020, Shihab et al., 2019, Shihab et al., 2020), concepts of operations (Nneji et al., 2017; Kotwicz et al., 2019), and issues associated with electric aircraft recharging for flight operations (Hamilton and German, 2019, Shihab et al., 2020).
在“市场与运营”类别中,大部分论文内容与“空中交通管理”(83 篇论文)和“航空运营”(80 篇论文)相关。 “空中交通管理”类别包括关注将大量城市空中交通整合到现有国家空中交通系统中的范式探索(Mueller 等,2017 年;Thipphavong 等,2018 年),基于主要机场当前运营情况对城市空中交通运营的限制(Vascik 和 Hansman,2017 年;Vascik 等,2018 年),以及评估和提高城市空中交通运营的空域密度和吞吐量(Goodrich 和 Barmore,2018 年;Lowry,2018 年)。 “航空运营”类别包括关注对城市空中运营商具有经济和实际意义的主题的论文,包括飞行计划(Stouffer 和 Kostiuk,2020 年)、飞行调度和派遣(Roy 等,2020 年;Shihab 等,2019 年;Shihab 等,2020 年)、运营概念(Nneji 等,2017 年;Kotwicz 等,2019 年)以及与电动飞机充电飞行运营相关的问题(Hamilton 和 German,2019 年;Shihab 等,2020 年)。

A significant number of papers in the “Market and Operations” category was focused on transportation studies and research to assess the potential of UAM for providing an effective and scalable means of reducing travel time in cities, assessments which lend to understanding the market potential of UAM and its value to society. Papers in the “Air Transportation Studies” category (39 papers) assessed specific types of novel UAM aircraft in on-demand or scheduled service, typically through the lens of one or several operational case studies in example cities, and papers in the “Multimodal Transportation Studies” category (14 papers) assessed connections between UAM and other transportation modes such as cars or public transport or at least discussed differences between UAM and other transportation modes. Examples of applied transportation studies include a series of papers on “suburban air mobility” with electric short takeoff and landing (eSTOL) in the south Florida region (Wei et al., 2018, Robinson et al., 2018, Justin and Mavris, 2019, Somers et al., 2019). Papers in the “Demand” category (20 papers) focused on topics such as assessing the potential market size for UAM and other forms of on-demand air mobility based on census data and choice models (Kreimeier et al., 2018, Roy et al., 2020, Ploetner et al., 2020), stated preference surveys to assess UAM demand (Binder et al., 2018, Garrow et al., 2020, Fu et al., 2019), and agent-based demand simulation (Rothfeld et al., 2018, Fu et al., 2019, Ploetner et al., 2020). Finally, the “Infrastructure” category (12 papers) includes papers focused on optimization-based site selection of new vertiports (Daskilewicz et al., 2018) and STOLports (Wei et al., 2020) to serve the maximum demand, as well as papers that assess capacity constraints of vertiports (Vascik and Hansman, 2019, Maheshwari et al., 2020).
“市场与运营”类别中的大量论文集中在交通研究上,评估无人机空中交通(UAM)在提供有效和可扩展的减少城市出行时间方面的潜力,这些评估有助于了解 UAM 的市场潜力和对社会的价值。在“航空运输研究”类别(39 篇论文)中,通过在示范城市中进行一个或多个运营案例研究的视角,评估了特定类型的新型 UAM 飞行器在按需或定期服务中的表现,而“多模式交通研究”类别(14 篇论文)则评估了 UAM 与其他交通方式(如汽车或公共交通)之间的联系,或至少讨论了 UAM 与其他交通方式之间的差异。应用交通研究的例子包括一系列关于在佛罗里达南部地区使用电动短距起降(eSTOL)的“郊区空中移动”论文(Wei 等,2018 年;Robinson 等,2018 年;Justin 和 Mavris,2019 年;Somers 等,2019 年)。 “需求”类别的论文(20 篇)关注的主题包括基于人口普查数据和选择模型评估 UAM 和其他形式的按需空中移动的潜在市场规模(Kreimeier 等,2018 年,Roy 等,2020 年,Ploetner 等,2020 年),声明偏好调查评估 UAM 需求(Binder 等,2018 年,Garrow 等,2020 年,Fu 等,2019 年),以及基于代理人的需求模拟(Rothfeld 等,2018 年,Fu 等,2019 年,Ploetner 等,2020 年)。最后,“基础设施”类别的论文(12 篇)包括关注新垂直港口(Daskilewicz 等,2018 年)和 STOL 港口(Wei 等,2020 年)的基于优化的选址,以满足最大需求的论文,以及评估垂直港口容量限制的论文(Vascik 和 Hansman,2019 年,Maheshwari 等,2020 年)。

In our meta-analysis of author affiliations, the 251 UAM-related articles from AIAA consisted of a total of 862 listed authors, many of whom were listed on multiple papers, resulting in 554 unique authors. Among the 554 unique authors, 44 percent are affiliated with an academic institution, and 31 percent are associated with NASA. The remaining 25 percent of authors are associated with U.S.-based and international companies and research agencies. The majority of authors in the AIAA database (83 percent) are affiliated with institutions in the U.S., and the country with the second-highest representation (7 percent) is Germany. As these statistics reveal, the majority of UAM research has been conducted by the U.S., and NASA has played a critical role in this research.
在我们对作者机构的元分析中,AIAA 的 251 篇与 UAM 相关的文章共涉及 862 位列出的作者,其中许多人在多篇文章中都有名字,因此总共有 554 位独特的作者。在这 554 位独特的作者中,有 44%与学术机构有关联,31%与 NASA 有关联。剩下的 25%的作者与美国和国际公司以及研究机构有关联。AIAA 数据库中的大多数作者(83%)与美国的机构有关联,其次是德国(7%)。正如这些统计数据显示的那样,大部分 UAM 研究是由美国进行的,NASA 在这项研究中起到了关键作用。

A similar meta-analysis was conducted with UAM articles returned from the Scopus search with AIAA publications excluded. The 61 UAM-related articles from the Scopus search consisted of a total of 175 listed authors and 141 unique authors. Of all 141 unique authors, 66 percent are affiliated with an academic institution, and 16 percent are affiliated with NASA. The remaining 18 percent of authors are affiliated with U.S.-based and international companies and research agencies. Similar to the results seen in the AIAA database, the majority of authors (52 percent) are affiliated with institutions in the U.S., and the country with the second-most representation in the Scopus search is Germany (20 percent). The country with the third-most representation is the Republic of Korea (4 percent). These statistics confirm the trends seen in the AIAA search—UAM research has been concentrated primarily among U.S.- and German-based researchers, and NASA has played a critical role.
从 Scopus 搜索中排除 AIAA 出版物,对返回的 UAM 文章进行了类似的元分析。来自 Scopus 搜索的 61 篇与 UAM 相关的文章共有 175 位列出的作者和 141 位独立作者。在这 141 位独立作者中,66%隶属于学术机构,16%隶属于 NASA。其余 18%的作者隶属于美国和国际公司以及研究机构。与 AIAA 数据库中的结果类似,大多数作者(52%)隶属于美国的机构,Scopus 搜索中代表性第二高的国家是德国(20%)。代表性第三高的国家是韩国(4%)。这些统计数据证实了在 AIAA 搜索中观察到的趋势——UAM 研究主要集中在美国和德国的研究人员之间,并且 NASA 发挥了关键作用。

3.3. Ground transportation publications

To categorize the content of ground-transportation articles, we first identified broad topics. Many of these topics are overlapping and represent envisioned synergies across new technologies. For example, papers that discuss a future in which a fleet of AVs operate on batteries would be classified under the high-level categories of “Electric Vehicles” and “Autonomous Vehicles.” Once we identified broad topics, we tagged themes within each topic area that were potentially relevant for UAM research. The content tags are shown in Fig. 4. Later sections present our review of these lower-level tags in depth, so we restrict our discussion here to one key observation: within the top-tier transportation journals identified on Hensher’s list (2019), there were only four articles published on UAM. This highlights the opportunity for the transportation planning community to take a more active role in research related to the design and operations of UAM systems and apply insights they have gained through related research in the EV and AV fields to UAM.
为了对地面交通文章的内容进行分类,我们首先确定了广泛的主题。其中许多主题是重叠的,并代表了新技术之间的协同效应。例如,讨论未来一支由电池驱动的自动驾驶车队的论文将被归类为“电动车辆”和“自动驾驶车辆”这两个高级分类。一旦我们确定了广泛的主题,我们在每个主题领域内标记了可能与 UAM 研究相关的主题。内容标签如图 4 所示。后面的章节将详细介绍我们对这些低级标签的审查,因此我们在这里限制我们的讨论只针对一个关键观察:在 Hensher(2019)的列表中确定的顶级交通期刊中,只有四篇关于 UAM 的文章发表。这凸显了交通规划社区在 UAM 系统的设计和运营方面扮演更积极角色的机会,并将他们在电动车辆和自动驾驶领域的相关研究所获得的见解应用于 UAM。

  1. Download : Download high-res image (214KB)
  2. Download : Download full-size image

Fig. 4. Themes from transportation journals’ table of contents search, January 2015 to June 2020.
图 4. 从交通期刊的目录搜索中得出的主题,时间范围为 2015 年 1 月至 2020 年 6 月。

The 478 ground transportation articles from the journals Transportation Research Part A (TR–A), TR–B, TR–C, TR–D, Transportation Science, and Transportation consisted of a total of 1,594 listed authors, many of whom were listed on multiple papers, resulting in 1,154 unique authors. Among the 1,154 unique authors, 84 percent are affiliated with an academic institution. The remaining 16 percent of authors are associated with U.S.-based and international companies and research agencies. Among authors associated with academic institutions, 26 percent are affiliated with institutions in the U.S. The country with the second-highest representation (13 percent) in the ground transportation journal database is China, closely followed by Germany (9 percent). As these statistics reveal, the majority of ground transportation research has been conducted by the U.S., but the authors are much more diverse in their affiliated countries than the UAM authors. Ground transportation authors are also much more commonly affiliated with academic institutions compared to UAM authors.
来自《交通研究 A 部分》(TR-A)、《交通研究 B 部分》(TR-B)、《交通研究 C 部分》(TR-C)、《交通研究 D 部分》(TR-D)、《交通科学》和《交通》期刊的 478 篇地面交通文章共涉及 1594 位作者,其中许多作者在多篇文章中都有署名,因此共有 1154 位独特的作者。在这 1154 位独特的作者中,84%与学术机构有关联。剩下的 16%的作者与美国和国际公司以及研究机构有关联。在与学术机构有关联的作者中,有 26%与美国的机构有关联。在地面交通期刊数据库中,第二大代表国家(占 13%)是中国,紧随其后的是德国(占 9%)。正如这些统计数据所显示的,大部分地面交通研究是由美国进行的,但作者的所属国家比 UAM 作者更多样化。与 UAM 作者相比,地面交通作者更常常与学术机构有关联。

4. Demand modeling 需求建模

To date, the UAM and ground transportation communities have taken different approaches with respect to modeling demand. The UAM community is currently focused on conducting high-level assessments to understand if there are viable markets for UAM and how mission requirements for these markets (which tie directly to aircraft design specifications) vary across different cities. Identifying where UAM could offer door-to-door travel time savings compared to other modes is a key part of these high-level assessments. To this end, macro-level data of economic activity, aggregate data of commuter flows, and census and other government data are often used to estimate UAM market demand.
迄今为止,城市空中交通(UAM)和地面交通领域在需求建模方面采取了不同的方法。目前,UAM 社区的重点是进行高层次评估,以了解 UAM 是否存在可行的市场,以及这些市场的任务需求(直接与飞机设计规格相关)在不同城市之间的差异。确定 UAM 相对于其他交通方式能够提供门到门的时间节约是这些高层次评估的关键部分。为此,通常使用宏观经济活动数据、通勤流量的综合数据以及人口普查和其他政府数据来估计 UAM 市场需求。

In contrast, the ground transportation community often conducts surveys to predict how individuals will respond to different operational, pricing, and policy measures. These surveys enable researchers to understand how opinions and intentions to adopt a new technology vary as a function of socioeconomic and sociodemographic (SED) characteristics, as well as attitudes and perceptions (e.g., is the individual tech-savvy?). Insights from these surveys can be helpful for identifying potential early adopters and designing marketing campaigns. Surveys also allow researchers to focus on specific questions, such as the willingness to travel with strangers in ride-hailing situations or the value of times across different modes as a function of trip purpose. These and other questions will be relevant to the UAM community as they start conducting detailed assessments of which particular consumers will use UAM and how much they are willing to pay.
相比之下,地面交通界经常进行调查,以预测个人对不同运营、定价和政策措施的反应。这些调查使研究人员能够了解意见和采用新技术的意愿如何随着社会经济和社会人口特征(SED)的变化而变化,以及态度和观念(例如,个人是否精通技术)。这些调查的见解对于确定潜在的早期采用者和设计营销活动非常有帮助。调查还使研究人员能够专注于特定问题,例如愿意在打车情况下与陌生人一起旅行的意愿,或者根据出行目的的不同模式下时间的价值。这些和其他问题将对 UAM 社区产生影响,因为他们开始进行详细评估,确定哪些特定消费者将使用 UAM 以及他们愿意支付多少费用。

This section provides an overview of demand studies from the UAM and EV/AV literatures, and summarizes key insights from the EV/AV literatures that can help inform future UAM research.
本节概述了来自 UAM 和 EV/AV 文献的需求研究,并总结了来自 EV/AV 文献的关键见解,这些见解可以帮助指导未来的 UAM 研究。

4.1. Review of UAM demand studies
4.1. UAM 需求研究的回顾

This section reviews three types of demand studies that have been conducted by UAM researchers: global market studies that have ranked cities worldwide for their potential to offer UAM services, studies that have compared potential travel times savings with UAM against other modes, and survey-based research.
本节回顾了由 UAM 研究人员进行的三种需求研究类型:全球市场研究,对全球城市进行排名以评估其提供 UAM 服务的潜力;将 UAM 与其他交通方式进行比较,评估潜在的出行时间节约;以及基于调查的研究。

4.1.1. Global market studies

Multiple studies, including those of Becker et al., 2018, Robinson et al., 2018, Booz Allen Hamilton (2018), KPMG (Mayor and Anderson, 2019), and NEXA Advisors (2019), have examined the potential for on-demand mobility for UAM across different cities. These studies used various qualitative and quantitative methodologies to measure the demand potential. For example, Becker et al. (2018) used a gravity model to forecast interurban air passenger demand for 2042 based on socioeconomic factors and generated a list of potential UAM markets. NEXA Advisors (2019) modeled demand for UAM for different use cases including an airport shuttle, a corporate campus shuttle, an on-demand air taxi service, medical and emergency operations and service, and regional air transport service. Various inputs were used including population and density, gross domestic product (GDP) per capita, age distribution, current commercial and business aviation activity, and presence of Fortune 1000 companies (NEXA Advisors, 2019).
包括 Becker 等人(2018 年)、Robinson 等人(2018 年)、Booz Allen Hamilton(2018 年)、KPMG(Mayor 和 Anderson,2019 年)和 NEXA Advisors(2019 年)在内的多项研究已经对不同城市的按需出行移动性(UAM)的潜力进行了研究。这些研究使用了各种定性和定量方法来衡量需求潜力。例如,Becker 等人(2018 年)使用重力模型预测了基于社会经济因素的 2042 年城际航空客运需求,并生成了潜在的 UAM 市场列表。NEXA Advisors(2019 年)对不同用例的 UAM 需求进行了建模,包括机场班车、企业园区班车、按需空中出租车服务、医疗和紧急救援服务以及区域航空运输服务。使用了各种输入,包括人口和密度、人均国内生产总值(GDP)、年龄分布、当前商业和商务航空活动以及财富 1000 强公司的存在(NEXA Advisors,2019 年)。

KPMG modeled UAM demand using inputs that included city GDP and GDP growth, city population and population growth, city population density, city change in income distribution through 2050, wealth concentration, and information about existing ground services (Mayor and Anderson, 2019). Mayakonda et al. (2020) estimated the UAM share of total passenger kilometers traveled for the cities identified in the KPMG report as a function of UAM ticket cost, travel time savings, and vertiport density. For UAM service offered at $1.50/km, they found UAM shares of 0.18–0.4 percent across different vertiport densities, but for UAM service offered at $0.30/km, these shares increased to 3.2–8.5 percent.
毕马威公司使用了包括城市 GDP 和 GDP 增长、城市人口和人口增长、城市人口密度、城市收入分配变化、财富集中程度以及现有地面服务信息在内的数据来模拟城市对城市空中交通需求(Mayor 和 Anderson,2019)。Mayakonda 等人(2020)根据毕马威报告中确定的城市,估计了城市空中交通在总乘客公里数中的占比,该占比是 UAM 票价、旅行时间节省和垂直起降机场密度的函数。对于每公里 1.50 美元的 UAM 服务,他们发现在不同的垂直起降机场密度下,UAM 的占比为 0.18-0.4%,但对于每公里 0.30 美元的 UAM 服务,这些占比增加到 3.2-8.5%。

The study by Booz Allen Hamilton selected 10 cities from a possible pool of 40 cities based on population and population density as case studies for a UAM analysis. The final 10 cities were selected based on qualitative criteria including ground transportation congestion, weather, and existing infrastructure and ground transportation patterns (Booz Allen Hamilton, 2018, Reiche et al., 2018).
博思艾伦汉密尔顿公司的研究选择了 10 个城市作为城市空中交通(UAM)分析的案例研究,这些城市是从 40 个可能的城市中根据人口和人口密度进行选择的。最终选择这 10 个城市是基于定性标准,包括地面交通拥堵、天气和现有基础设施以及地面交通模式(Booz Allen Hamilton,2018 年,Reiche 等,2018 年)。

Robinson et al. (2018) identified potential cities for UAM based on qualitative criteria such as the city’s level of sprawl, density, presence of water bodies (that could be used to construct barges for potential vertiports), number of airports currently in the city, population wealth, presence of high-tech industries, ground transportation congestion, ground transportation patterns, and weather. The U.S. cities identified as potential candidates for UAM for the reports discussed above are summarized in Table 2 and show that there is a large degree of overlap in the candidates identified in the previous literature, with the NEXA Advisors including more cities in their analysis, and thus having more smaller cities than the other studies.
Robinson 等人(2018 年)根据城市的扩张程度、密度、水域的存在(可用于建造可能的垂直港口的驳船)、城市现有机场数量、人口财富、高科技产业的存在、地面交通拥堵、地面交通模式和天气等定性标准,确定了无人机空中交通(UAM)的潜在城市。美国被认为是 UAM 潜在候选城市的报告如表 2 所示,表明在先前的文献中已经确定了许多重叠的候选城市,而 NEXA Advisors 在其分析中包括了更多的城市,因此比其他研究涵盖了更多的小城市。

Table 2. U.S. cities identified in the literature as having potential for UAM service.
表 2. 文献中被认为具有无人机空中出租车服务潜力的美国城市。

Empty CellStudies 研究Rank among U.S. cities in KPMG report in 2040
2040 年 KPMG 报告中美国城市的排名
Atlanta 亚特兰大Robinson, KPMG, NEXA 罗宾逊,毕马威,NEXA8
Baltimore 巴尔的摩NEXA
Boston 波士顿Robinson, KPMG, NEXA 罗宾逊,毕马威,NEXA12
Chicago 芝加哥Robinson, KPMG, NEXA 罗宾逊,毕马威,NEXA3
Dallas–Ft. Worth 达拉斯-沃思堡Robinson, BAH, KPMG, NEXA
Denver 丹佛BAH, NEXA 巴,内克萨
Detroit 底特律NEXA
Honolulu 檀香山BAH
Houston 休斯顿Robinson, BAH, KPMG, NEXA
Los Angeles 洛杉矶BAH, KPMG, NEXA BAH,毕马威,NEXA2
Miami 迈阿密Robinson, BAH, KPMG, NEXA
New York City 纽约市Robinson, BAH, KPMG, NEXA
Philadelphia 费城KPMG, NEXA 毕马威,NEXA9
Phoenix 凤凰BAH, KPMG, NEXA BAH,毕马威,NEXA11
San Diego 圣地亚哥NEXA
Seattle 西雅图NEXA
Silicon Valley 硅谷Robinson, BAH, KPMG, NEXA
Washington, D.C. 华盛顿特区BAH, KPMG, NEXA BAH,毕马威,NEXA10

Note: Additional cities included by NEXA are San Jose, Charlotte, Tampa, Nashville, Las Vegas, Salt Lake City, Raleigh–Durham–Chapel Hill, and Syracuse.
注意:NEXA 还包括的其他城市有圣何塞、夏洛特、坦帕、纳什维尔、拉斯维加斯、盐湖城、罗利-达勒姆-教堂山和锡拉丘兹。

Reference Key: Robinson = Robinson et al. (2018); BAH = Booz Allen Hamilton (2018) and Reiche et al. (2018); KPMG = Mayor and Anderson (2019); NEXA = NEXA Advisors (2019).
参考关键字:罗宾逊 = 罗宾逊等人(2018 年);BAH = Booz Allen Hamilton(2018 年)和 Reiche 等人(2018 年);KPMG = Mayor 和 Anderson(2019 年);NEXA = NEXA 顾问(2019 年)。

4.1.2. Door-to-Door travel time studies across modes
4.1.2. 跨模式的门到门旅行时间研究

Numerous studies have compared door-to-door travel times between UAM and conventional modes and examined the sensitivity of these door-to-door times to different parameters, such as access and egress times, and aircraft cruise speeds. Wei et al. (2018) conducted a door-to-door travel time comparison between personal cars and short takeoff and landing (STOL) aircraft. They found potential demand for STOL operations that have cruising speeds of 160 knots for individuals who have commutes in excess of 45 min based on a case study of the South Florida region, which includes Miami. As range decreases, access and egress times to and from the port become increasingly important (and the travel times savings associated with the air taxi decrease compared to auto) (Wei et al., 2018). Roland Berger (2018) found that air taxi trips need to be at least 15 to 25 km (about 9 to 16 miles) to provide travel time savings over existing modes.
许多研究比较了城市空中交通(UAM)和传统交通方式之间的门到门旅行时间,并研究了这些门到门时间对不同参数的敏感性,如进出时间和飞机巡航速度。魏等人(2018)对个人汽车和短距起降(STOL)飞机进行了门到门旅行时间比较。他们发现,在南佛罗里达地区(包括迈阿密)的案例研究中,对于通勤时间超过 45 分钟的个人来说,具有 160 节的巡航速度的 STOL 运营存在潜在需求。随着航程的减少,进出港口的时间变得越来越重要(与汽车相比,乘坐空中出租车的旅行时间节省减少)(魏等人,2018)。罗兰贝格(2018)发现,空中出租车的行程至少需要 15 到 25 公里(约 9 到 16 英里),才能比现有交通方式节省旅行时间。

Swadesir and Bil (2019) compare travel times, costs, and general convenience of using an air taxi service, bike, auto, and public transport for Melbourne, Australia. Consistent with the results from the South Florida study, Swadesir and Bil (2019) found that demand for an air taxi service is sensitive to access and egress times to the vertiport, as well as the times to board and disembark the aircraft. Based on an analysis of UAM service in Sioux Falls, South Dakota, USA, Rothfeld et al. (2018) found that UAM processing times have a larger influence on UAM adoption than the UAM vehicle cruising speed and that “the current focus on UAM vehicle capacity and speeds should be extended with UAM accessibility and shorter processing times.”
Swadesir 和 Bil(2019)比较了使用空中出租车服务、自行车、汽车和公共交通工具在澳大利亚墨尔本的旅行时间、费用和便利性。与南佛罗里达州的研究结果一致,Swadesir 和 Bil(2019)发现,对空中出租车服务的需求对于进出垂直起降机场的时间以及登机和下机时间非常敏感。根据对美国南达科他州苏福尔斯市的 UAM 服务的分析,Rothfeld 等人(2018)发现,UAM 处理时间对 UAM 的采用影响更大,而不是 UAM 车辆的巡航速度,并且“目前对 UAM 车辆容量和速度的关注应该扩展到 UAM 的可访问性和更短的处理时间。”

Antcliff, Moore, and Goodrich (2016) compared travel times for urban and suburban commutes between ground and air taxis in the Silicon Valley and found travel times that were three to six times lower for some commutes in the area. They also found that a key factor in improving door-to-door travel times for air taxis is to minimize preboarding times (e.g., waiting times and security clearance times) as well as the times to board and disembark the aircraft. Kreimeier, Strumpf, and Gottschalk (2016) assessed the viability of a UAM service in Germany for intercity travel and found that UAM market shares are highly sensitive to UAM prices as well as access and egress times.
Antcliff, Moore, and Goodrich (2016)在硅谷比较了城市和郊区之间地面出租车和空中出租车的通勤时间,发现该地区某些通勤时间比原来低三到六倍。他们还发现,改善空中出租车门到门旅行时间的关键因素是尽量减少登机前的时间(例如等待时间和安全检查时间),以及登机和下机的时间。Kreimeier, Strumpf, and Gottschalk (2016)评估了德国城际旅行的 UAM 服务的可行性,发现 UAM 市场份额对 UAM 价格以及进出时间非常敏感。

Other studies that have compared costs and travel times across modes include those by Roy et al., 2018b, Akhter et al., 2020; and Vascik, Hansman, and Dunn (2018). The latter compare door-to-door travel times for 32 reference missions in the Boston, Dallas, and Los Angeles areas to identify operational constraints. Their analysis focused on high-income commuter neighborhoods, which they defined as those with annual household incomes of at least $200K or as neighborhoods with average home valuations of at least $1M in Los Angeles and Dallas and at least $900K in Boston (Vascik, Hansman, and Dunn, 2018).
其他比较成本和旅行时间的研究包括 Roy 等人的 2018b 年研究,Akhter 等人的 2020 年研究以及 Vascik、Hansman 和 Dunn 的 2018 年研究。后者比较了波士顿、达拉斯和洛杉矶地区 32 个参考任务的门到门旅行时间,以确定运营限制。他们的分析重点放在高收入通勤社区上,他们将其定义为洛杉矶和达拉斯的年度家庭收入至少为 20 万美元或波士顿至少为 90 万美元的社区(Vascik、Hansman 和 Dunn,2018 年)。

4.1.3. Survey-Based UAM demand studies

Several survey-based studies of UAM have been conducted by consulting firms, aircraft manufacturers, and academics. For example, Booz Allen Hamilton (2018) conducted a survey that explored the potential for intercity and intracity UAM service. They sampled approximately 300 individuals in each of the following five cities: Houston, San Francisco, Los Angeles, New York City, and Washington, D.C. Airbus conducted a survey of 1540 individuals that compared public perceptions of UAM service among residents of Los Angeles, Mexico City, New Zealand, and Switzerland. Overall, 45 percent of respondents’ initial reactions to UAM were positive, with 42 percent believing UAM was safe or very safe. Interest in UAM was higher among males, those with higher educational attainment levels, those who frequently use ride-hailing services or public transportation, and those from Los Angeles or Mexico City (Yedavalli and Mooberry, n.d.). Deloitte conducted a survey of approximately 10,000 individuals representing the regions of the U.S., Canada, the U.K., France, China, Japan, and Australia and similarly found that nearly half of the respondents viewed autonomous UAM vehicles as a potentially viable solution to roadway congestion, but 80 percent had safety concerns (Lineberger, Hussain, and Rutgers, 2019).
咨询公司、飞机制造商和学者进行了几项基于调查的城市空中交通(UAM)研究。例如,Booz Allen Hamilton(2018)进行了一项调查,探讨了城际和城市 UAM 服务的潜力。他们在休斯顿、旧金山、洛杉矶、纽约市和华盛顿特区的每个城市中对大约 300 名个体进行了抽样调查。空中客车公司对 1540 名个体进行了一项调查,比较了洛杉矶、墨西哥城、新西兰和瑞士居民对 UAM 服务的公众看法。总体而言,45%的受访者对 UAM 的初步反应是积极的,42%的人认为 UAM 是安全的或非常安全的。对 UAM 的兴趣在男性、教育程度较高的人、经常使用打车服务或公共交通工具的人以及来自洛杉矶或墨西哥城的人中更高(Yedavalli 和 Mooberry,无日期)。德勤公司对代表美国、加拿大和英国地区的约 10,000 名个体进行了一项调查。法国、中国、日本和澳大利亚的调查结果显示,近一半的受访者认为自动驾驶的城市空中交通工具是缓解道路拥堵的潜在解决方案,但 80%的人对安全性表示担忧(Lineberger, Hussain, and Rutgers, 2019)。

On the academic side, Fu, Rothfeld, and Antoniou (2019) modeled the choice among private car, public transportation, autonomous ground taxi, and autonomous air taxi using multinomial logit, nested logit, and mixed logit models based on a stated preference survey of 248 respondents from the Munich metropolitan area. Two trip purposes were considered and combined into a single estimation dataset: daily commuting and a non-commuting private trip. The authors estimated values of times for these four modes as 27.55, 27.47, 32.57, and 44.68 €/hour respectively, which correspond to11 33.89, 33.79, 40.06 and 54.96 USD/hour, respectively.
在学术方面,Fu、Rothfeld 和 Antoniou(2019)使用多项 Logit、嵌套 Logit 和混合 Logit 模型对慕尼黑都市区 248 名受访者进行了一项偏好调查,模拟了私家车、公共交通、自动驾驶地面出租车和自动驾驶空中出租车之间的选择。考虑了两种出行目的,并将其合并为一个估计数据集:日常通勤和非通勤私人出行。作者估计了这四种模式的时间价值分别为 27.55、27.47、32.57 和 44.68 欧元/小时,相当于 33.89、33.79、40.06 和 54.96 美元/小时。

Based on a survey conducted by Uber of 2607 residents from Dallas–Ft. Worth (DFW) and Los Angeles (many of whom were drawn from the Uber customer database), Song, Hess, and Decker (2019) estimated a latent class model and found values of time ranging from $11.15 to $36.78 for different travel time components. On average, across two latent classes the access time, egress time, flight time, and in-vehicle travel time in $/hour were found to be 26.03, 34.43, 20.75, and 13.94, respectively.
根据 Uber 对达拉斯-沃思堡(DFW)和洛杉矶的 2607 名居民(其中许多人来自 Uber 客户数据库)进行的调查,Song、Hess 和 Decker(2019)估计了一个潜在类别模型,并发现不同出行时间组成部分的时间价值在 11.15 美元至 36.78 美元之间。平均而言,两个潜在类别中的接入时间、离开时间、飞行时间和车内旅行时间的价值分别为 26.03 美元/小时、34.43 美元/小时、20.75 美元/小时和 13.94 美元/小时。

Binder et al. (2018) and Garrow et al. (2019) conducted two surveys of high-income commuters residing in Atlanta, Boston, DFW, San Francisco, and Los Angeles. The first survey, which contained 2499 responses, examined competition with current modes, and the second survey, which contained 1405 responses, was expanded to include competition with autonomous ground vehicles. Results from the first survey showed that individuals who were male, tech-savvy, and frequent users of ride-hailing services were more likely to take an air taxi for commuting. (Boddupalli, Garrow, and German, 2020). Results from the second survey were consistent with the first survey in that males, tech-savvy, and frequent users of ride-hailing services were more likely to take an air taxi. In addition, those who had positive attitudes toward collective modes (i.e., transit, ride-hailing, etc.) and those who felt time pressured were more likely to take an air taxi (Garrow, Roy, and Newman, 2020). In both surveys, the authors found significant heterogeneity in individuals’ value of time (VOT), and in the second survey that included AVs, the authors found that compared to the VOT for a conventional auto, the median VOTs for an AV and air taxi were 15 percent lower and 9 percent higher, respectively.
Binder 等人(2018 年)和 Garrow 等人(2019 年)对居住在亚特兰大、波士顿、DFW、旧金山和洛杉矶的高收入通勤者进行了两次调查。第一次调查共收到 2499 份回复,调查了与现有交通方式的竞争情况;第二次调查共收到 1405 份回复,扩大了范围,包括与自动驾驶地面车辆的竞争。第一次调查的结果显示,男性、科技精通和经常使用打车服务的人更有可能乘坐空中出租车通勤(Boddupalli、Garrow 和 German,2020 年)。第二次调查的结果与第一次调查一致,即男性、科技精通和经常使用打车服务的人更有可能乘坐空中出租车。此外,那些对集体交通方式(如公共交通、打车等)持积极态度和感到时间紧迫的人更有可能乘坐空中出租车(Garrow、Roy 和 Newman,2020 年)。 在两项调查中,作者发现个体的时间价值(VOT)存在显著的异质性。在第二项调查中,包括自动驾驶汽车(AVs)的情况下,与传统汽车的 VOT 相比,自动驾驶汽车和空中出租车的中位数 VOT 分别降低了 15%和增加了 9%。

Based on a survey of 221 individuals, the majority of which resided in Europe, Al Haddad et al. (2020) estimated multinomial and ordered logit models, where the ordering corresponds to the time of adoption, and interpreted the results in the context of the Technology Acceptance Model. Among the 221 respondents, 22 percent stated they would adopt UAM in the first year, 37 percent in the second or third year of implementation, 14 percent during the fourth and fifth year, and 3 percent during the sixth year; 3 percent stated they would never adopt the service and 22 percent indicated they were unsure on their adoption time horizon of UAM. Based on a survey of 4700 individuals conducted in 2019, Ljungholm and Olah (2020) found that 14 percent of respondents would be “ready and comfortable to ride in a flying taxi” right now, 18 percent within the next year, 21 percent within the next five years, 20 percent in more than 10 years’ time, and 14 percent would never be comfortable. While the timing of adoption across these studies varies, what is clear is that UAM adoption by all consumers will not be instantaneous.
根据对 221 名个体的调查,其中大多数居住在欧洲,Al Haddad 等人(2020 年)估计了多项式和有序逻辑模型,其中排序对应于采用的时间,并在技术接受模型的背景下解释了结果。在 221 名受访者中,22%表示他们将在第一年采用 UAM,37%在实施的第二或第三年采用,14%在第四和第五年采用,3%在第六年采用; 3%表示他们永远不会采用该服务,22%表示他们对 UAM 的采用时间尚不确定。根据 2019 年对 4700 名个体进行的调查,Ljungholm 和 Olah(2020 年)发现,14%的受访者“现在已经准备好并愿意乘坐飞行出租车”,18%的受访者在未来一年内,21%的受访者在未来五年内,20%的受访者在十年以上的时间内,14%的受访者永远不会感到舒适。尽管这些研究中的采用时间有所不同,但明确的是,所有消费者对 UAM 的采用不会是瞬间的。

Finally, Han, Yu, and Kim (2019) examined customers’ decision-making processes for adopting electric airplanes for traditional commercial flights. Based on a survey of 321 airline customers in the U.S. who had used an airline for traveling within the last year, they found that reducing consumers’ perceived risk and increasing new product knowledge was critical to increasing trust and positive attitudes toward electric airplanes and their willingness to pay.
最后,Han、Yu 和 Kim(2019 年)研究了顾客在传统商业航班中采用电动飞机的决策过程。根据对美国 321 名航空公司顾客的调查,这些顾客在过去一年内曾使用航空公司进行旅行,他们发现降低消费者的感知风险和增加新产品知识对于增加对电动飞机的信任和积极态度以及支付意愿至关重要。

4.2. Review of EV and AV demand studies
4.2. 电动车和自动驾驶需求研究的回顾

Within the ground transportation literature, there have been more than 200 studies over the past five years that have focused on demand. Some of these studies focus on understanding how demand for EV, AV, carsharing, and/or ride-hailing services varies as a function of sociodemographic and socioeconomic (SED) characteristics, as well as different attitudes, beliefs, and personality factors. About 50 studies within the ground transportation literature have focused on how adoption of new EV and AV technologies will increase over time. These studies include an assessment of barriers to adoption, analysis of the differences between early adopters and late adopters, and extension and application of different theoretical frameworks used to predict the timing of adoption across a population. Finally, there are two topics that have been explored in the ground transportation literature that are particularly relevant for UAM: studies that have examined individuals’ value of time, defined as the amount of money individuals are willing to spend for travel time savings, and studies that have examined individuals’ willingness to ride in vehicles with individuals they know or strangers. This section reviews these demand-related topics in depth.
在地面交通文献中,过去五年已经有 200 多项研究集中在需求方面。其中一些研究着重于理解电动汽车(EV)、自动驾驶汽车(AV)、共享汽车和/或打车服务的需求如何随社会人口特征、经济社会特征以及不同的态度、信念和个性因素而变化。地面交通文献中约有 50 项研究集中在新的电动汽车和自动驾驶技术的采用将如何随时间增加。这些研究包括评估采用障碍、分析早期采用者和晚期采用者之间的差异,以及扩展和应用用于预测整个人群采用时间的不同理论框架。最后,地面交通文献中探讨的两个与城市空中交通(UAM)特别相关的主题是:研究个人时间价值,即个人愿意为节省旅行时间而花费的金额;以及研究个人愿意与认识的人或陌生人一起乘坐车辆的意愿。 本节深入讨论了这些与需求相关的主题。

4.2.1. SED characteristics and segmentation
4.2.1. SED 特性和分割

Within the ground transportation literature, there are many publications that focus on understanding how SED characteristics influence transportation choices. Segmentation studies that compare the travel behavior of different populations defined by sociodemographic, socioeconomic, and/or geographic characteristics are common. Examining how travel behavior varies across different populations is important from a public policy perspective, as it helps better target limited resources to meet demand, helps ensure that the travel needs of mobility-restricted individuals are met, and helps ensure that policies are equitable across different population segments and geographical areas.
在地面交通文献中,有许多出版物专注于理解 SED 特征如何影响交通选择。通过比较不同人口群体的旅行行为,根据社会人口学、经济学和/或地理特征进行分割研究是常见的。从公共政策的角度来看,研究不同人群的旅行行为变化非常重要,因为它有助于更好地定位有限资源以满足需求,确保满足行动受限个体的出行需求,并确保政策在不同人口群体和地理区域之间是公平的。

Here, we loosely use the term segmentation to identify studies that examined how consumer preferences for AVs, EVs, and/or sharing programs vary across demographic, socioeconomic and/or geographic segments. Several studies have found that interest in AVs, EVs, and sharing technologies is associated more with individuals who are younger, more educated, have higher incomes, and are male (e.g., see Dong et al., 2019b, Hudson et al., 2019, Kopp et al., 2015, Liu et al., 2019b, Potoglou et al., 2020, Shabanpour et al., 2018a, Spurlock et al., 2019, Vij et al., 2020, Wang and Zhao, 2019).
在这里,我们宽泛地使用“分割”一词来识别研究,这些研究考察了消费者对自动驾驶汽车(AVs)、电动汽车(EVs)和/或共享计划的偏好在人口统计、社会经济和/或地理分割上的变化。一些研究发现,对 AVs、EVs 和共享技术的兴趣更多地与年轻人、受教育程度较高、收入较高和男性相关(例如,参见 Dong 等,2019b;Hudson 等,2019;Kopp 等,2015;Liu 等,2019b;Potoglou 等,2020;Shabanpour 等,2018a;Spurlock 等,2019;Vij 等,2020;Wang 和 Zhao,2019)。

There are subtle differences across studies related to gender and income. For example, while many studies have found that women are more risk-adverse and less likely to adopt AVs (Wang and Zhao, 2019, Kaltenhäuser et al., 2020), Spurlock et al. (2019) found that women are less likely to adopt new transportation technologies except for ride-hailing. While Young and Farber (2019) found that ride-hailing is generally a wealthier, younger phenomena, Spurlock et al. (2019) found that higher-income individuals are disproportionately represented among current adopters of new ground vehicle technologies and that low- to middle-income individuals are just as likely to have adopted pooled ride-hailing. Kim (2015) found that carsharing in low-income and high-income areas of NYC were similar.
有关性别和收入的研究存在细微差异。例如,虽然许多研究发现女性更加风险规避,不太可能采用自动驾驶汽车(Wang 和 Zhao,2019 年,Kaltenhäuser 等,2020 年),但 Spurlock 等人(2019 年)发现女性除了叫车之外,不太可能采用新的交通技术。虽然 Young 和 Farber(2019 年)发现叫车通常是富裕、年轻人的现象,但 Spurlock 等人(2019 年)发现高收入个体在当前新型地面车辆技术的采用者中占比较大,而低至中等收入个体采用共享叫车的可能性与之相当。Kim(2015 年)发现纽约市低收入区和高收入区的共享汽车使用情况相似。

Several studies focused on the travel behaviors of those with disabilities or older adults, who are two populations for which AVs and sharing services could potentially help increase mobility. For example, Bennett, Vijaygopal, and Kottasz (2019) investigated how attitudes toward AVs differ for those with physical disabilities and those without physical disabilities in the U.K. Harper et al. (2016) used the U.S. National Household Travel Survey to predict potential trip increases for older adults and individuals with travel-related medical conditions. Faber and van Lierop (2020) examined preferences for AVs among older adults in Utrecht, the Netherlands. Other studies focused on better understanding particular (and often narrow) market segments. For example, Lee and Mirman (2018) explored parents’ perspectives on using AVs to transport their children. Ghasri, Ardeshiri, and Rashidi (2019) compared how perceptions toward EVs vary among younger adults, i.e., Gen X, Gen Y (Millennials) and Gen Z, in New South Wales, Australia, and found that the Millennials showed interest in adopting EVs. Alemi et al. (2019) used a survey of Millennials from California and found that those who frequently use smartphone apps to manage other aspects of their travel (e.g., checking traffic) or who frequently travel by plane for leisure purposes were more likely to use a ride-hailing service.
几项研究关注残疾人或老年人的出行行为,这两个群体是自动驾驶汽车和共享服务可能有助于增加流动性的人群。例如,Bennett、Vijaygopal 和 Kottasz(2019)调查了英国残疾人和非残疾人对自动驾驶汽车的态度差异。Harper 等人(2016)利用美国全国家庭出行调查预测了老年人和有出行相关医疗条件的个体的潜在出行增加。Faber 和 van Lierop(2020)研究了荷兰乌得勒支老年人对自动驾驶汽车的偏好。其他研究则着重于更好地了解特定(通常是狭窄的)市场细分。例如,Lee 和 Mirman(2018)探讨了家长们使用自动驾驶汽车接送孩子的观点。Ghasri、Ardeshiri 和 Rashidi(2019)比较了澳大利亚新南威尔士州年轻人(即 X 世代、Y 世代(千禧一代)和 Z 世代)对电动汽车的感知差异,并发现千禧一代对采用电动汽车表现出兴趣。Alemi 等人。 (2019 年)对加利福尼亚州的千禧一代进行了一项调查,发现那些经常使用智能手机应用程序来管理旅行其他方面(例如,查看交通情况)或经常以休闲为目的乘飞机旅行的人更有可能使用打车服务。

Several studies investigated geographic differences. Huang and Qian (2018) explored how preferences for EVs differ across cities with different population sizes in China. They found that consumers in smaller cities are more sensitive to EV purchase price and subsidies. Illgen and Höck (2018) explored the potential for carsharing services in rural regions of Switzerland, and Rotaris and Danielis (2018) explored the potential for ride-hailing services in the Friuli Venezia Giulia region, Italy, “a region characterized by small-sized towns and less-densely populated rural areas.” Based on a comparison of individuals in Germany, India, Japan, Sweden, the U.K., and the U.S., Potoglou et al. (2020) found that Japanese consumers are generally willing to pay for AVs, whereas European consumers need to be compensated for automation. Finally, Liu, Khattak, et al. (2019) used the U.S. National Household Travel survey to investigate geographic differences in the ownership of alternative-fueled vehicles and found higher ownership rates among high-income households in states in the southeast or northwest, and higher ownership rates among seniors in states in the northeast and northwest.
几项研究调查了地理差异。黄和钱(2018 年)探讨了中国不同人口规模城市之间对电动汽车的偏好差异。他们发现,较小城市的消费者对电动汽车的购买价格和补贴更为敏感。Illgen 和 Höck(2018 年)探讨了瑞士农村地区共享汽车服务的潜力,而 Rotaris 和 Danielis(2018 年)则探讨了意大利弗留利威尼斯朱利亚地区的打车服务潜力,该地区以小型城镇和人口稀少的农村地区为特点。Potoglou 等人(2020 年)通过比较德国、印度、日本、瑞典、英国和美国的个体发现,日本消费者普遍愿意为自动驾驶汽车付费,而欧洲消费者需要得到补偿才能接受自动化。最后,刘、Khattak 等人(2019 年)利用美国全国家庭出行调查研究了替代燃料车辆的拥有情况的地理差异,并发现东南部或西北部的高收入家庭拥有率较高,而东北部和西北部的老年人拥有率较高。

4.2.2. Attitudes, Beliefs, and personality factors

Within the ground transportation literature, a wide body of literature focused on understanding how individuals’ attitudes, beliefs, personality, and similar factors influence travel behavior choices, including the adoption of new technologies. Table 3 in the appendix summarizes 19 ground transportation studies that have examined the influence of individuals’ attitudes toward EVs, AVs, or sharing programs. Sixteen of the studies used surveys, two used interviews, and one reviewed the literature. The studies were conducted across a range of nations including Australia, Canada, China, Europe, the U.S., and South Korea.
在地面交通文献中,有大量的文献专注于理解个人的态度、信念、个性和类似因素如何影响出行行为选择,包括对新技术的采用。附录中的表 3 总结了 19 项地面交通研究,这些研究考察了个人对电动汽车、自动驾驶汽车或共享计划的态度对其影响。其中 16 项研究采用了调查,两项采用了访谈,一项回顾了文献。这些研究涉及到澳大利亚、加拿大、中国、欧洲、美国和韩国等多个国家。

Table 3. Ground transportation studies that have examined the influence of customer attitudes towards EVs, AVs, or sharing programs.
表 3. 研究了顾客对电动汽车、自动驾驶汽车或共享计划态度影响的地面交通研究。

Study 学习Location 位置Sample Size 样本大小Key Findings 主要发现
Axsen, Goldberg and Bailey (2016)
Axsen, Goldberg 和 Bailey(2016 年)
Canada 加拿大94 plug-in EV owners and 1754 conventional auto owners
94 个插电式电动汽车车主和 1754 个传统汽车车主
Early adopters of EVs have higher levels of environmental concerns and higher engagement in environment- and technology-oriented lifestyles.
Bansal and Kockelman (2018)
Bansal 和 Kockelman(2018 年)
Texas, USA 德克萨斯州,美国1088About 50% of respondents will likely time their AV adoption in conjunction with their friends. Environmental friendliness and cost savings were factors in current carsharing users.
大约 50%的受访者可能会与朋友一起选择采用自动驾驶技术。环保和节省成本是目前共享汽车用户的考虑因素。
Bennett, Vijaygopal, and Kottasz (2019)
贝内特、维贾戈帕尔和科塔斯(2019 年)
UK444 physically disabled; 353 with no physical disability
444 身体残疾;353 无身体残疾
Higher levels of interest in new technology associated with intention to use AVs.
Biresselioglu et al. (2018)
Biresselioglu 等人(2018 年)
Review article with focus on Europe
N/AMotivation to purchase EVs are influenced by environmental, economic, and technical benefit.
Cherchi (2017)Denmark 丹麦2363Social conformity effects (e.g., word of mouth) were just as important as vehicle characteristics on intention to purchase EV.
Degirmenci and Breitner (2017)
Degirmenci 和 Breitner(2017 年)
Germany 德国40 interviews 40 次面试Environmental performance of EVs is a stronger predictor of EV purchase intention than price value and range confidence.
Hohenberger et al. (2016)
Hohenberger 等人(2016 年)
Germany 德国1603Anxiety associated with AVs can be mitigated through providing safety-related information.
Huang and Qian (2018) 黄和钱(2018 年)South Jiangsu region, China
348Social conformity effects (word of mouth, peer influence) positively influenced consumer preference for EVs; risk-aversion negatively influenced EV preference.
Kim et al. (2019) 金等人(2019 年)Georgia, USA 美国乔治亚州2890Respondents who would select air over AV tended to be more tech-savvy.
Kim et al. (2015) 金等人(2015 年)Seoul, Korea 首尔,韩国533 participants in an EV carsharing program
533 名电动汽车共享计划参与者
Individuals with higher environmental concerns and higher concern for what others think are more likely to purchase an EV.
Lane et al. (2018) Lane 等人(2018 年)United States 美国1080Respondents preferring a battery EV over a plug-in EV were drawn to its environmental and technical appeal.
Li and Kamargianni (2019)
李和卡马尔加尼(2019 年)
Taiyuan, China 太原,中国3486Pro-environmental attitudes are positively associated with an intention to use bike-sharing.
Liu, Ma, and Zuo (2019)
刘、马和左(2019 年)
China 中国213 college students 213 名大学生Highlighting the environmental advantages of AVs and increasing public trust in AVs may increase societal acceptance of AVs.
Potoglou et al. (2020) Potoglou 等人(2020 年)Germany, India, Japan, Sweden, U.K., and U.S.
6033Individuals self-identifying as having a pro-environmental identity and as being innovators were more in favor of automation and AVs.
Smith et al. (2017) 史密斯等人(2017 年)Perth, Australia 澳大利亚珀斯440Individuals who always selected EVs as preferred choice among six trade-off questions were more concerned with the environment.
Sovacool et al. (2019) Sovacool 等人(2019 年)Denmark, Finland, Iceland, Norway, Sweden
5067 online surveys and 257 interviews
5067 份在线调查和 257 次面试
Pro-environmental and safety attitudes are positively associated with women’s preferences for EV vehicles.
Sweet and Laidlaw (2019)Toronto and Hamilton, Canada3201Individuals who are first to try out a new product and live a hectic life are positively associated with interest in using AVs.
Tsouros and Polydoropoulou (2020)Greek Islands of Lesvos and Chios550Tech-savvy individuals are more likely to purchase vehicles with higher levels of automation, and pro-environmental individuals are more likely to purchase hybrids.
Wang et al. (2020)China426, primarily university studentsEarly adopter of new technology, environmental awareness, and perceived usefulness is positively associated with intention to use ridesharing.

Three key themes emerge from these studies. First, individuals with pro-environmental attitudes are more likely to prefer (and by extension adopt, use, or purchase) EVs and/or AVs over conventional vehicles (Axsen et al., 2016, Biresselioglu et al., 2018, Kim et al., 2015, Potoglou et al., 2020, Smith et al., 2017, Sovacool et al., 2019, Tsouros and Polydoropoulou, 2020). Environmental performance of EVs was a stronger predictor of EV purchase intention than price and range confidence in a study by Degirmenci and Breitner (2017). Given a choice between a hybrid and battery-EV, respondents preferring a battery-EV were drawn to its environmental appeal (Lane et al., 2018). Pro-environmental attitudes were also positively associated with intention to use bike-sharing (Li and Kamargianni, 2019), ride-hailing (Wang et al., 2020), and carsharing (Bansal and Kockelman, 2018, Kim et al., 2015). Liu, Ma, and Zuo (2019) found that highlighting the environmental advantages may increase social acceptance of AVs.

Second, tech-savvy individuals who have higher levels of interest in new technology, a technology-oriented lifestyle, and/or are individuals who are the first to try out a new product were found to be early adopters of EVs (Axsen et al., 2016, Biresselioglu et al., 2018), have higher intentions of using AVs (Bennett et al., 2019, Potoglou et al., 2020, Sweet and Laidlaw, 2019), and were more likely to purchase AVs with higher levels of automation (Tsouros and Polydoropoulou, 2020). Individuals from the U.S. who preferred battery-EVs over plug-in EVs were also drawn to its technological appeal (Lane et al., 2018). Early adopters of new technology were positively associated with the intention to use ride-hailing services (Wang et al., 2020). In a study that included both AV and commercial air, respondents who selected air over AV tended to be more tech-savvy (Kim et al., 2019). For individuals who are anxious about using new AV technologies, one study found that this anxiety could be mitigated through providing safety-related information (Hohenberger, Spörrle, and Welpe, 2016). Sovacool et al. (2019) found that safety attitudes were positively associated with women’s preferences for EV vehicles.

Third, multiple studies have found that social effects are important to AV and EV adoption. Bansal and Kockelman (2018) found that about 50 percent of respondents would time their adoption of AVs in conjunction with their friends. Huang and Qian (2018) and Kim, Ko, and Park (2015) found that social conformity effects (such as word-of-mouth and peer influence) positively influenced consumer preference for EVs. Cherchi (2017) found that word-of-mouth effects were just as important as vehicle characteristics on the intention to purchase EVs.

4.2.3. Acceptance and adoption

Numerous papers in the ground transportation field have examined general barriers to EV adoption (e.g., see Berkeley et al., 2017, Berkeley et al., 2018, Kim et al., 2018) and applied or extended theoretical models used to predict the timing of adoption for EVs, AVs, and sharing services. Several review papers have been written, including one by Becker and Axhausen (2017) who reviewed surveys regarding AVs with a focus on methodologies and results as they pertain to acceptance of AVs, and a second by Rezvani et al. (2015) who reviewed the drivers for and barriers against adoption of plug-in EVs and provided an overview of the theoretical perspectives that have been used. In this section, we present an overview of papers that provided general overviews of barriers toward adoption of new ground technologies and papers that modeled the timing of adoption of new ground technologies.

Cunningham et al. (2019) conducted a survey to gauge public acceptability and opinions of AVs within Australia and found that the majority of Australians are currently not willing to pay more for a fully autonomous vehicle than a conventional car. Raj et al. (2020) examined the barriers to AV adoption and found that the lack of customer acceptance is the most prominent barrier.

Multiple authors have applied or extended Davis’ Technology Acceptance Model (TAM) to show that perceived usefulness and perceived ease of use are use predictors or behavioral intentions to have or use new ground technologies (Davis, 1989). These include studies by Globisch et al. (2018) and Wolff and Madlener (2019) that examined acceptance of EVs in commercial fleets, and studies by Panagiotopoulos and Dimitrakopoulos (2018) and Lee et al. (2019) that examined acceptance of AVs and found that perceived usefulness, perceived ease to use, perceived trust and social influence helped predict behavioral intentions to have or use AVs. Zhang, Tao, et al. (2020) extended the TAM to show that at the beginning of AV commercialization, perceived ease of use and perceived usefulness help describe intention to use, but social influence and initial trust contributed most to explain whether users would accept AVs or not. Zhang et al. (2019) showed that initial trust could be enhanced by improving perceived usefulness and reducing perceived safety risk. Adnan et al., 2018b, Khastgir et al., 2018; and Xu, Zhang, et al. (2018) also found that trust is important to AV acceptance and that experience with AVs could increase trust (Xu, Zhang, et al., 2018), as well as providing knowledge about the AV system’s true capabilities and limitations (Khastgir et al., 2018). Du et al. (2019) found that information about AVs provided to respondents before they participated in a driving simulator experiment helped increase trust in and preference for AVs. Wang et al., 2018a, Wang et al., 2018b used an extended TAM to show that consumers’ lack of knowledge and risk perceptions could be barriers to the acceptance of EVs, and Wang et al. (2020) used an extended TAM to show that personal innovativeness, environmental awareness, and perceived usefulness are positively associated with the intention to use ride-hailing services, whereas perceived risk is negatively associated with intention to use and perceived usefulness.

Other theoretical frameworks have been used to model adoption and timing of new ground vehicle technologies. For example, Adnan, Nordin, Amini, and Langove (2018) used the Theory of Planed Behavior (Ajzen, 1985) to examine the adoption of plug-in hybrid vehicles in Malaysia, and Wang et al. (2016) used this theory to examine adoption of hybrid EVs in China. Roger’s Diffusion of Innovations Theory (Rogers, 2003) has been used in multiple studies. Kröger et al. (2019) examined potential AV market penetration in the U.S. and Germany; Prieto et al. (2017) examined diffusion of carsharing services in London, Madrid, Paris, and Tokyo; and Zhang, Schmöcker, et al. (2020) examined diffusion of a one-way carsharing system in Tokyo. Shabanpour, Shamshiripour, and Mohammadian (2018) modeled the timing of AVs that considers individuals’ desires to innovate and need to imitate the rest of society, and Talebian and Mishra (2018) predicted the adoption of connected AVs and found that information individuals receive from peers was a key influence of adoption.

Two studies have extended discrete choice models to incorporate timing effects associated with adoption. El Zarwi et al. (2017) integrated discrete choice and TAM models to predict the adoption timing of a one-way carsharing service. They found that adoption is influenced by social influences, network effects (e.g., placement of stations), level of service attributes, and sociodemographics and that placing a carsharing location outside a major technology firm induced the highest expected increase in the monthly number of adopters. Liu and Cirillo (2018) used a generalized dynamic discrete choice model to predict the initial and repeat purchases of alternative fuel vehicles that accounts for technology improvements and changes in prices over time.

In summary, studies based on TAM have confirmed that that perceived usefulness and perceived ease of use, perceived trust, and social influence help explain the adoption of EVs, AVs, and ride-hailing services, whereas perceived risk is a barrier to adoption. Providing safety information and general information about the technical capabilities and limitations of new transportation technologies are strategies that authors have identified for increasing comfort in the new technologies. Within the UAM, only one study has applied the TAM framework. Al Haddad et al. (2020) extended the TAM framework and confirmed the importance of safety and trust and affinity to automation in the timing of the adoption of UAM.

4.2.4. Value of time

UAM offers the potential for travel time savings. Several studies within the ground transportation literature have focused on evaluating the value of time for AVs. As air taxis enter the market, they may be competing with AVs, thus the findings from these studies are particularly relevant for UAM demand and pricing studies.

Multiple studies have noted that VOT is related to productivity and two theoretical papers have shown that the VOT for AVs will be less than the VOT in a conventional ground vehicle. Correia et al. (2019) presented a theoretical model for VOT, noting that “full automation will enable passengers to perform other, non-driving, related tasks while traveling to their destination. This may substantially change the way in which passengers experience traveling by car, and, in turn, may lead to considerable changes in [VOT].” Pudāne and Correia (2020) adapted this model, showing that “if automated vehicles provide identical work or leisure experience to out-of-vehicle locations, then the opportunity costs of travel time are erased and the (VOT) equals the intrinsic costs of travel, which is strictly smaller than the VOT in a conventional vehicle.”

Several empirical studies have confirmed this theoretical result. Based on a survey of approximately 500 individuals from the Netherlands, Correia et al. (2019) found that the average VOT for an AV with an office interior12 (5.50€/hr; $6.16/hr USD) was lower than the VOT for a conventional car (7.47€/hr; $8.37/hr USD); no significant differences in VOT were found between the AV that contained a leisure interior and a conventional car. Based on a survey of approximately 500 individuals from Germany, Kolarova et al. (2019) found an average value of travel time savings (VTTS) reduction of 41 percent for the AV compared to a conventional car for commuting trips; no significant changes in the average VTTS were found for leisure or shopping trips. Gao et al. (2019) found that VOT was 13 percent lower when being driven in a ride-hailing service than a personal car and, further, that mentioning the ability to multi-task explicitly led to a much lower VOT, approximately half that of driving oneself. However, noting that the ride-hailing service was driverless led to a 15 percent higher VOT compared to driving a personal car, “which may reflect a lack of familiarity and comfort with driverless technology at present” (Gao et al., 2019).

These findings are important, as they suggest that the UAM community should not use an average VOT, but rather incorporate a distribution of VOTs across the population that accounts for “non-adopters.”

4.2.5. Willingness to share rides with strangers

As noted by Kolarova et al. (2019), prior results in the literature have shown that using a shared autonomous vehicle alone and sharing the journey are perceived as two distinct mobility options (Krueger et al., 2016), which may be due to psychological barriers or discriminatory attitudes associated with sharing a ride with a stranger (Correia and Viegas, 2011, Middleton and Zhao, 2019). For example, based on focus groups of older adults in the province of Utrecht, the Netherlands, Faber and van Lierop (2020) found that participants had a strong interest in using AVs in their daily life and that the option to travel with friends was an important factor in having a positive attitude toward AV adoption. Lavieri and Bhat (2019) noted that an important obstacle to ride-hailing adoption is the user’s willingness to share rides with strangers and “recent studies indicate that travelers are hesitant about being in an automobile environment with unfamiliar faces, due to a desire for personal space, an aversion to social situations, distrust, and concerns about security and privacy (see, for example, Tahmasseby et al., 2016, Morales et al., 2017, Amirkiaee and Evangelopoulos, 2018).” Based on a 2017 survey of 1607 commuters in the Dallas–Ft. Worth–Arlington Metropolitan area, Lavieri and Bhat (2019) examined individuals’ willingness to share trips with strangers in an AV. They found that privacy is a main deterrent to pooled ride-hailing service, with non-Hispanic whites being more privacy sensitive than individuals of other ethnicities. However, they found that respondents are less sensitive to the presence of strangers when in a commute trip compared to a leisure-activity trip and found evidence that the travel time added to the trip to serve other passengers may be a greater barrier to the use of shared services compared to the presence of a stranger.

Conversely, a study of Australians found that riding with strangers was more onerous than the added trip time. Based on a survey conducted in 2018 of 3985 Australians that asked for their preferences for a ground on-demand transportation system, Vij et al. (2020) found that consumers are willing to pay13, on average, AUD $0.28/km (USD $0.33/mi) more to avoid sharing a vehicle with other passengers, AUD $0.17/km (USD $0.20/mi) more for door-to-door service, and AUD $0.10/km (USD $0.12/mi) to be able to book the service in real time as opposed to having to book the service several hours in advance. All trip purposes were included in their analysis.

The willingness to travel with strangers may be related to ride-hail usage and whether individuals in general like to interact with other people. Based on a database of 6.3 million Lyft trips taken in Los Angeles County in 2016, Brown (2020) found higher rates of ride-hailing usage among frequent Lyft users compared to moderate and less-frequent users, which “suggests either that repeat users seek more economical service options and/or repeated ride-hail use increases or is associated with peoples’ comfort in sharing cars with strangers.” Based on focus groups of individuals from Denmark, Nielsen et al. (2015) found that some Danish negatively perceive ride-hailing with strangers due to “social awkwardness,” whereas other Danish positively perceive ride-hailing with strangers due to the ability to “socialize” with others.

The willingness to travel with strangers may also be related to modes, given individuals are more used to traveling with strangers by air than in an automobile. In the UAM context, Garrow, Roy, and Newman (2020) found that the willingness to ride with strangers varied across the AV and air taxi modes, with those ages 18–24 less willing to travel with strangers in an AV than those ages 25–64 in an AV and that the willingness to travel with strangers was about the same for the air taxi (across all ages) as for those ages 25–64 in an AV. Finally, it is important to note that while many studies point to the willingness to pay to travel with strangers, the result is not consistent across all studies. Based on a survey of approximately 500 individuals from Germany, Kolarova et al. (2019) did not find any differences between using a shared AV alone or with others.

4.3. Bringing it all together—Demand modeling insights and research directions for UAM

As seen from the literature review, demand modeling within the UAM and EV/AV domains have focused on different objectives. Within the UAM field, the primary focus has been on determining if UAM is a viable concept—e.g., will enough people be willing to fly in these new air taxis and can the service be supported across different cities? Within the ground transportation field, EVs, AVs with lower levels of automation, and sharing services have already been implemented, allowing researchers to focus on understanding SED characteristics of early adopters or how individuals respond to different policy incentives and operational policies.

To the extent that individuals who are interested in EV, AV, and/or sharing modes will also be interested in air taxis, we would expect that early adopters of air taxis will be more likely to be male, have higher incomes, have pro-environmental attitudes and/or be tech-savvy, technology-oriented lifestyles and be the first to try out new products. These expectations have been confirmed in surveys of U.S. commuters by Boddupalli et al. (2020) and Garrow, Roy, and Newman (2020).

The EV and AV literature have several findings that are relevant for the UAM community. To date, there has been a significant amount of research in the EV and AV literature that has looked at the timing of when adoption occurs, but only one paper in the UAM area, by Al Haddad et al. (2020). These technology adoption models can provide valuable information on the role of trust, safety, and perceived usefulness on the adoption of UAM. The literature across both the air and ground transportation areas show mixed reactions in the population with respect to autonomy. Finding ways to increase individuals’ trust in autonomy would be a valuable direction for future research. For example, we may find that it is important to provide demonstrations of what it would be like to fly in a UAM using virtual reality and/or to provide safety information to increase individuals’ comfort levels with the new technology. The role of social effects (like trusting perceptions of friends and family) has been shown to play a role in adoption of ground vehicle technologies and could be investigated in the context of UAM.

Unlike with ground transportation modes, individuals are more likely to expect to travel with strangers in an aircraft. It is, thus, unclear whether the same effects seen for ride-hailing services will apply to UAM. One study, by Garrow, Roy, and Newman (2020) did find that younger commuters were less likely to take a UAM with strangers compared to older commuters. However, there is a research need to understand if the willingness to travel with strangers in a UAM aircraft varies across nations and trip purposes.

Perhaps one of the most interesting findings from the AV literature is that the VOT for commuters will decrease when ground AVs enter the market due to the ability for commuters to use their time more productively. From the UAM perspective, this is important as it suggests that AVs will compete more heavily with air taxis than with conventional autos and that additional travel time savings will be required for the air taxi mode relative to the AV. Potential productivity gains in an AV compared to an air taxi have not been explored in the literature, and there is a need to determine what levels of productivity would be achievable in a UAM vehicle and how productivity varies as a function of ride quality, trip duration, and other factors. Given the VOT decreases seen for AV ground research, better understanding of VOT decreases for UAM vehicles—particularly as they relate to commute trips—is an important area of future research.

Another interesting avenue for future research would be to explore how AVs and air taxis will compete across different trip purposes as an air taxi system evolves and adoption rates increase across both new modes.

5. Integration with existing modes and infrastructure

UAM has the potential to transform urban travel by providing faster connections among residential, business, sports, medical, and other facilities. To achieve this goal, air taxis will need to fly close to and/or over high-density population areas and integrate with existing city infrastructure—including other modes of transportation, the electric grid, and the NAS. As such, there will be many questions that the aviation community will need to address with respect to how we can safely integrate UAM into existing infrastructure while ensuring equitable access. This section reviews infrastructure-related topics that have been investigated by the UAM and EV/AV areas.

5.1. Review of UAM infrastructure studies

Several UAM researchers have focused on infrastructure-related issues, mostly in the context of UAM operations. This section highlights the tight couplings researchers have observed among vertiport placement, operations, demand, and energy requirements.

5.1.1. Vertiport Placement, Design, and airspace integration

Multiple terms have been used for vertiports, including vertipads, vertistops, and skyparks (Vascik and Hansman, 2017). In this paper we will refer to vertiports for eVTOL operations and STOLports for operations that involve short takeoff and landing flights. Multiple types of locations have been suggested as possible infrastructure that could be used to integrate vertiports into cities, including rooftops with parking lots and/or parking decks (Kreimeier et al., 2018, Robinson et al., 2018, Uber Elevate, 2016), vacant land, floating barges, pre-existing airports and helipads (Robinson et al., 2018), the land adjacent to highways and/or in cloverleaf interchanges, parking lots at places of worship that may be used only on weekends, large stadiums or concert venues that are unused for large portions of the year, the corner of a parking lot in large superstores or malls, and technology campuses (Uber Elevate, 2016).

Several studies have examined optimal locations for vertiports to serve different types of demand. Most of these studies are focused on finding which census tracts and/or larger geographic area would be ideal locations, instead of actual siting. For example, Lim and Hwang (2019) investigated how competitive eVTOL would be for commuters in the Seoul metro area by increasing the number of vertiports from 2 to 36; Daskilewicz et al. (2018) found vertiport locations that maximize population-cumulative potential travel time savings compared to driving in San Francisco and Los Angeles; and German et al. (2018) formulated an optimization problem to find vertiport locations for a cargo demand application in the San Francisco Bay area. As part of a broader study that identified eight operational constraints that could limit or prohibit UAM service, Vascik, Hansman, and Dunn (2018) found that the three most stringent constraints concerned community acceptance of aircraft noise, vertiport availability, and air traffic control scalability.

In terms of vertiport designs, several architectural firms have presented visions (Uber Elevate, 2020b). For example, Vascik and Hansman (2019) considered how different vertiport designs (defined by the number of touchdown and liftoff pads, number of aircraft gates, and number of aircraft staging areas/parking spaces) and the layout of these designs (which include linear, satellite, pier, and remote apron topologies) impact vertiport capacity envelopes. They found that the ratio of gates to touchdown and liftoff pads is a key design parameter, that aircraft staging areas can provide significant benefits, and that vertiports with multiple touchdown and liftoff pads can greatly increase throughput.

Several studies have examined how constraints on the paths aircraft use to take off and land from a vertiport restrict the number of locations that can be used for siting vertiports. Conceptually, even though eVTOL aircraft can hover, they typically climb from and approach a vertiport at an angle to conserve energy reserves and to operate in safe areas of their flight envelopes (e.g., see Yilmaz et al., 2019). The same design criteria and guidance used for helipads can be used as a starting point for siting of vertiports, e.g., the departure and approach paths must be free of obstacles and consider historic wind patterns. Two FAA documents that are particularly relevant in this context include FAA Advisory Circular AC-150/5390-2C Heliport Design (FAA, 2012) and FAA Instrument Procedures Handbook FAA-H-8083-16B (FAA, 2017).

To date, we could find no published studies that explicitly examined the optimal placement of vertiports for eVTOL operations that considered port design criteria; however, work is in progress by Tarafdar et al. (2020) that uses Zillow's Assessor and Real Estate Database (ZTRAX) to identify parcel-level characteristics important for siting (Zillow, 2020). Several studies have examined the optimal placement of STOLports for short takeoff and landing operations in urban and suburban areas of South Florida, which includes the Miami metro area (Robinson et al., 2018, Justin and Mavris, 2019, Somers et al., 2019, Wei et al., 2020). Robinson et al. (2018) found that an average density of 1.66 STOLports per square mile can be achieved with 300-ft-long runways. Subsequent studies by Justin and Mavris, 2019, Somers et al., 2019 and Wei et al. (2020) built on this initial analysis by accounting for obstacles and historic wind patterns and formulating an optimal facility location problem among potential sites.

Finally, several authors have pointed out how the placement of vertiports needs to integrate with airspace restrictions. Verma et al. (2019) looked at near-term routes for UAM based on current-day helicopters routes in DFW, and Vascik and Hansman (2017) used radar trajectory data recorded by the FAA Airport Surface Detection Equipment Model X (ASDE-X) from LAX to identify areas where it may be feasible to route future UAM operations due to the low volume of conventional operations. Air taxi trips into a major commercial airport pose particular challenges due to the need to coordinate trajectories with existing commercial operations. Vitalle et al. (2020) looked at route design for eVTOL aircraft transporting passengers into Tampa International Airport (TPA). They consider three possible vertiport locations at or near the airport: a helipad located two miles from the airport, the rooftop of the economy parking lot in the main terminal area, and the rooftop of the rental parking garage. The rooftop of the economy parking lot is ideal from the passengers’ point of view in that it reduces the time to reach their commercial gates and remains inside a secure area; however, this location presents greater challenges with designing trajectories, as aircraft would need to land between two parallel runways.

Vascik and Hansman (2020) developed an approach to analytically identify terminal airspace that is procedurally segregated from large aircraft operations and may be appropriate for new airspace cutouts for eVTOL operations. They applied the methodology to the 34 largest metro areas in the U.S. and found that, on average, 65 percent of a city’s population was accessible to vertiports operating under visual flight rules and without air traffic control (ATC) limitations. However, on average, only 34 percent of long-duration commuter workplace locations could be accessed by UAM. Further, a very large variation in accessibility measures existed across the metro areas. The authors found that providing access to special-use airspace, and especially temporary flight restrictions for sporting events, increased commuter workplace access to 54 percent for the median U.S. city.

5.1.2. Battery and electric grid considerations

There is a fundamental trade-off between battery size and mission length. On one hand, bigger batteries have more energy, which can translate into longer missions. However, with the increase in battery size comes additional aircraft weight, which can translate to increased acquisition cost. Based on current battery technology, eVTOL aircraft will likely need to be partially or fully recharged after each mission. Given current battery-charging technologies, the time to perform this charging is likely to deter high aircraft utilization, particularly during peak demand periods. The amount of electricity required to power an electric fleet of aircraft is not trivial and will likely have significant impacts on the electric grid, which may not be able to be supported by the current electric grid.

The issues are described by Kohlman and Patterson (2018) as follows:

“if UAM vehicles are to be all-electric, as many are proposing, there will be new demands placed on the electrical grid infrastructure that must be understood. Additionally, vehicle-level characteristics such as the recharge time or energy used for a flight will have direct impacts on the efficiency, cost, and ultimate viability of UAM networks. For example, if vehicles must be charged for long periods of times between missions, a very large number of charging stations will be required at vertiports and many vehicles may be required to meet demand for UAM services.”

Further, the cost of grid upgrades to support UAM operations is not trivial. A recent report by Black & Veatch estimates the cost to extend an existing service line to support 31 MW chargers to be between $75K and $100K; the cost for a new feeder line to support up to 83 MW chargers to be between $2.6K and $1.3M per mile; a new transformer bank over 10 MW to support over 15 chargers to be between $3M and $11M; and a new substation bank over 20 MW to support 30 chargers to be between $40M and $80M (Stith, 2020).

The impact of charging on operations and the number of required charging stations has been noted by other authors. In a study of cargo operations in the San Francisco Bay Area, German et al. (2018) found that for a lift + cruise eVTOL concept model and a tiltrotor aircraft model, charging times with a 300 kW charger ranged from 12.5 to 19.1 min and 16.0 to 23.1 min, respectively. When the charger was increased to 400 kW, these charge times decreased to 9.5 to 14.4 min and 12.1 to 17.4 min, respectively.

The impacts of UAM operations on the electric grid were clearly demonstrated in a study by Justin et al. (2017). Based on an examination of electric aircraft for regional distances, they generated power profiles for stations where Cape Air and Mokulele Airlines operate. Cape Air’s network included 525 daily flights to 43 airports primarily in the New England area using mostly twin-engine piston-powered Cessna 402s. Mokulele’s network included 120 daily flights to airports primarily in the Hawaiian Islands using 11 single-engine turboprop Cessna 208s. They found very high peak powers at the airlines’ busiest airports, i.e., for Cape Air the peak power exceeded 1 MW in Nantucket Memorial (ACK) airport and in Boston Logan International (BOS) airport, which is the order of magnitude of the demand of approximately 1,000 households. For Mokulele, the peak-power at Molokai airport (MKK) was 517 kW, which is about 1/20th of the total generation capability for the entire island of Molokai (Justin et al., 2017). The authors explored various operational strategies to reduce peak-power demands and the cost of electricity, and found that a strategy that includes optimizing battery recharging with battery swaps can achieve reductions on the order of 20 percent compared to a power-as-needed strategy.

What is clear from these and other publications is that the power requirements on the electric grid are not trivial, and significant opportunities exist to optimize the deployment of charging and fast-charging stations. Furthermore, given that electricity prices vary across cities and providers, the optimal battery recharging solution will likely be city-dependent.

5.2. Review of EV and AV infrastructure studies

This section provides an overview of the types of questions that researchers and policy makers have investigated when integrating a new mode into the existing ground transportation network, and the role of parking availability on mode choice. This discussion is followed by a detailed review of a topic that is particularly relevant to UAM: integration with the electric grid.

5.2.1. Integration with existing modes and the ground transportation network

As new technologies and transportation modes enter the market, transportation planners need to understand whether these modes will complement or compete with existing modes. For example, is ride-hailing complementary with public transit in providing first-mile and/or last-mile access, or does ride-hailing replace public transit trips? Transportation planners are also interested in longer-term impacts, such as whether carsharing reduces car ownership or influences households’ residential location choices. Finally, transportation planners often model “rebound effects,” which occur when new technologies result in increased travel that can have negative environmental impacts, e.g., through the generation of more trips or longer trips. For example, will AVs reduce the need for parking but generate longer trips due to the fact they can drop passengers off and travel back home and/or travel to a less expensive and more remote location to park?14 As shown in Fig. 3, more than 30 studies exist within the ground transportation literature that look at how EV, AV, and/or ride-hailing services will integrate or compete with existing infrastructure. In addition, the U.S. Department of Transportation has funded studies that have developed planning tools for on-demand mobility that consider UAM (Shaheen et al., 2020).

From a UAM perspective, what is most relevant about these studies is not necessarily the results, but rather the underlying motivations for why transportation planners are asking these questions. Within the U.S., urbanized areas that have a population of greater than 50K are required to have a metropolitan planning organization (MPO) that is responsible for establishing a long-term transportation improvement plan (TIP) that sets transportation investment priorities in the area (FTA, 2019). Major federal transportation authorization bills, such as the Moving Ahead for Progress in the 21st Century Act (MAP-21) and the Fixing America’s Surface Transportation (FAST) Act, establish regulations that MPOs must follow in order to receive transportation-related funding. These regulations require that the selection of projects be based on performance metrics, equity considerations, and other criteria (e.g., see US DOT, 2013). In the case of congestion management plans, the regulations state that an MPO’s TIP “must include regional goals for reducing peak-hour vehicle miles and improving transportation connections and must identify existing services and programs that support access to jobs in the region … [23 U.S.S. 134(k)(3)]” (FHWA, 2016). Other federal legislation is critical to transportation planning and funding priorities, including the Clean Air Act (CAA), Clean Water Act (CWA), National Environmental Policy Act (NEPA), National Historic Preservation Act (NHPA), and Americans with Disabilities Act (ADA) (EPA, 2020a, EPA, 2020b, NPS, n.d., US DOJ, 2020). From a UAM perspective, it is important to note that as we integrate this new mode into our cities, government funding for infrastructure improvements will likely be tied to these or similar regulations.

Given the focus (not only in the U.S., but in many countries throughout the world) on transitioning to clean energy and reducing the negative impacts of the transportation sector on carbon emissions, many studies that look at integration of new ground technologies with existing modes and infrastructure consider metrics that tie to these goals, including total vehicle miles traveled (VMT) and greenhouse gas emissions. For example, Jones and Leibowicz (2019) examined these issues in the context of shared AVs, and Shen et al. (2018) examined these issues in the context of an integrated AV and public transit system for Singapore. Bansal et al. (2016) examined long-term adoption of shared AVs in Austin, Texas, and simulated long-term adoption under different scenarios to help assess sustainability impacts. Ai et al. (2018) found through siting EV charging stations near public transit in Chicago, Illinois, that commuters can reduce up to 87 percent of personal VMT and 52 percent of carbon emissions, and Muñoz-Villamizar et al. (2017) evaluated environmental impact and delivery cost implications of using an all-electric fleet of delivery vehicles in Bogotá, Colombia.

One of the key findings from the ground transportation literature that is directly applicable to UAM research is the role of parking availability on mode choice. Within the ground transportation field, several researchers have explored the relationships among mode choice, work departure time, ground transportation congestion, and availability of parking at the work destination (e.g., see Tian, Sheu, and Huang, 2019; Wang et al., 2019). Intuitively, we expect that individuals are more likely to take a ride-hailing service and/or transit modes compared to a conventional auto if there is limited parking availability at their destination. This has important implications for UAM. First, if parking in business centers is expensive and/or limited, using an air taxi for commuting will be more competitive with auto. Similarly, if on-site parking at airports15 reaches capacity at certain times of the day and/or days of the week, then using an air taxi to travel to and from the airport will be more competitive with auto (although maybe not as competitive with ride-hailing services).

It is also important to note that UAM may interact with other modes in both positive and negative ways, and the ultimate impacts on energy usage and sustainability have yet to be determined (ARPA-E, 2021). Congestion in the ground transportation systems at the network level may improve as the air system provides additional capacity for transporting individuals and could reduce the number of long-distance ground vehicle trips. This is particularly relevant if UAM is able to replace single-occupancy ground transportation trips with shared-occupancy air trips, thereby leading to improvements in energy use and reductions in greenhouse gas emissions. In the context of SAVs, multiple authors have evaluated the energy and environmental impacts of automated vehicles (e.g., see Greenblatt and Shaheen, 2015; Lee and Kockelman, 2019; Fleming and Singer, 2019, Brown and Dodder, 2019). Shaheen and Bouzaghrane (2019) note that “while the literature on potential SAV impacts on travel behavior and the environment is still developing… [researchers] speculate that SAVs would result in a 55% reduction in energy use and [about a] 90% reduction in greenhouse gas (GHG) emissions.” In the context of UAM, multiple strategies for reducing energy usage and GHG emissions are being investigated that include vehicle improvements (e.g., through the design of electric propulsion systems and batteries) as well as connectivity and automation strategies (e.g., through trajectory optimization or optimized network coordination strategies with air and ground traffic systems). Representative studies include Pradeep and Wei, 2019, Shihab et al., 2019 and Kasliwal et al. (2019).

5.2.2. Integration with the electric grid

Within the ground transportation literature, multiple studies have examined how and where to place recharging infrastructure, how plug-in EVs (PEVs) and consumers’ charging behaviors will impact the electric grid, and how policy strategies can be used to help reduce peak loads on the electric grid. In the area of charging infrastructure site selection, optimization-based and activity-based approaches for selecting recharge sites have been developed (Dong et al., 2014, Yang et al., 2017, Li et al., 2020, Lin and Greene, 2011). These approaches are typically formulated to maximize the feasibility of daily use of EVs within a community. Additional work has focused on site selection for intercity trips (Xie et al., 2018, Xie and Lin, 2021).

Significant research has also focused on the impact of charging on the electric grid and how this impact is affected by driver behavior relating to charging. Hardman et al. (2018) reviewed the literature as it pertains to infrastructure requirements for PEVs and found that PEV charging will not impact electric grids in the short term but may need to be managed long term. Marmaras et al. (2017) modeled the impact of EV driver charging behavior on the transportation and electric grid networks. They found that EV driver behavior has “direct and indirect impacts on both the road transport network and the electricity grid.” They examined consumer charging preferences (e.g., normal charging at home, normal or fast charging at a public charging station) and offered operational strategies to help shift peak loads at public charging stations. Additional studies focused on the impact of driver behavior on EV demand, charging infrastructure usage, and corresponding impacts on the electric grid include Tal et al. (2014), Chakraborty et al. (2019), and Lee et al. (2020). Considering the importance of recharging costs on driver behavior, multiple studies have also investigated the economics of recharging based on the cost of electricity for fast charging and associated infrastructure costs (Muratori et al., 2019a, Muratori et al., 2019b, Borlaug et al., 2020).

Significant research has also been conducted related to policies and strategies to manage the impact of charging on the electric grid. These studies have shown that controlled charging of EVs, including time-of-day pricing, can better balance loads on the electric grid and impact power grid loads, voltage, frequency, and power losses (Bailey and Axsen, 2015, Daina et al., 2017, Latinopoulos et al., 2017, Xu et al., 2017). Luo et al. (2020) went one step further by jointly designing charging station and solar power plants with time-dependent charging fees to improve management of transportation and power systems.

Some studies have analyzed interactions between the electric grid and e-mobility and how these interactions are dependent on policy choices. These studies are difficult to conduct in some countries due to limited information that is publicly available about the electric grid (e.g., capacity and loads as a function of different times of the day); therefore, it is common to produce EV charging profiles and examine how these profiles are affected by different policies (such as changing time of day pricing), e.g., see Delgado et al. (2018) for a study in Portugal. A notable exception is a study by Kannan and Hirschberg (2016) who used a detailed energy model developed for Switzerland and found that the cost effectiveness of e-mobility depends on policy decisions in the electric sector.

Within the ground transportation literature, there has been considerable interest in vehicle-to-grid technologies and policies in which energy is stored in EVs and returned to the grid when it is needed, generating revenues for the EV owner (e.g., see Kester et al., 2019, Nourinejad et al., 2016, Sovacool et al., 2019). Within a UAM context, vehicle-to-grid approaches are likely not a viable option, given the high costs of aerospace-grade battery packs and the battery degradation that would occur through the charging and discharging cycles. Additionally, this use case would likely further add to the challenge of certificating UAM battery packs with national aviation regulatory agencies. However, the concept could be adapted to UAM applications by using batteries that are no longer viable for use onboard the aircraft to store energy on the ground at or near vertiports, e.g., by charging these ground batteries during less-expensive off-peak hours and then using them to charge the flight batteries in the UAM aircraft during peak-period operations.

Given the high cost of batteries, several studies have examined how different recharging strategies, including battery swapping and fast charging, can be optimized to help regulate the charge profile and enhance battery life (e.g., Amjad et al., 2018, Sweda et al., 2017, Pelletier et al., 2018, Keskin and Çatay, 2016, Liao et al., 2016, Qin et al., 2016, Wu and Sioshansi, 2017, Widrick et al., 2018). The study by Pelletier and colleagues offers one of the more comprehensive optimization models and incorporates realistic charging processes, time-dependent energy costs, battery degradation, grid restrictions, and facility-related demand charges for a fleet of electric freight vehicles. They found that fast chargers may be required for vehicle operation flexibility when longer routes are performed.

Optimal charging strategies for carsharing and fleet vehicles have received a lot of attention in ground literature in part because EV charging has a significant impact on EV downtime. For example, in a study by Roni et al. (2019) they noted that in free-floating EV carsharing fleets “downtime due to charging, including time spent traveling to and waiting in queues at charging stations in a sparse charging infrastructure network is a major barrier to sustainable operations.” The authors found that fleet vehicle charging time comprises 72–75 percent of the total downtime spent on charging trips and that adding new charging stations reduced total charging trip travel time but did not significantly reduce total downtime. These results are relevant for UAM because they show that a significant operational bottleneck is related not only to battery recharging but to queuing for battery charging. Shen et al. (2019) and Amjad et al. (2018) provided review articles that cover EV charging operations and optimization approaches.

5.3. Bringing it all together—Infrastructure insights and research directions for UAM

Across both the UAM and EV/AV literature, station placement has been shown to be a critical factor influencing overall system performance. The sheer volume of publications in the EV area that have focused on charging infrastructure or charging type is noteworthy—about one out of every five EV papers we inventoried addressed these topics, as indicated in Fig. 3. Within the EV community, significant attention has been placed on understanding individuals’ charging behavior and strategies for shifting charging patterns to reduce the peak period load on the electric grid. This is relevant from a UAM perspective, as it suggests that the transportation community is already experiencing challenges associated with charging a ground EV fleet. Some EV research has suggested that current electric grids won’t be able to support future EV ground vehicle charging needs. Needless to say, if we are not in a position to handle charging of a ground EV fleet, how are we going to handle charging a UAM fleet that will likely require even faster charge times? There is a clear research need to better understand the power profiles of UAM fleets and develop strategies for how to optimally charge UAM fleets without overwhelming the electric grid. Another interesting topic would be to jointly examine the power profiles for UAM fleets and EV ground fleets, as both technologies will be competing for a limited amount of electricity.

As will become more evident in the next section that focuses on operations, the placement of vertiports is closely coupled with operations and other factors. All of this points to the need to develop high-fidelity simulation models for UAM operations that capture interactions among vertiport locations, vertiport topology, demand, pricing, dispatching, and airspace restrictions.

In comparing the UAM and AV/EV literatures, we could find no mention of the role of parking availability on air taxi mode choice, and we suggest this could be an interesting factor to include in future air taxi mode choice studies, particularly studies that included ride-hailing, air taxis, and traditional autos as potential modes. As we integrate UAM into our cities, it will be important to work with local planning organizations to ensure that any infrastructure investments that require public funding align with regulations these organizations need to follow.

6. Operations

This section reviews operations-related topics explored by the UAM and EV/AV research communities and identifies results from the EV/AV areas that can help inform future UAM research.

6.1. Review of UAM operations studies

As the UAM community designs an air taxi system capable of high-volume throughput integrated in urban areas, many operations-related questions arise. One of the first steps in the analysis process is to design a concept of operations (ConOps), which is essentially a plan for how UAM operations can be safely integrated into the national airspace system. In June 2020, NASA released its ConOps vision, which includes UAM corridors in the sky in which aircraft could operate without the direct involvement of air traffic control (ATC) (Bradford, 2020). Given the importance of ensuring safe operations within the existing NAS, it is not surprising that the UAM community has focused significant attention on ATC-related issues (as shown in the meta-analysis, Fig. 2).

Given a concept of operations, researchers can assess whether a particular aircraft design can successfully and economically perform a given mission, and if it cannot, make modifications to the aircraft design (e.g., see Clarke et al. 2019). To determine whether a mission can be performed successfully for an electric-powered aircraft, researchers need to model the mission’s power and energy requirements, which imply the peak current and total capacity required by the battery. For example, Kulkarni et al. (2018) developed an on-board battery monitoring and prognostic architecture for batteries on electric-propulsion aircraft. Alnaqeb et al., 2018, Prabhakar et al., 2020 developed models to predict mission-based energy and performance metrics; Donateo and Ficarella (2020) proposed a modeling approach for the degradation of the battery performance during its aging; and Pradeep and Wei (2019) developed energy-efficient trajectory plans for a multirotor eVTOL. Hamilton and German, 2017, Hamilton and German, 2019 optimized airspeeds for electric aircraft operations to maximize energy feasibility in the schedule by balancing energy expended during cruise and energy replenished during recharge.

Several researchers have investigated the relationships among aircraft design parameters and operational requirements such as cruise speed and hover time. For example, using the Uber eCRM-001 eVTOL common reference model (Uber Elevate, 2020a), Ha et al. (2020) jointly optimized aircraft design parameters in addition to operational parameters to achieve a 9.66 percent decrease in required hover power. Other researchers have examined the potential for retrofitting existing aircraft with an electric propulsion system to determine if such aircraft could profitably operate for pilot-training applications (Olson, 2015) or short-haul UAM intracity commuter trips in U.S. cities (Kotwicz et al., 2019).

UAM clearly will not be successful without a ConOps that safely integrates aircraft into the NAS and aircraft that can complete the required missions. Thus, it is not surprising that much of the research by the UAM community has been focused on mission performance and related areas, such as battery design and battery modeling. However, as the vision for UAM ConOps and aircraft designs has become clearer, the UAM community is starting to focus on more complex operational issues that include dispatching algorithms and pricing approaches, which are similar topics explored by the EV and AV communities. For example, Roy et al. (2019) examined how existing infrastructure, resources, and operational strategies could be leveraged with improvements in battery and autonomy for regional air mobility. Roy, Crossley, et al. (2018) jointly optimized aircraft designs, operations, and revenue management, and Roy et al. (2020) developed a dispatch model to optimally schedule UAM flights for a shuttle service to an airport that has both scheduled and on-demand customers. Shihab et al. (2019) developed a model to decide whether to offer on-demand or scheduled flights, and how to dispatch the fleet and schedule operations based on simulated market demand. Munari and Alvarez (2019) assigned aircraft to on-demand requests while accounting for maintenance events, allowing flight upgrades in order to reduce operational and repositioning costs. Narkus-Kramer et al. (2016) examined trade-offs associated with battery-powered, remotely piloted semi-autonomous personal aircraft, and found that profitability is closely tied to high network utilizations (which result in fewer deadhead and repositioning flights) and high daily utilization (or higher average hours flown). Finally, Stouffer and Kostiuk (2020) designed a dispatching tool for UAM operations that enables a dispatcher to plan a UAM flight and check for issues before filing a flight plan.

The majority of the papers to date focused on dispatching and scheduling algorithms have presumed a deterministic framework, but two important consideration in UAM applications are that scheduling and dispatching algorithms may need to be done in real time or using a rolling horizon framework to account for delays and uncertainties, and these algorithms may need to be applied at a network level. As Thipphavong et al. (2018) noted, “due to limited energy reserves, UAM aircraft must have assurances prior to takeoff that their destination landing site will be available when they arrive. The tight coupling between arrivals and departures across the vertiports in a UAM network points to the possible need for continuous network-wide scheduling as a first-order control method for real-time, on-demand resource management.”

6.2. Review of EV and AV operations studies

EVs have been integrated into many communities throughout the world. As a result, researchers have been able to both develop and validate models using case studies. In the process, researchers have gained many insights regarding how system performance and profitability are affected by fleet size, demand, pricing, and reservation and dispatching strategies. Many of these insights are relevant to the UAM community, particularly given similarities in the directional demand patterns that both ground and UAM communities seek to serve.

From an operations perspective, developing strategies to serve one-way demand while maintaining profitability has been a particularly vexing problem for vehicle-sharing companies. Many travel patterns exhibit strong uni-directional flows, especially during peak periods. For example, in many cities morning rush hour traffic is created by commuters traveling from the suburbs into the city centers to work. Before COVID-19, airports that served predominately business travelers would see peaks of passengers traveling to the airport for Monday morning flights and peaks of passengers leaving the airport Thursday evening and/or Friday to return home. One-way demand patterns result in the need to increase the number of vehicles available to serve peak directional demand (e.g., see Hörl et al., 2019) and/or increase the need to reposition empty vehicles. Staging vehicles to serve peak demand and/or attempting to temporally or spatially shift demand to nearby pick-up and drop-off locations are some strategies that have been explored to serve one-way demand profitably (e.g., see Ströhle et al., 2019).

Within the ground transportation literature, many researchers have focused on the vehicle relocation problem, often in the context of one-way demand systems. Illgen and Höck (2018) provide a review of methods used to relocate vehicles in carsharing networks. Representative studies include those by Wang, Liu, and Ma (2019) and Wang, Yang, and Zhu (2018), who examined one-way electric carsharing systems; Warrington and Ruchti (2019), who studied Philadelphia’s public bike-sharing program; and Vasconcelos et al. (2017), who studied a carsharing service in Lisbon, Portugal, and found that relocating vehicles generated an additional 19–24 percent in profits for operators.

Given that “the cost associated with vehicle relocation operations represents a significant proportion of the total operating cost” (Boyacı and Zografos, 2019), many researchers have developed methods for better predicting demand and for tailoring operational strategies to minimize relocation costs while maintaining high service levels. Wen et al. (2019) examined dispatching polices with different types of demand information for an AV shared system and found that individual demand information from in-advance requests improves performance, but the degree of performance depends on the spatial disparity of requests. Boyacı and Zografos (2019) examined temporal and spatial flexibility regarding pick-up and drop-off of vehicles in a one-way electric carsharing system and found that spatial flexibility has a stronger effect than temporal flexibility, but both temporal and spatial flexibility can increase profitably of the system by serving more customers with fewer relocation needs. Hyland and Mahmassani (2018) compared different dispatching policies for an AV service and found that the optimal dispatching policy is a function of demand, with more sophisticated dispatching polices generating higher revenues during the peak demand period and simple dispatching policies (i.e., assigning passengers sequentially to nearest idle AV) working well in low demand periods. This result is consistent with the findings based on a case study of Zurich, Switzerland, that investigated different operational policies for an AV shared mobility system and found that operational policies had a significant impact on vehicle assignment and repositioning, heavily influencing system performance of wait times and cost (Hörl et al., 2019). Both Hyland and Mahmassani (2018) and Hörl et al. (2019) found that the utilization of intelligent demand forecasts and dispatching and rebalancing algorithms were crucial elements of profitability. In addition, Hyland and Mahmassani (2020) found that increases in the mean curbside pick-up time for a shared AV system significantly degrades operational performance in terms of user in-vehicle travel time and user wait time.

The role of advance reservations in the profitability of sharing services is nuanced. On one hand, advance reservations provide more certainty with respect to future demand and allow the operator to position vehicles in advance to the locations where customers have requested service. However, if operators take vehicles out of service too far in advance to guarantee availability for reservations, then vehicle utilization and the ability to serve on-demand requests may decrease, resulting in a less profitable system. As Molnar and Correia (2019) pointed out “while it is convenient for customers to be able to do one-way trips and drop off vehicles anywhere in a service area, this makes it difficult to offer reservations in advance” and there is a need to explore ways to increase advance reservation times by relocating vehicles to shortly before reservation pick-up times.

Several researchers have explicitly focused on the issue of advance reservations and traveler flexibility. Wu et al. (2019) examined the role of guaranteed advance reservations for a free-floating carsharing service in London and found that individuals are willing to pay £0.54 per journey ($0.75 USD)16 for a guaranteed advance reservation. Duan et al. (2020) examined a system in which individuals can either request immediate rides or reserve an AV taxi service in advance, and optimized a model that considers vehicle-to-passenger assignment with empty vehicle rebalancing. They found that when the number of vehicles is adequate and reservations are made further ahead of time, the completion rate of requests and revenue improve. Allahviranloo and Chow (2019) examined a system in which individuals can buy future time slots for AV and are guaranteed service. They found the spatial temporal distribution of demand impacts the solution to the fleet sizing problems.

Several researchers have jointly optimized fleet size and trip pricing for sharing systems. Xu, Meng, and Liu (2018) jointly optimized EV fleet size and trip pricing for a one-way carsharing service that considers vehicle relocation and personnel assignment based on a case study of Singapore. Jorge et al. (2015) used a theoretical case study network of 75 carsharing stations in Lisbon, Portugal, and found that trip pricing can increase profits through more balanced systems; optimal profits are on average 23 percent higher than base prices and serve 18 percent less demand.

Finally, several researchers have noted that the optimal operational policies and/or deployment of charging stations will evolve over time as demand increases. Ghamami, Zockaie, and Nie (2016) found that ignoring delay induced by charging congestion led to suboptimal configuration of charging infrastructure, with effects potentially more prominent as demand increased over time for PEVs. Wu and Sioshansi (2017) found challenges in planning placement of public fast-charging stations for EV due to uncertainty in future demand with initial expansion concentrated around the urban core. Dong, Ma, et al. (2019) found that as additional charging stations are built, the optimal locations start in central London and gradually expand out to suburban areas of London. Zhang, Schmöcker, et al. (2020) found when expanding one-way carsharing stations, demand growth is higher around transit hubs and public facilities than in residential areas.

6.3. Bringing it all Together—Operations insights and research directions for UAM

Based on prior research from the ground transportation literature, it is clear that system performance and profitability is driven by multiple factors, including the spatial and temporal distribution of demand, fleet size, pricing, and operational policies, and that there are strong couplings across these factors. As Repoux et al. (2019) eloquently stated, “The interaction between all parameters and settings in carsharing is complex and highly non-linear. It re-emphasizes the importance for any practitioner to identify the most effective elements (namely fleet size, station capacities, rental rules) as well as the ones specific to the system’s environment and demand.” From a UAM perspective, this highlights a critical need to jointly optimize interactions among fleet, demand, pricing, and dispatching policies. The fact that many use cases for UAM (such as commuting and trips to the airport) exhibit strong directional or one-way demand patterns will likely put further pressure on the profitability of UAM networks. One key difference between the UAM and EV/AV communities relates to the need for real-time optimization and dispatching algorithms. The penalty for running out of battery energy is much more severe in air applications than ground applications; simply stated, an eVTOL aircraft cannot run out of battery power for safety reasons. Consequently, approaches that synchronize takeoffs and landings at a vertiport in real time or under a rolling horizon framework will likely be much more critical (e.g., see Kleinbekman et al., 2020).

The experiences from the ground transportation literature with respect to the potential reduction in utilization caused by guaranteeing advance reservations is particularly relevant for the UAM community, given many customers may expect a high level of availability for their flights. Results from the ground transportation literature that find profitability can be significantly increased by rejecting demand requests is similarly problematic for UAM applications, given customer retention and wide-scale adoption will likely be strongly tied to reliability and availability of air taxis. Some strategies to increase reliability used in public transportation, such as a guaranteed ride home, may be valuable for UAM applications, e.g., if the UAM service cannot fly, the passengers would be given priority and guaranteed a ride via a ground transportation mode (like ride-hailing) for a similar or reduced price as UAM. Based on a survey of 2500 commuters in the U.S., Boddupalli et al. (2020) found that individuals were 1.8 times more likely to take an air taxi if a guaranteed ride home were provided.

All of these factors point to the trade-off between system performance and system cost—that is, we can over-design a system by ensuring extra aircraft in the fleet are available to serve peak periods and most customer demand requests, but serving all customer demand requests will likely be prohibitively costly. For example, the former CEO of NetJets, a private business jet company with fractional ownership, noted that in order to be profitable, he needed to cover 98 percent of all requested trips, and that serving 100 percent of all requested trips eliminated profits (Berger, 2001, as quoted in Mane and Crossley, 2007). Findings related to intelligent operational strategies and pricing policies that have been able to improve performance in ground transportation offer promising directions for the UAM community.

Optimizing over different time horizons will be important, particularly given the higher costs of establishing vertiports and charging stations for UAM applications than for EV applications. In addition, planning the deployment of vertiports and charging stations in ways that provide equitable access to citizens will be important if public funding is used for this infrastructure.

7. Conclusions

Research and interest in UAM have grown exponentially over the past five years, but significant questions remain with respect to whether UAM will become the next disruptive technology in urban transportation. As seen in the meta-analysis of UAM publications, much of the emphasis to date has been focused on fundamental questions. How do we design an eVTOL aircraft? How can we create more energy-dense batteries to support eVTOL missions? How do we design the airspace so that high-volume eVTOL operations can occur simultaneously with commercial and drone operations? Will there be demand for an eVTOL air taxi service and, if so, which business cases make the most sense—commuting, business shuttles to an airport, or other trip purposes? In contrast, research in EV/AV and sharing technologies for ground transportation is further along, and researchers and communities have experiences in designing and implementing EV fleets, some of which are part of ride-hailing or carsharing applications.

This paper conducted a meta-analysis of UAM, EV, and AV research published over the past five years (i.e., 2015 to 2020) to compare and contrast their research thrusts. By conducting an in-depth review of articles related to demand modeling, operations, and integration with existing infrastructure, we gleaned insights that can inform future UAM research directions.

From a demand perspective, if UAM follows trends seen in EV adoption, we would expect early adopters of UAM to more likely be male, have higher incomes, and have tech-savvy and pro-environmental attitudes; however, differences in adoption across countries is expected, with Asian countries having greater pro-technology inclinations. Importantly, the EV/AV literature has consistently found that individual preferences vary greatly and a polarization often occurs in which some individuals are enthusiastic about the new technology and willing to pay for automation and other technology features, while other individuals are negative about the new technology and state they will never adopt it. One of the reasons the EV community has focused so much research in the technology adoption area is because EV use and adoption rates have not been as high as researchers expected. The UAM community should pay particular attention to this phenomenon, as it suggests that modeling when individuals will adopt UAM will be important for demand estimations and that there is a research need to better understand how to help potential consumers feel more comfortable with the technology. Applying insights from the EV/AV area, this could include designing messages and information campaigns about the safety and limitations of UAM vehicles, and it may involve marketing campaigns that focus on recommendations from trusted family and friends. From a technology adoption perspective, it will be important to model how adoption rates for UAM evolve as AVs enter the market. Based on the theoretical and empirical results reported in the EV/AV literatures that find values of time decrease (and potentially significantly) for commute trips due to the fact individuals can be more productive in an AV compared to a conventional car, we expect that the introduction of AVs into the market will erode demand for commuter air taxis.

Our review of articles focused on infrastructure- and operations-related topics revealed strong couplings among multiple factors, including the spatial and temporal distribution of demand, fleet size, pricing, vertiport placement, vertiport topology, airspace restrictions, and operational policies. Further, many of the articles focused on one or more of these topics showed significant impacts on system performance. An important direction for the UAM research community is to develop high-fidelity simulation models that take these and potentially other factors into account. Given demand profiles today will not be reflective of demand profiles in the future (due to different adoption rates, spatial changes in populations, the introduction of competing technologies such as AVs), it will be important to conduct these simulations over different time periods to ensure results are robust over time.

It will be important for the UAM community to understand how UAM operations will impact the electric grid (and if the grid can even support UAM operations). Given insights from the EV literature that suggest the electric grid will already be stressed handling ground EV charging requirements, jointly considering EV and UAM power profiles may be important to ensure the electric grid can support both EV and UAM charging needs.

As with any analysis, there are limitations to be noted. The classification of keywords we associated with each article is arguably subjective; however, the classification enabled us to identify high-level trends across the fields. Given that researchers may be interested in identifying themes that we did not cover in this paper, we compiled a supplemental spreadsheet file, which is available online as a compendium to this paper. Our intention for this spreadsheet is to help facilitate the ability of other researchers to quickly identify keywords and/or to use the DOI links provided to more quickly identify to find papers relevant to their own research areas. Additionally, it is important to note that the publications in the AIAA database include both peer-reviewed journal publications as well as non–peer-reviewed conference proceedings.

It is important to note that our review was conducted pre-COVID-19 and that the future of transportation is at this time unclear. Some trends suggest that demand for UAM may actually increase. For example, as individuals move out of cities and into suburbs and work from home multiple days per week, they may be more interested in using an air taxi to commute to work on the days they need to travel to the office. Other trends suggest that demand for UAM may decrease. For example, if business travel decreases, the overall demand for business trips to commercial airports will decrease and fewer individuals would likely take an air taxi to the airport. It is also important to recognize that the momentum we have seen on UAM development may stall as the effects of COVID-19 continue to ripple through the industry. For example, on September 18, 2020, Boeing announced that it was suspending work at its NeXt innovation unit, which is the business division that was responsible for its UAM efforts (Gates, 2020).

In conclusion, it is our hope that both the air and ground transportation communities will find this article to be a valuable resource document, generate discussions as to potential research directions in UAM, and encourage interdisciplinary research in UAM. Never before have we attempted to fly so many air vehicles in our cities—and achieving this goal will not be a problem solved in isolation by the aerospace community.

CRediT authorship contribution statement

Laurie A. Garrow: Conceptualization, Data curation, Formal analysis, Methodology, Writing – original draft. Brian J. German: Conceptualization, Data curation, Formal analysis, Methodology, Writing – original draft. Caroline E. Leonard: Conceptualization, Data curation, Formal analysis, Visualization, Writing – review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.


The authors thank Sharon Dunn who copy edited the document prior to submission.

Appendix A. Supplementary material

The following are the Supplementary data to this article:

Download : Download spreadsheet (459KB)

Supplementary data 1.