Learning Module: *New* Shared Autonomous Vehicles


The adoption of AV technologies has the potential to impact the urban environment significantly in both positive and negative ways. It is important for cities to think through the complex interactions between vehicle autonomy and, for example, pedestrians, transit, cycling, scooters, ridesourcing, parking, local business, and public space. Streets will need to be planned, and in some cases reconfigured, to balance mobility, safety, economic prosperity, environmental responsibility, equity, and quality of life. Forward-thinking cities have an opportunity to leverage AV technological innovation to achieve public goals rather than merely allowing their communities to be overwhelmed by whatever comes next.

This learning module will focus on the impacts of AV technology on shared-use and on-demand modes, and it will consider them in the context of complex urban environments. Specifically, it will explore how AVs can be leveraged as a shared mode of transportation to both increase mobility access and reduce carbon emissions. According to the Federal Transit Administration (FTA), shared mobility includes public transportation, taxis and limos, carsharing, bikesharing, e-scooter sharing, shuttles, carpooling, ridesharing and commercial delivery vehicles.[1]

Although the technology for fully-autonomous driverless cars has not yet been realized, progress toward this goal is taking place rapidly. As a result, this case study examines what hurdles that process may encounter, and what the eventual outcomes are likely to be. It also explores some key considerations raised by the prospect of the proliferation of shared AV fleets, including what technological challenges remain, what legal frameworks will be required, and how transportation as a whole may be impacted.

Why Focus on Shared AVs?

Most AVs are electric vehicles (EVs), and EVs have lower operational costs long-term: Most AVs are EVs, meaning that they are powered by electrically-charged batteries, rather than traditional internal combustion engines (ICE). In part, this is because many features of AVs (such as electrical control systems rather than mechanical systems) are easier to implement on EVs due to EVs’ fewer moving parts.[2]  Although EVs currently cost more to purchase initially than their gas-burning counterparts, in the longer term, EVs have lower fuel and maintenance costs.[3]

Fleets tend to have higher utilization rates, making EVs (and their lower operational costs) more appealing to fleet operators: Services that require fleets of vehicles – whether to move people or products – tend to have higher utilization rates of their vehicles, relative to individual vehicle owners. This makes intuitive sense, given that fleets often need to run continuously throughout a day, whereas privately owned cars often sit idle in between morning and evening commutes. Because of these higher utilization rates, vehicles that have lower operational costs long-term – like EVs – are typically more appealing to fleet operators and transportation agencies.[4]

In particular, AVs tend to be especially appealing to fleet operators: In addition to the relatively lower total cost of ownership that EVs in general offer, electrically-powered AVs in particular offer additional benefits to fleet operators. For example, electric AVs are anticipated to improve safety, and they can provide a solution to driver shortages; it is for these reasons that commercial fleets are expected to be among the first to take advantage of AVs.[5] Furthermore, entirely autonomous technology that involves no driver will likely be limited to Operational Design Domains (ODD) such as weather conditions, geofenced areas, and designated times of day for several years as the necessary technology and infrastructure improve.[6] Due to these ODDs that will restrict AV operation, AVs are anticipated to be deployed primarily in fleets rather than as a private means of transportation when first launched broadly.

Transportation service operators are eager to explore shared modes that can sustainably transport growing numbers of people: To meet global emissions reduction goals, the transportation sector will require both a shift away from use of ICE vehicles, and a decrease of total VMT.  Because AVs are predominantly electrically-powered, their proliferation will help limit the use of traditional gas-burning engines. Furthermore, the greater use of shared modes in place of personal vehicles will serve to reduce the number of single-occupancy cars on the road and total miles driven. Without both greater electrification and adoption of shared modes of travel, the sector will not reach emission-reduction targets needed to achieve climate change goals, such as reducing global emissions by 45% from 2010 levels by 2030.[7]

UC Davis Diagram of Three Revolutions in Urban Transportation

Source: UC Davis. https://steps.ucdavis.edu/three-revolutions-landing-page/

Transportation agencies and private operators alike are interested in how shared AVs can be leveraged as a sustainable, efficient mode of travel. In addition to meeting the transportation sector’s need for more environmentally-friendly transportation and reduced VMT, fleets of shared AVs could be especially well-suited for certain types of mobility services. Currently, university and hospital campuses are some of the earliest adopters of AV fleets, because the vehicles are restricted to fairly short routes and interaction with other vehicles is limited. The Shared-Use Mobility Center is tracking and maintaining a list of these increasingly numerous AV pilot projects around the U.S. which will be published in late Spring 2020. Broader adoption of AVs for shared modes of transportation is anticipated in the coming years, which will most likely include freight and deliveries, public transit, microtransit and TNC and taxi services. This case study focuses on current efforts to adopt and regulate shared AV fleets in these areas, and it aims to serve as a resource for organizations seeking to advance this trend.



  • Autonomous vehicles (AVs): A classification system based on six different levels, ranging from fully manual (“Level 0”) to fully automated (“Level 5”) systems, was published in 2014 by the Society of Automotive Engineers (SAE) International, and since 2016 it has been the operating classification under the National Highway Traffic Safety Administration (NHTSA).[8] A “Level 1” system offers driver assistance, with features like adaptive cruise control and parallel parking guidance. A partially automated “Level 2” system is one in which the system can take control over critical functions like acceleration, breaking or steering, but with continual “hands on” supervision by a human driver. NHTSA and SAE refer to vehicles at a “Level 3” and above as having Automated Driving Systems (ADS). A conditionally automated “Level 3” system allows drivers to safely relinquish supervision of the vehicle in limited circumstances. These “eyes off” vehicles were made available to consumers with the 2018 Audi A8; the vehicle’s “Traffic Jam Pilot,” currently only enabled on vehicles sold in Germany, can self-navigate at speeds up to 37 miles per hour when on a road with a median.  Highly automated “Level 4” systems, currently being tested in some US states, allow drivers to completely relinquish navigation to the vehicle except in certain functions, like parking, or under certain inclement weather conditions. The parameters of when Level 3 or 4 ADS can or cannot operate without a human driver are called the vehicle’s Operational Design Domain (ODD), which includes factors such as road conditions, lighting, and weather. Fully automated “Level 5” systems are capable of all driving functions and could, in theory, operate without a steering wheel or any other mechanism of human control. AVs for the foreseeable future will be between Levels 2 through 4—which is to say, automated to varying degrees and limited by an ODD.  The timeline for deployment of Level 5 systems remains uncertain.
SAE Automation Levels diagram

Source: SAE 

  • Fixed-route transit: According to the FTA, fixed-route transit is a “system of transporting individuals (other than by aircraft), including the provision of designated public transportation service by public entities and the provision of transportation service by private entities, including, but not limited to, specific public transportation service, on which a vehicle is operated along a prescribed route according to a fixed schedule” [49 CFR 37.3]. This commonly refers to public transit that operates on established routes and schedules, such as buses, trains, trams and shuttles.
  • Microtransit: Microtransit is a transportation service that falls between fixed-route transit and ridesourcing operations. It resembles traditional demand-responsive transit (DRT), such as dial-a-ride, in that it often uses similar vehicles (passenger vans or cutaways) but with improved dispatching and routing, enabled by various mobile technologies, especially location access through smartphones or mobile tablets. Rides may be arranged through a smartphone app, in addition to more traditional phone- or web-based booking systems. According to the SUMC-authored TCRP Report 188, microtransit is defined as an “app-enabled private multi-passenger transportation service that serves passengers using dynamically generated routes, and may expect passengers to make their way to and from common pick-up or drop-off points.”[9] Some common microtransit vendors in the U.S. include Transloc, DemandTrans, Via, and Transdev. Increasingly, microtransit is being explored by public agencies, as well as private operators. Microtransit is an area of great interest for proponents of AVs, because AVs are commonly viewed as a mode well-suited for microtransit service. For more on microtransit in general, see SUMC’s Microtransit Learning Module.
  • Ridesourcing: Ridesourcing involves the use of online platforms to connect passengers with drivers and automate reservations, payments, and customer feedback. Typically, a transportation network company provides the online platform and manages drivers. Importantly, ridesourcing differs from ridesharing in that ridesharing involves trips where both driver and passenger(s) share the same destination: that includes travel types such as carpooling and vanpooling.
  • Vehicle miles traveled (VMT): According to the University of Texas A&M Transportation Institute, VMT “measures the amount of travel for all vehicles in a geographic region over a given period of time, typically a one-year period. It is calculated as the sum of the number of miles traveled by each vehicle.”[10] Because VMT measures travel demand, it is useful in determining where resources are most needed, and it is an important measure to monitor and forecast. For example, the FTA’s Office of Highway Patrol Information releases regular forecasts for VMT nationally. In order to meet transportation sector emission-reduction targets, total VMT will need to be reduced, and greater use of AVs as a shared mode of travel is one avenue to achieve that aim. An alternate measure, “passenger miles traveled” or PMT, takes into account that shared rides may provide more mobility access.
  • Mobility as a Service (MaaS): MaaS is a transportation model that integrates various forms of transportation – including both fixed-route and on-demand services – into comprehensive digital platforms that “integrate end-to-end trip planning, booking, electronic ticketing, and payment services across all modes of transportation, public or private”.[11] Rather than each transportation mode being viewed and experienced in its own silo, MaaS is a more user-centric paradigm that seeks to incorporate all modes into a more interconnected system that benefits customers. Future mobility platforms will need to consider ways to incorporate shared AV fleets into their models.
  • Self-driving vs. Driverless: Although the terms “self-driving” and “driverless” are often used interchangeably, self-driving usually refers to vehicles that fall in the SAE Level 3-4 range, where the vehicle is capable of handling some or most conditions, but a human driver is required to take control in certain conditions. A driverless vehicle is fully autonomous Level 5 vehicle.
  • 5G vs. dedicated short-range communication (DSRC) technology: 5G is the mobile broadband that is rapidly being rolled out by mobile carriers and cities around the globe. 5G is particularly important for AVs, as it will enable AVs to connect to other vehicles around them more readily than previous broadbands, as well as to certain 5G-compatible infrastructure, such as smart traffic lights and signals. However, there are uncertainties about 5G’s readiness, as well as the logistics of how the sim cards necessary for 5G use in vehicles will be distributed and paid for.[12] Unlike 5G which is a cellular-based technology, DSRC technology uses wifi to connect to surrounding vehicles and infrastructure. While both technologies enable vehicles to sense and react to intelligence around them, they may interfere with one another on the road, meaning that one technology is likely to be become standard down the line. Currently, some auto manufacturers (like BMW) are advocating for the widespread use of 5G, and others (like Volkswagon) are advocating for use of DSRC.


Potential Benefits of Shared AVs

  • Fewer Greenhouse Gas Emissions: Because most AVs are anticipated to be electrically-powered, their widespread adoption in place of traditional ICE vehicles could lead to a reduction in transportation-related emissions – a critical component in reaching global climate change goals. Furthermore, greater use of shared modes of transportation like shared AVs will help reduce the number of single-occupancy vehicles on the road.
  • Lower Maintenance & Operating Costs:
    • Maintenance: Because most AVs will likely be electric, they are expected to require lower maintenance costs long-term. Instead of having over 2000 moving parts that occasionally break down and need maintenance as traditional ICE vehicles do, EVs have fewer moving pieces, and therefore require less time off the road and in the shop. Furthermore, fully-autonomous vehicles will not have a driver at the wheel, thereby reducing the likelihood of a costly accident caused by human error.
    • Operating: AVs have the potential to have lower per-mile operating costs than current modes. For one, insurance costs are expected to decline significantly as AVs become more prevalent due to the anticipated decrease in accidents overall, although repair costs may be higher when accidents do occur[13]; a 2018 KPMG report estimates that total auto insurance industry losses could be reduced by 60% by 2050. The costs allocated to driver earnings will also decrease. However, agencies and companies will need to plan for the re-allocation of personnel to other functions, as well as new workforce training.
  • Safety: The USDOT estimates that 94% of all traffic accidents are caused by or linked to human error.[14] In theory, fully-autonomous, Level 5 AVs will be able to operate without any human control, thereby removing a major contributor to road accidents. Even Level 4 AVs, which are currently being tested around the country and would allow drivers to relinquish navigation to the vehicle except in certain circumstances, would likely reduce the total number of human-caused errors. When all is functioning correctly, AVs are anticipated to have faster response times than human drivers, to be constantly scanning their surroundings, and to not get distracted by things like the radio or phones. This would be expected to lead to greater safety for passengers in the AVs, in vehicles nearby, and for cyclists and pedestrians. This does not mean that the proliferation of AVs will eliminate road deaths or crashes altogether, as evidenced by the pedestrian who was killed in Arizona in 2018 by one of Uber’s self-driving test vehicles.[15]
  • Vehicle Productivity: According to the well-regarded 2011 book from Donald Shoup, private vehicles spend 95% of their time parked. Because shared AVs could potentially be used to transport people to work during rush hour, and then be reallocated to freight or delivery services during off-peak times, such vehicles would likely spend less time parked unproductively.[16]
  • Greater mobility for people with transportation difficulties: For people with transportation difficulties such as seniors and people with disabilities, AVs could serve as a safe transportation option that could enable such populations greater mobility. This could be increasingly important as the Baby Boomer generation ages and may grow less comfortable behind a wheel. One 2017 study estimates that AVs could facilitate greater job opportunities for 2 million people with disabilities.[17] However, people with mobility challenges often require accommodations specifically tailored to their transportation needs, including wheelchair accessibility and programming that serves limited visual or audio capacity. As a result, designing AV vehicles and shuttles so that they are accessible for seniors and persons with disabilities will benefit all users and avoid costly retrofitting of vehicles after the fact.  The Alliance of Automobile Manufacturers released report on findings from its workshop series on AVs and accessibility that serves as a useful resource.
  • Decreased car ownership: A 2016 report by the Florida State University Department of Urban & Regional Planning for the Florida DOT projects that private car ownership in the US is expected to decline in the coming decades, and private AV ownership is expected to follow this same pattern. Specifically, the report predicts that by 2060, 70% of AVs on the road will be privately owned – down from 90% in 2040.[18] Similarly, a 2017 report from the Rocky Mountain Institute anticipates that private car ownership rates in the US will peak in 2020, and will fall in the subsequent years, largely due to the increasing cost-advantage and availability of shared AVs.[19] Decreased car ownership overall could lead to a reduction in transportation-related emissions, fewer cars on the road and reduced incidence of collisions.
  • More green space, less road expansion: In addition to decreased car ownership, the proliferation of AV fleets could also result in a number of outcomes that urban planners deem desirable. First, because AVs will be able to drive closer together due to their faster response time and more predictable behavior, highway congestion is anticipated to lesson. This would mean a decreased need to expand existing roadways or pave new ones. Furthermore, AVs will not require wide lanes or medians to travel safely, potentially providing more space for pedestrian and cycling infrastructure. Similarly, because AV fleets can be reallocated between various purposes, they will likely spend less time parked unproductively. This will mean less space dedicated to parking, and potentially more to green space.[20] For example, a 2018 World Economic Forum report on AVs in Boston suggested that the city would require approximately half as many on- and off-street parking spots once AVs were incorporated into the modal mix.[21]
McKinsey diagram of AV deployment timeline and corresponding benefits

Source: McKinsey & Company; Projected timeline of AV deployment and corresponding benefits

Emerging Concerns & Considerations

  • Safety & Liability: According to a 2018 study by the American Automobile Association, 63% of US drivers say they would be afraid to ride in a fully self-driving car.[22] In part, that number reflects the uncertainty felt by many Americans regarding whether an AV will be truly capable of responding appropriately to all contexts it might encounter. Crashes – like that of the killing of a pedestrian by an Uber self-driving test vehicle in Arizona in 2018, or the low-speed accident between a supervised AV shuttle and delivery truck in Las Vegas in 2017 – have likely kept conversations relating to AV safety prevalent in the public discourse. Interestingly, a 2019 study found that public perception of present and future AV safety is far less optimistic in Western nations when compared with developed Asian nations.[23]  Regardless of current attitudes, there are serious questions relating to AV safety that will need to be resolved at various governmental levels before AVs can be adopted broadly. For example: 1) what road and climate conditions will require a driver to take over controls in semi-autonomous vehicles (such as Level 4), and who is responsible for regulating those conditions and adherence to them? 2) What happens when an AV‘s navigation system malfunctions, putting others at risk, and who is held liable? And 3) can AVs be effectively programmed to respond appropriately to unanticipated conditions, such as a pedestrian unexpectedly crossing a poorly-lit street, as was the case in the fatal Arizona accident? State and federal agencies are actively working to ensure that legislation addresses these and other safety-related concerns, but consensus has not been reached about “how safe is safe enough”. One particular challenge for AV technology is posed by pedestrians and cyclists, which are more difficult for sensors to pick up, and can be behave unpredictably. In fact, a 2020 report released by the Insurance Institute for Highway Safety suggests that even if all vehicles on the road were AVs, the vehicles’ more accurate perception would only be able to eliminate about one third of crashes. Furthermore, a 2019 report from the Georgia Institute of Technology concluded that every AV programming system studied was less accurate in detecting dark-skinned pedestrians compared to lighter-skinned pedestrians, raising additional questions about the safety of AVs relative to race and the size and racial make-up of the datasets from which programmers are drawing.
  • Employment Impacts: For many public transit employees, AVs can be seen as more of a potential risk than benefit. For example, if cities replace even a single bus route with an automated fleet of vehicles that do not need human supervision, that will likely mean a decrease in the number of total drivers required by the agency. Similarly, the anticipated reduction in accidents resulting from greater adoption of personal AVs could lead to a significant change in the auto insurance industry. According to an analysis by KPMG in 2015, the US personal automotive insurance sector could shrink by 40 percent within 25 years, with the number of accidents per vehicle dropping by 80 percent; similarly the Bank of England has predicted a fall in motor insurance premiums of 21–41 percent by 2040.[24] This potential reduced need for drivers and insurance providers could negatively impact the job market.
  • Publicly-funded AV infrastructure:
    • Roadways: Another concern regarding AVs is the potential dedication of public funds to support AV needs and infrastructure. For example, cities may be urged to develop AV-only lanes on already-busy highways, but they may receive pushback against using city tax dollars for such projects. This is especially true if AVs simply replace ICE vehicles among private car owners, rather than reduce car ownership altogether by shifting travelers to shared modes. If that were to occur, access to those safer, faster AV-only lanes would be limited to those who could afford AVs. The proliferation of electric vehicles suggests this could be a valid concern; in California, for example, the wealthiest cities in the state were also found to have the highest rates of EV market share.[25] Other related roadway infrastructure would also require public resources: Finland, for example, is already working to integrate smart road signs and traffic lights to accommodate AV operators, and the country is planning to repaint all continuous yellow road lines with white, which is easier for AVs to detect.[26]
    • Curb Access: Similarly, cities will likely be faced with the prospect of designating passenger zones for AV fleets to pick up and drop off passengers, but between increased home deliveries, TNC rides, and shared bike and scooter parking, the demand for curb access is already rising. AV fleets will likely add to this growing list of modes needing allocated space at the curb.
    • Remote Parking Lots: When realized, fully autonomous AVs will be able to drop passengers off at their desired destinations, and then drive elsewhere to park. Because such parking lots would not need to be located in close proximity to popular destinations, many anticipate the development of parking lots specifically designed to house AVs that will be located in more remote or less developed parts of a city.[27] However, determining an acceptable location for such parking lots may become a politically-charged battle, with more influential neighborhoods winning the NIMBY fight. There is also the potential for a significant decrease in revenue from parking taxes received by the city.
    • Charging Infrastructure: Additionally, because most AVs are also EVs, cities would also need to invest heavily in expanding their public electric charging infrastructure in order to support the needs of an electric AV shared fleet. The US is well behind many other developed nations in this regard; the International Council on Clean Transportation estimates that among the 100 largest metropolitan areas, all but 12 are predicted to face charging shortages unless AC Level 2 charging at workplaces increases by sevenfold and destination charging by threefold.[28]
  • Resulting increase in VMT: One of the major concerns regarding the widespread adoption of AVs is that car owners will simply switch from traveling via private non-automated vehicle to private automated vehicle. This would not lead to a reduction in the number of vehicles on the road, nor to a reduction in road congestion or vehicle infrastructure needed, even if vehicles could travel closer together.[29] A 2019 article published in Transport Policy argues that, due to their ability and incentive to cruise between trips rather than pay for parking, AVs will “blur[] the boundary between parking and travel”, and even traditional congestion pricing will be limited in its effectiveness in dense urban areas.[30] This possibility is particularly concerning given that VMT for delivery vehicles is already anticipated to rise to 78 billion miles annually in the US by 2040 as consumer demand for same-day or same-hour delivery expands.[31]
  • Capability of AVs to meet needs of shared fleets: Currently, many AV models are not built to meet the needs of multiple travelers in the same way that traditional shared modes are. For example, electrically-powered AVs will need to be able travel long distances without needing a recharge, which many are currently not capable of doing. Similarly, questions remain about the cost to retrofit AV models to provide more trunk space or accommodate oversized mobility devices like wheelchairs, which many shared fleets will require to meet customers’ needs.[32]
  • Data security: Cybersecurity is already a concern for non-autonomous vehicles; in 2015, there were 1.4 million recalls in the U.S. alone due to cybersecurity vulnerabilities.[33] With AVs that use cellular networks and/or wifi to connect to their surroundings, there are many more areas of connectivity vulnerable to attack, and these opportunities are likely to grow as the systems used to connect to the surrounding world (such as LiDAR, DSRC, 5G, RADAR, cameras and GPS) become more widespread on the road.[34] See the image below from the European Union Agency for Cyber Security for some of the networks with which AVs will interact.
EU Agency for Cyber Security Diagram of Connected Cars' Networks

Source: EU Agency for Cyber Security

Current State of Shared AV Pilot Programs & Partnerships

General market trends:

Although countries and cities are currently at different stages of AV deployment and development, there are some trends shared globally. As recently as 2016, many companies and individuals excited about the sector anticipated the operation of fully autonomous, driverless cars on public roads in just a few years’ time. However, the anticipated timeline seems to have been extended, as developers and programmers tackle the significant hurdles related to large-scale AV deployment. In fact, in 2019, the CEO of Ford asserted that “[w]e overestimated the arrival of autonomous vehicles”.[35] The cost of developing and testing sensors is decreasing, which may help accelerate this timeline, but the arrival of AVs operating widely on public roads remains unclear. A 2019 McKinsey assessment anticipates the mass deployment of Level 4 and 5 AVs in China in about a decade, while a 2020 report by the Victoria Transport Policy Institute anticipates Level 5 vehicles may be commercially and legally available globally by the late 2020s, but that the benefits of AV technology will only be realized when AVs are mainstream, likely in the 2050s or 2060s.[36] Because transportation is more highly regulated than some other industries, the testing and legislation necessary to deploy them broadly may slow the market penetration.

McKinsey Diagram of AV Deployment in China over time

Source: McKinsey & Company; diagram mapping estimated timeline of AV deployment in China. https://www.mckinsey.com/industries/automotive-and-assembly/our-insights/how-china-will-help-fuel-the-revolution-in-autonomous-vehicles

Generally, most experts anticipate that some of the earliest major deployments of autonomous technology on roads will involve intercity trucks that are able to be linked together along designated corridors. Testing of such projects is already underway, for example with the EU-funded ENSEMBLE Consortium planning to launch a multi-brand truck platooning demonstration by 2021 and the successful EU Truck Platooning Challenge in 2016.

Furthermore, according to a 2019 report released by the Society of Actuaries, entirely autonomous technology that involves no driver will likely be limited to ODDs such as weather conditions, geofenced areas, and designated times of day for several years to come as the necessary technology and infrastructure improve.[37]  Due to these ODDs constraining their operation, AVs are anticipated to be deployed primarily in fleets rather than as a private means of transportation.[38] The report even suggests that some states may prohibit private ownership of AVs altogether.

Select Private Sector Leaders:

An estimate released by GM’s AV unit, Cruise, in early 2020 values the entire global AV industry at $8 trillion. A similar analysis from Intel in 2017 placed the estimated value at $7 trillion. As a result, it is perhaps not surprising that companies are eager to gain a competitive advantage regarding AV technology, testing and deployment. The following list is by no means exhaustive, but it includes some of the major private-sector players pushing the needle on AV deployment.

  • Amazon: delivery robot Scout currently operating in Washington & California.
  • Apple: acquired Drive.ai in 2017; testing Lexus fleet in California. Voluntary Safety Self-Assessment submitted to NHTSA 2019 available here. Drive.ai operating autonomous shuttles in Texas.
  • Aptiv: testing in Boston, Las Vegas, Pittsburg, Singapore, among others.
  • Baidu: testing self-driving taxis in China.
  • Continental: Primarily delivery robots, testing in Germany.
  • Cruise (General Motors): testing in Las Vegas; Voluntary Safety Self-Assessment submitted to NHTSA 2019 available here. Considered one of leaders in AV development for cars.
  • EasyMile: major driverless technology vendor, testing globally.
  • Ericsson: Operating 2 AV shuttle buses in Stockholm, Sweden.
  • Ford: testing in Miami, DC and Austin; Voluntary Safety Self-Assessment submitted to NHTSA 2019 available here.
  • Huawei & Audi: Audi’s A8 was first hands-free driving enabled vehicle released; working with Huawei on developing technology for Audis’ sold in China.
  • May Mobility: Operating in Detroit, MI area and Providence, RI.
  • Mercedes-Benz, Daimler & BMW: testing in Germany, US and China; developing Level 4 technology.
  • Nuro: first company to receive autonomous vehicle exemption from NHTSA from three federal motor vehicle standards; testing grocery delivery vehicles in Houston, TX.
  • Nvidia: ADS used by several companies around the globe. Voluntary Safety Self-Assessment submitted to NHTSA 2019 available here.
  • Tesla: Developing and testing models.
  • Toyota: Toyota Research Institute-Advanced Development developing AV software; planning to offer shuttles in Japan during 2021 Olympics.
  • Volvo: Testing Level 2+ vehicles near Stockholm; collaborating with Uber, Baidu, and Nanyang Technological University and Singapore’s Land Transport Authority to launch autonomous bus; also heavily involved in truck platooning testing.
  • Volkswagon: Developed its own subsidiary, Volkswagon Autonomy (VWAT); investing in Argo AI, AV software start-up company; also working with IBM.
  • Uber: testing Volvo XC90; had been testing in Arizona, California, Texas; now limited to San Francisco area. Voluntary Safety Self-Assessment submitted to NHTSA 2018 available here.
  • Waymo (Google): Tests in Los Angeles, CA and Phoenix, AZ; first to test fully autonomous, driverless vehicles in the US. Voluntary Safety Self-Assessment submitted to NHTSA 2019 available here. Considered one of leaders in AV development for cars.

As of April 1, 2020, 19 companies had submitted their Voluntary Safety Self-Assessment to NHTSA, and all are publicly available here.

Additional Resources:


Planning for the future of Shared AVs

Key Components of AV Readiness

According to an annual report released by KPMG on AV readiness at the national level, there are several key components a country needs to have in place locally in order to be sufficiently prepared for AV proliferation.  First, in regards to national infrastructure, some of the elements that will be most important to support shared AVs are sufficient numbers of EV charging stations, overall technological infrastructure, such as mobile internet and 4G or 5G coverage, and high-quality roads. Specific roadway infrastructure that many agencies are exploring include smart traffic signals and parking and curb management.  Second, having national and local policies and legislation in place will also be critical to ensure safe deployment of AV fleets; KPMG suggests that AV regulations pertaining to safety, government-funded AV pilots, and data-sharing policies are most central. Many agencies are also establishing AV-focused departments or agencies to handle such regulation. See the following section, “Featured Policies” for some national, state- and city-level AV policies. Lastly, to be fully prepared for the mass deployment of AVs, countries need to have a well-established marketplace that facilitates innovation. Markers of such a marketplace include AV firms headquartered in the country, patents filed locally, and a high EV market share.

KPMG Diagram of forces contributing to AV evolution

Source: KPMG; Diagram of key elements contributing to AV evolution. https://www.casact.org/education/crashcourse/2018/KPMG.pdf

Steps to Ease the Transition from Traditional Types of Service to AVs

There are four types of service for which shared AVs are anticipated to be adopted most readily.

Freight & Deliveries: Urban freight is likely to be one of the first areas where AV technology will be deployed broadly. For example, AV technology company Nuro was the first company to receive autonomous vehicle exemption from NHTSA from three federal motor vehicle standards for its delivery vehicle, while truck platooning efforts are already underway across the globe. In order to support these efforts, states that have not already legalized truck platooning can ensure that their roads permit it, and they can stay abreast of the lessons learned from testing projects like the on-going EU-funded ENSEMBLE Consortium and the 2016 EU Truck Platooning Challenge. NATCO’s Blueprint for Autonomous Urbanism also encourages agencies to enable various carriers to consolidate facilities and incentivize smaller fleet vehicles. At the city level, city leaders will need to explore opportunities to better manage curb access, particularly as it relates to AV delivery vehicles. Examples of such management tools are included in the following section: “Steps to Ensure Adequate Safety and Infrastructure” under the subsection Curb Management.

Fixed-route transit: In addition to freight and deliveries, fixed-route transit is another area where AV technology could prove beneficial – if leveraged proactively. In anticipation of the development of technology that can support high-frequency, fixed-route AV buses and trains, planners should pursue opportunities to prioritize transit use over other, less sustainable modes of travel. One avenue to prioritize transit and encourage use of shared modes prior to the arrival of AVs is to explore establishing dedicated lanes and signal priority, in particular where density, congestion, and ridership are highest. In the short term, a commitment to prioritize these core transit routes may mean protecting them from competition from ridesourcing, especially where ridesourcing traffic adds congestion that could undermine transit service. Conversely, cities can consider subsidizing ridesourcing for trips with origins or destinations along eliminated or low-ridership bus routes. Such policies can build use of and familiarity with fixed-route transit service, which can later facilitate greater use of transit when it is potentially safer, faster and more environmentally-friendly by leveraging AVs.

On-demand AV microtransit: Unlike core transit routes, lower-ridership routes are likely to be better served by autonomous on-demand microtransit. For these routes – and particularly at certain times of day – smaller vehicles serving more dynamic routes will likely be more effective than larger vehicles on fixed-routes; the smaller vehicles would take up less space and be less disruptive to other traffic, and they could be platooned when appropriate. In order to prepare, cities can look for opportunities to replace unproductive bus routes with microtransit, and they can deploy incentives (such as VMT tolling) to promote shared-ride microtransit instead of single-occupancy ridesourcing. Furthermore, most ongoing AV pilot programs in the US currently are microtransit programs operating on college and hospital campuses and military bases; lessons from such pilots regarding planning and regulation are likely to emerge in the coming years that can be applied to systems serving the general public.

Ridesourcing: Cities are already wary of use of single-occupancy ridesourcing – with vehicles operated autonomously or not – replacing use of public transit. According to several reports, private vehicle ownership is anticipated to decline in the coming years, which may make ridesourcing an even more appealing form of travel. Additionally, Mobility as a Service platforms will likely blur the line between ridesourcing and microtransit, further complicating regulation efforts. However, ridesourcing leveraging AVs can be a cost-effective alternative to traditional paratransit and non-emergency medical transportation, expanding mobility for the physically impaired and the elderly in both urban and rural environments. City-level policies that could help prepare for the advancement of AVs in ridesharing include: subsidizing ridesharing trips that complement transit (by providing a first/last mile link or replacing bus service); taxing ridesharing trips that compete with transit; VMT tolling; and exploring opportunities to launch ridesharing programs for people with transportation challenges.


Steps to Ensure Adequate Safety and Infrastructure

In addition to the approaches listed above that cities can pursue to better facilitate the use of AVs among traditional types of service, there are also several policies that pertain to road infrastructure and traveler safety that cities can pursue.

Strategic routing: When planning for fixed-route transit and microtransit utilizing AVs, one of a city’s responsibilities will be to develop through routes where AVs do not compete with pedestrians. This is because AVs will likely be programmed to stop whenever a pedestrian steps in front of them, and transit vehicles may not be able to stop abruptly for pedestrians without injuring on-board passengers. For example, an incident in Columbus, Ohio in early 2020, in which a passenger on an AV traveling at low speeds was hurt when the AV had to make an emergency stop, led to the NHTSA ordering the fleet of vehicles to stop carrying passengers.[39] While this is a current concern, technology continues to evolve. It is also important to note that separating pedestrians and AVs completely would lead to a very different kind of street – one that differs starkly from the type of streets for which Complete Streets policies have been advocating.

Smart signals & sensors: Traffic signal modernization should make it increasingly possible for cities to direct and manage the flow of traffic by redistributing green time strategically, including redirecting cars away from routes where space is needed for dedicated transit lanes and toward routes designed to keep AVs moving. The combination of dedicated lanes and signal priority at intersections could create a network of high-capacity transit lines that are virtually immune to congestion. To enable AVs in particular, cities are exploring signals with communication capabilities, smart lighting systems, camera-enabled equipment on vehicles and infrastructure, and smart crosswalk detectors. The Vancouver & Surrey Smart Cities Challenge Finalist Application, submitted in March 2019, is a useful resource to see what intelligent transportation system technologies are planned to prepare for AVs and safer multi-modal corridors.[40]

Advanced travel demand modeling and planning: In addition to dynamic signal controllers that can compensate background traffic for each instance of transit signal priority, cities can also explore high-fidelity travel demand modeling that takes into account the changing dynamics of traveler needs, preferences and options. Furthermore, many cities are pursuing approaches aimed at optimizing the flow of people rather than vehicles; the Portland Bureau of Transportation’s “Central City in Motion Plan” is an example of such an approach. Barcelona’s effort to make superblocks, is another example of a city prioritizing sustainable, environmentally-friendly, multi-modal transportation over traditional automobile travel; within the city’s planned “superblocks” comprised of several city blocks together, vehicle travel is restricted or prohibited, thereby facilitating greater space for pedestrians, micromobility and parks within their boundaries, as well as less congestion and interaction with difficult-to-predict pedestrians on the roads surrounding them – key for eventual AV deployment.

Parking: Overall demand for parking is likely to decline sharply as privately-owned vehicles are replaced by fleets of vehicles that serve many people, are in motion much more and are parked much less. However, because AVs have the capacity to continue to circle in traffic rather than pay for parking, there is the potential for AVs to actually increase congestion and VMT. Cities can address this changing demand for parking through pricing incentives, some of which include:

  • the elimination of subsidized on-street parking to drive down ownership of rarely used vehicles;
  • the elimination of parking requirements for new developments until demand for the existing supply is effectively managed; and
  • the implementation of congestion pricing to discourage AVs from circling in traffic. This could involve higher prices for empty vehicles, or time-based and distance- or energy-based charges.[41]

In addition, cities should be aware that parking structures specifically designed for AVs will likely require different layouts. For example, the University of Oregon’s Urbanism Next Center anticipates that the size of both parking spaces and aisles will adjust to the changing size and shape of AVs, garages will accommodate electrical charging needs, and compact loading lines could replace the traditional layout of singular rows and aisles to increase efficiency among shared AV storage.[42]

Micromobility: Micromobility (bikes, scooters, and whatever comes next) is expected to continue to expand. To accommodate for these modes, some cities both in Europe and America have begun lowering speed limits on streets in built-up areas to 30 km/hr (19 mph). Studies indicate that doing so increases the flow of vehicles while reducing emissions and crashes.[43] In regards to AVs, these lower speed limits can help AVs interact more safely with travelers using micromobility, especially in dedicated low-speed lanes. Cities can also facilitate decreased conflict with slow lanes by prioritizing micromobility over driving and parking, and placing fixed-route transit in the center lanes.

Curb management: Due to increased rates of home deliveries, TNC rides, and shared bike and scooter parking, the demand for curb access is already rising. The advent of AVs will add to this competition, as AVs will need access to the curb in order to pick-up and drop-off passengers and deliveries safely. To address this trend, cities should explore ways to charge for curb access, and they can replace on-street parking altogether with loading zones in trip-generating areas like shopping streets, job centers, and tourist attractions. Two examples of city-level efforts to proactively address the changing demands placed on curb access include Los Angeles’ “Code the Curb” effort to build a database of information on parking regulations, signs, and curb paint throughout the city; and Washington D.C.’s program to improve management of the city’s more than 500 curbside loading zones, which resulted in a new, comprehensive plan in 2017. Other resources for curb management in the age of AVs include:

Data & Cybersecurity: One of the greatest concerns regarding the advent of AVs is how to ensure that the resulting trip and passenger data is appropriately managed. One avenue that many cities are already pursuing is to develop a single trip-planning and booking app that integrates all modes of transportation. Such an app allows the public agency managing the platform to generate a greater range of ridership data, which can be combined with advanced travel demand modeling to optimize public spending on transportation infrastructure and operations. It also eliminates complicated or incompatible payments between systems, which makes it easier for people to choose public transit options.[44]

In addition, cities need to assess what kind of data to which they have access, and to explore how AVs will influence such data. NATCO’s Blueprint for Autonomous Urbanism 2.0 defines transportation data in two categories: first, journey data, which includes information on how people or goods move between points; and second, asset data, which includes information on infrastructure like curbs, parking spaces and bus stops. Going forward, GPS, vehicle-to-everything (V2X) technology, and growing digital databases will likely expand both journey and asset transportation data available to public agencies, making it possible for cities to see in real-time whether parking spaces are in use, the speed at which traffic is moving, and other valuable information. In addition to planning for ways to gather and securely store the increasing data available to them, cities should also take steps to ensure that their transportation staff is trained to assess this data to help plan for future investments.

NATCO and the Open Transport Partnership also launched the SharedStreets data standard in 2018, which aims to provide an information-sharing platform that works for cities and companies to import previously incompatible data regarding traffic safety, traffic monitoring, and curb management, among others. Open source tools like this – as well as others, like Open Curbs – are crucial to helping public and private parties access and evaluate the growing amounts of data available to them.  The Los Angeles Mobility Data Specification (MDS) is another data standard for mobility service providers developed by the Los Angeles Department of Transportation, and dozens of transportation agencies across the country have adopted it for their application programming interfaces (APIs). The Shared-Use Mobility Center also released a white paper in 2019 evaluating the lessons learned from the FTA’s Mobility-on-Demand Sandbox program regarding ways to determine the right data-sharing approach for a shared mobility project.

State- or federal-level regulation may also prove critical to ensuring cybersecurity following the deployment of AVs. For example, the European Union recently passed its General Data Protection Regulation, and one of its many components stipulates that personal traveler data will need to be separated from the data needed to guide vehicles; such a requirement will impact how cities and private companies alike will need to plan for data gathering and storage. The European Union Agency for Cybersecurity has also released a useful guide regarding cybersecurity of connected cars, including both automated and non-automated vehicles.


Many countries and cities around the globe are already taking these and other steps to plan appropriately for AVs. At the international level, Singapore’s Centre of Excellence for Testing and Research of Autonomous Vehicles is collaborating with The Netherlands’ Organisation for Applied Scientific Research to conduct research regarding how systems should operate and ways to establish an international standard for AVs. Similarly, The Netherlands, Germany and Belgium have established a program to accelerate “truck platooning” across their shared borders. Nearby, the Law Commissions of England and Wales and Scotland are collaborating to develop a shared, comprehensive legal framework for AVs throughout the Kingdom. See the following section, Featured Policies, for some examples of specific policies already in place around the globe and in the US.


Featured Policies

International & National Policies and Programs: 

  • European Union (EU): In May 2018, the European Commission published its strategy for addressing automated mobility on roads. The communication outlines the EU’s vision for automated and connected mobility, ways to strengthen the EU’s existing technology and infrastructure to better support such mobility, opportunities to bolster the European market for AVs, and its next steps to regulate the sector. The Commission also published guidelines in 2019 for the EU approval process for AVs, including what manufactures can anticipate from regulators and how national regulations can become more harmonious. In addition to developing AV standards and policies, the Commission is also co-funding related research and infrastructure. The European Parliament has also published this assessment, which details a shared EU approach to liability rules and insurance for connected and autonomous vehicles. The EU is also facilitating the designation and implementation of 5G cross-border corridors, which will permit automated cross-border travel between nations and will allow for the study of safety, connectivity and digital technologies.
  • The Netherlands: According to KPMG’s annual ranking of AV readiness in both 2018 and 2019, the Netherlands was the nation most prepared for large-scale AV deployment in the world. One major factor in the Netherlands’ preparedness for AVs is that the country has established several policies to help move the needle on AV research and deployment. For example, the Dutch government announced in 2019 that it was developing a “driving license” for AVs that would permit the government to set standards and certify autonomous models for road use. The Dutch Vehicle Authority also has a clear process to issue an exemption for self-driving vehicles to operate on roads after companies seeking to test them demonstrate their safety and submit a formal application.[45] In regards to legislation, the Experimenteerwet zelfrijdende auto (or law governing the experimental use of self-driving vehicles) was passed in 2018 legalizing monitored unmanned AV testing on public roads – previously a driver or safety operator was required to be present in the vehicle. The Dutch, German and Belgian governments are also collaborating on “truck platooning” programs along cross-border corridors.
  • United Kingdom (UK): In 2015, the UK established its Centre for Connected and Autonomous Vehicles, part of the Departments for Transport and for Business, Energy & Industrial Strength. This Centre conducts research regarding AVs, and it provides funding for projects to develop understanding of vehicle technologies, communications, data security and infrastructure, among other research areas.[46] The Law Commission of England and Wales is also working with the Scottish Law Commission to develop a comprehensive regulatory framework for AVs; this three-year project will conclude after the third public consultation is completed in 2020 with a final legal framework recommendation report released in 2021.[47]
  • Finland: In 2017, the Finnish Ministry of Transport and Communications passed the three-phase Act on Transport Services, the objective of which was to accelerate digitalization of the transport sector and facilitate more streamlined regulations. Part of this act involved permitting the use of vehicles being controlled remotely. Finland’s revised Road Traffic Act, which will go into effect in June 2020, also serves to reduce regulatory burdens, as well as to better integrate data from existing infrastructure like roads signs and lights for AV operation. The Act also provides for the Ministry of Transport and Communications to repaint the country’s yellow dividing lanes on roads to white to better facilitate AV deployment.
  • Germany: In 2018, Germany passed ethical regulations for AVs, developed by the Ethics Commission of the Federal Ministry of Transport and Digital Infrastructure. Some components of the report include the guideline that human life should always be prioritized over property, and that driverless cars should utilize technology similar to airplanes’ “blackboxes” to determine accident causes.
  • Singapore: This city-state is one of the global leaders in city-level AV policy development. In early 2019, Singapore’s Land Transportation Agency (LTA) published Technical Reference 68, which is a set of provisional standards for AV deployment focusing on vehicle behavior, vehicle functional safety, cybersecurity, and data formats.[48] It also opened in 2017 the Centre of Excellence for Testing and Research of Autonomous Vehicles at Nanyang Technological University, which helps researchers gather data and explore the impacts of factors such as traffic lights, hills, and rain on AV travel. Some autonomous shuttles and buses are already being tested on public streets.
  • United States: At the federal level, the US National Highway Traffic Safety Administration (NHTSA), an agency of the USDOT, has taken the lead regarding research and AV guidance. In September 2016, the USDOT published the Federal Automated Vehicles Policy, and the following year, it released the Automated Driving Systems: A Vision for Safety 2.0 (ADS 2.0), which is a non regulatory approach to help auto manufacturers and operators approach deployment of AVs Level 3-5. Preparing for the Future of Transportation: Automated Vehicles 3.0 (AV 3.0) was released in 2018, which expands the scope of the ADS 2.0 report, aims to reduce policy uncertainty, and outlines a process for working with the USDOT. The agency is currently accepting public feedback on its latest report, Ensuring American Leadership in Automated Vehicle Technologies: Automated Vehicles 4.0 (AV 4.0). In addition, the federal government encourages companies to submit annual voluntary safety self-assessments, which can be found here.

NHTSA is responsible for establishing the Federal Motor Vehicle Safety Standards (FMVSS), and companies are then tasked with certifying that they meet those standards. When a passenger was injured on board an EasyMile autonomous shuttle during an emergency stop in 2020, it was the NHTSA that stepped in to temporarily prohibit all programs around the country using this vendor’s shuttles from transporting passengers. Despite this move, NHTSA’s approach to regulation has been relatively relaxed compared to other developed nations; according to the National League of Cities, the agency has embraced a “permissive environment marked by regulatory restraint and heavy trust in AV developers”.[49]

Several companies have encouraged NHTSA to revise its existing vehicle safety standards so that different standards apply to AVs that have no manual controls; currently, the FMVSS specify 73 standards that must be present on vehicles to be sold in the US, but many level 4 or 5 AVs would not comply with standards relating to steering wheels, brake pedals, and other human-centric requirements.[50] The NHTSA issued its first  autonomous vehicle exemptions to those standards in February 2020 to California-based Nuro, whose autonomous package-delivering vehicle was ruled exempt from three federal motor vehicle standards.[51]

NHTSA Table of Federal vs State Responsibilities

Source: NHTSA; current assignment of roles and responsibilities between NHTSA and states pertaining to vehicle safety standards.  https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/documents/13069a-ads2.0_090617_v9a_tag.pdf

Historically, federal agencies were largely responsible only for establishing requirements for vehicle standards pertaining to safety and equipment. As a result, much was left to the discretion of individual states regarding how to best regulate other matters, including licensing, insurance, liability, traffic law, and safety inspections. The above chart from the NHTSA’s ADS 2.0 report clarifies how the federal government has allocated these regulatory responsibilities in the past.  Some NHSTA reports have provided recommendations to states on how to best manage driver regulation and other AV-related issues, but such recommendations remain nonbinding, and there are some who believe the federal government should take a larger role in regulating AVs than they have previously in regulating non-automated travel. Congress has attempted to pass national AV legislation, but such efforts have failed to gain consensus on issues ranging from the appropriate federal role in driver management to cybersecurity concerns. An overview of such issues is detailed in the Congressional Research Service’s February 2020 report, Issues in Autonomous Vehicle Testing and Deployment

Examples of US State Level Policies & Programs:

  • Arizona: In 2015, the Governor of Arizona issued an executive order establishing the state’s Self-Driving Vehicle Oversight Committee and supporting the testing and operation of self-driving vehicles. The Governor issued a second executive order in 2018 permitting the use of AVs on Arizona roads without a person present behind the wheel. To date, more than a dozen companies have conducted AV-related tests in Arizona, including Waymo, Intel, Uber and GM, taking advantage of the supportive testing environment and clear weather. Companies seeking to test or operate AVs must first submit AV Testing Statement and Certification to the Arizona DOT, which has outlined step-by-step requirements for testing with and without human drivers.
  • California: In 2012, California became the third state (after Nevada and Florida) to legalize AVs on state roads through the SB 1298. In 2016, the state passed AB 1592, which authorized Contra Costa Transportation to launch an AV pilot project. The law included some stipulations, including that the AV operate at speeds less than 35 miles per hour. Through an Adoption of Regulations in 2018, the California Department of Motor Vehicles also established procedures and requirements to which companies need to adhere in order to conduct testing and operation on state roads. For example, the state has an AV Testing Program for testing programs with a driver, an AV Testing Program for programs without a driver, and an AV Deployment Program. A special permit is also required for companies seeking to collect revenue from AV programs.
  • Florida: In 2012, Florida passed legislation permitting AVs, making it just the second state to do so behind Nevada, and in June 2018, the Governor signed into law new legislation that was intended to make Florida an attractive testing ground for AV companies. The law permits testing of fully autonomous cars without a human safety driver present with very limited restrictions, and the law does not require an additional permit to collect revenue from passengers. The state is also investing in SunTrax, its AV testing facility near Orlando, with the second phase of construction anticipated to be completed by 2021.
  • Oregon: In April 2018, the Oregon DOT was named the state’s lead agency regarding AVs, and it established the Oregon Taskforce on AVs. The taskforce consists of law enforcement representatives, legislators, cybersecurity experts and transportation sector professionals, among others, and its objective is to develop recommendations for automated vehicle legislation. In September 2018, the taskforce submitted to the Oregon legislature its first report with its recommendations for AV-related legislation. This report explored matters pertaining to licensing and registration, law enforcement and crash reporting, cybersecurity, and insurance and liability. In September of 2019, the taskforce submitted its second report, which focused on land use, road and infrastructure design, public transit, workforce changes, and state responsibilities relating to cybersecurity and privacy.
  • Washington: In May 2017, the Governor of Washington signed an executive order encouraging AVs on public roads. Companies seeking to test AVs in the state henceforth needed certification from the Department of Licensing. However, there has been some criticism that the Department’s requirements to allow testing are insufficient, and that there is no data sharing requirement, meaning the Department does not have information on how many cars are being tested at any given time among those currently permitted.[52] In 2018, the state also established the Washington State AV Working Group, which aims to prepare legislation for AV deployment. The group’s 2019 annual report is available here. More generally, Washington state’s Commute Trip Reduction Law – initially passed in 1991, with various updates in subsequent years – sets certain standards for how employers and employees must work together to reduce the number and length of drive-alone commute trips made to their worksite. The policy defines single-occupancy trips to include certain types of ridehailing, and promotes the most effective trip-reduction strategies. Such approaches will be extremely helpful to ensure that the adoption of AVs increases mobility and shared rides, rather than facilitates greater single-occupancy travel and greater VMT.
  • Additional resources for decision makers and planners at the state level:

Considerations at the City-Level

Cities have a wide range of authority to enact laws that might impact AV operations and deployment. For example, some cities may choose to prohibit AVs in school zones, or require certain data sharing practices from operators. One issue that will likely be of increasing concern to cities is the extent to which state and federal laws might preempt such authority. Just as some cities have experienced with the proliferation of TNCs and related state-level legislation, there are some AV-related matters that states may regulate, thereby prohibiting cities from regulating those matters as they might otherwise choose. Cities should be proactive in working with state and federal legislators to ensure their authority to manage AV operations to meet uniquely urban contexts is appropriately protected.

Additional resources for decision makers and planners at the city-level:



The widespread adoption of AV technology holds the potential to significantly change the way humans travel. However, to ensure those changes are beneficial, public agencies and companies alike must consider how incorporating AVs into the modal mix may impact pedestrians, cyclists, parking, businesses, public space, the environment, quality of life, and much more. Agencies will need to stay abreast of best practices emerging from the many on-going pilot programs around the globe, and remain proactive in ensuring that AV-related legislation supports their long-term aims.



[1] Federal Transit Administration (2020 February 28). Shared Mobility Definitions. https://www.transit.dot.gov/regulations-and-guidance/shared-mobility-definitions

[2] McCauley, Ryan (2017 March 15). Government Technology: “Why Autonomous and Electric Vehicles are Inextricably Linked.” https://www.govtech.com/fs/Why-Autonomous-and-Electric-Vehicles-are-Inextricably-Linked.html

[3] Kamiya, George, Teter, Jacob (2019 March 28). International Energy Agency: “Shared, autonomous…and electric?” https://www.iea.org/commentaries/shared-automated-and-electric

[4] Kamiya, George, Teter, Jacob (2019 March 28). International Energy Agency: “Shared, autonomous…and electric?”  https://www.iea.org/commentaries/shared-automated-and-electric

[5] Fleet Europe (2019 February 19). “How AVs will revolutionise fleet and OEM revenues by 2030.” https://www.fleeteurope.com/en/autonomous/europe/analysis/how-avs-will-revolutionise-fleet-and-oem-revenues-2030?a=JMA06&t%5B0%5D=KPMG&t%5B1%5D=Autonomous&t%5B2%5D=EVs&curl=1

[6] Mudge, Richard, Kornhauser, Alain (2019 October). Society of Actuaries: An Update on the Outlook for Automated Vehicle Systems. https://www.soa.org/globalassets/assets/files/resources/research-report/2019/automated-vehicle-update.pdf

[7] Intergovernmental Panel on Climate Change (2018 October 8). “Summary for Policymakers of IPCC Special Report on Global Warming of 1.5°C approved by governments.” https://www.ipcc.ch/2018/10/08/summary-for-policymakers-of-ipcc-special-report-on-global-warming-of-1-5c-approved-by-governments/

[8] US Department of Transportation (USDOT) (2018 October). Preparing for the Future of Transportation: Automated Vehicles 3.0. https://www.transportation.gov/sites/dot.gov/files/docs/policy-initiatives/automated-vehicles/320711/preparing-future-transportation-automated-vehicle-30.pdf

[9] Transportation Cooperative Research Program (2017 May 3). Transportation Research Board TCRP Report 188: Shared Mobility and the Transformation of Public Transit. http://www.trb.org/TCRP/Blurbs/174653.aspx

[10] Texas A&M Transportation Institute (2016). Methodologies Used to Estimate and Forecast Vehicle Miles Traveled. https://static.tti.tamu.edu/tti.tamu.edu/documents/PRC-2016-2.pdf

[11] Goodall, Warwick, Fishman, Tiffany Dovey, Bornstein, Justine, Bonthron Brett (2017). Deloitte Review: “The Rise of mobility as a service.” https://www2.deloitte.com/content/dam/Deloitte/nl/Documents/consumer-business/deloitte-nl-cb-ths-rise-of-mobility-as-a-service.pdf

[12] Yoshida, Junko (2019 October 25). EE Asia Times: “The DSRC vs 5G Debate Continues”. https://www.eetasia.com/news/article/The-DSRC-vs-5G-Debate-Continues

[13] McMahon, Jeff (2017 April 17). Forbes: “Big Fuel Savings from Autonomous Vehicles.” https://www.forbes.com/sites/jeffmcmahon/2017/04/17/big-fuel-savings-from-autonomous-vehicles/#59874ff44390

[14] US Department of Transportation (2017 October 6). Press Release: USDOT Releases 2016 Fatal Traffic Crash Data. https://www.transportation.gov/briefing-room/nhtsa0117

[15] Keeling, Brock (2018 March 22). Curbed San Francisco: “Uber self-driving car kills a pedestrian in Arizona [Update].” https://sf.curbed.com/2018/3/19/17139856/uber-self-driving-car-kills-pedestrian-person-death

[16] Medium (2018 August 24). “Transport Network Impacts of AVs and Emerging Technologies.” https://medium.com/predict/transport-network-impacts-of-avs-and-emerging-technologies-94cdc1c11c6

[17] Claypool, Henry, Bin-Nun, Amital, Gerlack, Jeffrey (2017 January). The Ruderman Family Foundation: Self-Driving Cars: The Impact on People with Disabilities. https://rudermanfoundation.org/white_papers/self-driving-cars-the-impact-on-people-with-disabilities/

[18] Chapin, Tim, Stevens, Lindsay, Crute, Jeremy, Crandall, Jordan, Rokyta, Anne, Washington, Alexandria (2016 April). Florida Department of Transportation: Envisioning Florida’s Future: Transportation and Land Use in an Automated Vehicle World, page 3. https://fpdl.coss.fsu.edu/sites/g/files/imported/storage/original/application/abfcc477779d0bc0ea825c8011011939.pdf.

[19] Johnson, Charlie, Walker, Jonathan (2016). Rocky Mountain Institute: Peak Car Ownership Report. https://rmi.org/insight/peak-car-ownership-report

[20] Sisson, Patrick (2016 February 25). Curbed: “How Driverless Cars Can Reshape our Cities”. https://www.curbed.com/2016/2/25/11114222/how-driverless-cars-can-reshape-our-cities

[21] World Economic Forum (2018 June). Reshaping Urban Mobility with Autonomous Vehicles: Lessons from the City of Boston, page 4. http://www3.weforum.org/docs/WEF_Reshaping_Urban_Mobility_with_Autonomous_Vehicles_2018.pdf

[22] Naughton, Keith (2018 January 24). Bloomberg News: “American Drivers Begin Warming Up to Riding in Self-Driving Cars.” https://www.bloomberg.com/news/articles/2018-01-24/american-drivers-begin-warming-up-to-riding-in-self-driving-cars

[23] Moody, Joanna, Bailey, Nate, Zhao, Jinhua (2019). MIT Department of Urban Studies & Planning: Public Perceptions of Autonomous Vehicle Safety. http://dusp.mit.edu/publication/public-perceptions-autonomous-vehicle-safety-international-comparison

[24] KPMG International (2019). 2019 Autonomous Vehicles Readiness Index. https://assets.kpmg/content/dam/kpmg/xx/pdf/2019/02/2019-autonomous-vehicles-readiness-index.pdf

[25] Nutsey, Nic (2018 May). The International Council on Clean Transportation: California’s continued electric vehicle market development, page 4. https://theicct.org/sites/default/files/publications/CA-cityEV-Briefing-20180507.pdf

[26] KPMG International (2019). 2019 Autonomous Vehicles Readiness Index. https://assets.kpmg/content/dam/kpmg/xx/pdf/2019/02/2019-autonomous-vehicles-readiness-index.pdf

[27] Sisson, Patrick (2016 August 8). Curbed: “Why high-tech parking lots for autonomous cars may change urban planning.” https://www.curbed.com/2016/8/8/12404658/autonomous-car-future-parking-lot-driverless-urban-planning

[28] Nicholas, Michael, Hall, Dale, Lutsey, Nic (2019 January). The International Council on Clean Transportation: Quantifying the Electric Vehicle Charging Infrastructure Gap Across U.S. Markets. https://theicct.org/sites/default/files/publications/US_charging_Gap_20190124.pdf

[29] World Economic Forum (2018 June). Reshaping Urban Mobility with Autonomous Vehicles: Lessons from the City of Boston, page 4. http://www3.weforum.org/docs/WEF_Reshaping_Urban_Mobility_with_Autonomous_Vehicles_2018.pdf

[30] Millard-Ball, Adam (2019 March). Transport Policy: The autonomous vehicle parking problem, 75(1). Pages 99-108. https://www.sciencedirect.com/science/article/abs/pii/S0967070X18305924

[31] KPMG International (2019). 2019 Autonomous Vehicles Readiness Index. Page 6. https://assets.kpmg/content/dam/kpmg/xx/pdf/2019/02/2019-autonomous-vehicles-readiness-index.pdf

[32] Kamiya, George, Teter, Jacob (2019 March 28). International Energy Agency: “Shared, autonomous…and electric?” https://www.iea.org/commentaries/shared-automated-and-electric

[33] Garza, Harold (nd). Southwest Research Institute: Security Considerations for Connected Autonomous Vehicles.  https://www.sans.org/cyber-security-summit/archives/file/summit-archive-1525720275.pdf

[34] Ippolito, Pier Paolo (2020 January 2). Medium Towards Data Science: “Future of Cyber Security for Connected and Autonomous Vehicles.” https://towardsdatascience.com/future-of-cyber-security-for-connected-and-autonomous-vehicles-4c553def6d50

[35] Naughton, Keith (2019 April 9). Bloomberg News: “Ford CEO Tamps Down Expectations for First Autonomous Vehicles.” https://www.bloomberg.com/news/articles/2019-04-09/ford-ceo-tamps-down-expectations-for-first-autonomous-vehicles

[36] Litman, Todd (2020 March 24). Victoria Transport Policy Institute: Autonomous Vehicle Implementation Predictions. https://www.vtpi.org/avip.pdf

[37] Mudge, Richard, Kornhauser, Alain (2019 October). Society of Actuaries: An Update on the Outlook for Automated Vehicle Systems. https://www.soa.org/globalassets/assets/files/resources/research-report/2019/automated-vehicle-update.pdf

[38] Mudge, Richard, Kornhauser, Alain (2019 October). Society of Actuaries: An Update on the Outlook for Automated Vehicle Systems. https://www.soa.org/globalassets/assets/files/resources/research-report/2019/automated-vehicle-update.pdf

[39] Marshall, Aarian (2020 February 29). Wired: “The Feds Ban a Self-Driving Shuttle Fleet From Carrying People.” https://www.wired.com/story/feds-ban-self-driving-shuttle-fleet-carrying-people/?bxid=5cec24fbfc942d3ada068985&cndid=57041884&esrc=DailyNLPromo?utm_ter&source=EDT_WIR_NEWSLETTER_0_TRANSPORTATION_ZZ&utm_brand=wired&utm_campaign=aud-dev&utm_mailing=WIR_Transportation_030220&utm_medium=email&utm_source=nl&utm_term=WIR_Transportation

[40] City of Vancouver and City of Surrey (2019 March 5). Smarter Cities Challenge Finalize Application. https://www.smartertogether.ca/wp-content/uploads/2019/03/smart-cities-challenge-smarter-together-finalist-proposal.pdf

[41] Millard-Ball, Adam (2019 March). Transport Policy: The autonomous vehicle parking problem, 75(1). Pages 99-108. https://www.sciencedirect.com/science/article/abs/pii/S0967070X18305924

[42] Whitney, Jenna (2018 May 9). University of Oregon Urbanism Next: Changing Parking Infrastructure with Autonomous Vehicles. https://urbanismnext.uoregon.edu/category/inputs/ownership/

[43] Small, Andrew (2019 August 8). CityLab: “Why Speed Kills Cities.” https://www.citylab.com/transportation/2019/08/low-speed-limit-vehicle-safety-crash-data-traffic-congestion/588412/

[44] National Association of City Transportation Officials (2019 September). Blueprint for Autonomous Urbanism: Second Edition. https://nacto.org/publication/bau2/transit/

[45] Government of Netherlands (nd). Self-driving vehicles. https://www.government.nl/topics/mobility-public-transport-and-road-safety/self-driving-vehicles

[46] Centre for Connected & Autonomous Vehicles (2018). UK Connected & Autonomous Vehicle Research & Develoment Projects 2018. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/737778/ccav-research-and-development-projects.pdf

[47] Law Commission of the United Kingdom (nd). Automated Vehicles. https://www.lawcom.gov.uk/project/automated-vehicles/

[48] Land Transport Authority of Singapore (2019 January 31). Joint Media Release by the Land Transport Authority (LTA), Enterprise Singapore, Standards Development Organisation & Singapore Standards Council: Singapore Develops Provisional National Standards to Guide Development of Fully Autonomous Vehicles. https://www.lta.gov.sg/content/ltagov/en/newsroom/2019/1/2/joint-media-release-by-the-land-transport-authority-lta-enterprise-singapore-standards-development-organisation-singapo.html

[49] Perkins, Lucy, Dupuis, Nicole, Rainwater, Brooks (2018). National League of Cities: Autonomous Vehicle Pilots Across America. Page 6. https://www.nlc.org/sites/default/files/2018-10/AV%20MAG%20Web.pdf

[50] Fraade-Blanar, Laura, Kalra, Nidhi (2017). RAND Corporation: Autonomous Vehicles and Federal Safety Standards: An Exemption to the Rule? https://www.rand.org/content/dam/rand/pubs/perspectives/PE200/PE258/RAND_PE258.pdf

[51] Canis, Bill (2020 February 11). Congressional Research Service: Issues in Autonomous Vehicle Testing and Deployment. https://crsreports.congress.gov/product/pdf/R/R45985

[52] Drew, James (2019 September 23). Government Technology: “Washington State in the Dark Over Public AV Testing.” https://www.govtech.com/fs/automation/Washington-State-in-the-Dark-Over-Public-AV-Testing.html

Related Items