Excerpts from Bloomberg-McKinsey on preparing for AVs (2017)

New Markets

AVs could unleash unprecedented new demand for automobile travel among three long-underserved groups: non-drivers, the elderly, and the disabled. For many people, the leap from partial to fully autonomous vehicles will open doors to vastly increased mobility. The most detailed look at how these groups (FIG. 12) could shape future demand for AV-based travel, a 2016 Carnegie Mellon University (CMU) study, holds some startling projections. If these three groups take to AVs en masse, they could boost overall vehicle travel (in the United States) by as much as 14 percent. Add in children under 18, a tech-savvy market that mobility service startups are already tapping, and this figure could rise even higher.

Senior Citizens
Persons age 65 and older are a second new source of demand for travel by AV. Focusing in just on the nondisabled, drving elderly, the CMU study calculated a smaller but still significant impact of about 46 billion additional VMT annually. The elderly population in most developed countries with cities participating in the Bloomberg Aspen Initiative on Cities and Autonomous Vehicles will experience an historic expansion during exactly the same
period (2020–2040) as the AV transition shifts into high gear. (FIG. 14)

Many adults never learn to drive, even in highly auto-dependent nations like the United States, where licensing rates actually started declining before the advent of AVs. (FIG. 13) But if adult non-drivers (age 19 and over) travel by AV as much as drivers do with cars today, they could contribute some 196 billion additional annual vehicle miles traveled (VMT) in the United States, about a 4 percent overall increase.

The Disabled
A third new market for AV-based mobility is adults over the age of 19 with temporary or permanent medical conditions or disabilities that make it difficult to travel outside the home. The CMU study found that this group, while much smaller than the healthy elderly population, would have an even greater impact on travel demand, more than 55 billion VMT annually. 

Gradual Upgrades for Government Vehicles
Commercial and government fleets will be automated at very different rates. At first, the impact on cities will be limited to urban expressways and industrial zones. But over time, the footprint of automated truck fleets will expand into more congested parts of cities. Taxis are also on a fast track to full
automation, and this will have a much greater impact on employment and city streets. Analysts at Barclays estimate that by eliminating the cost of taxi drivers’ labor, full automation could slash uberPOOL’s current fares for shared rides ($1.00 to $1.50 per mile) to just eight cents per mile. If even a small portion of these fare reductions is achieved in practice, use of shared AV taxis could grow rapidly.

Public transit has a long history of automation, where it has been pursued 26,000 miles annually and are replaced every 3 or 4 years. (FIG. 19) In comparison, the average private car in the United States is more than ten years old. But government fleets are the slowest of all to be replaced, as they are subject to much longer working lifetimes and are replaced subject to planning and capital investment cycles. The US Postal Service, for instance, is currently testing prototypes to replace some 142,000 long-life vehicles first put into service in 1987, some three decades ago. A variety of purpose-built AVs could take over city streets long before private cars even hit the market.

Commercial and government vehicles are a major presence on city streets, accounting for more than 25 percent of traffic. (FIG. 18) Building on decades of investment in navigation, communication, safety, and diagnostics technology, the road is clear for widespread conversion of urban vehicle fleets to autonomous driving in the coming years.

Trucks, Taxis, and Transit 
Long-haul trucks will likely be the first widely automated fleets. Even without the elimination of labor costs, the case for automated trucking is being made on fuel savings alone. Truck platooning, a technology for closely-spaced,

Modernize plans for expressways, pivoting from expansion to modernization and management. 

Transportation planning at all levels should refocus on modernizing existing expressways with instrumentation for new technology.

Develop and implement robust datasharing requirements for new vehicle technology.

Support safer, more efficient, environmentally sustainable freight systems.

Include transportation professionals from cities. Future visioning for automated vehicles should begin from the inside out, from the centers of our economy, looking at land use as well as transportation.

Policies at the Federal and State levels for infrastructure funding must be revised to reflect the restructuring of the transportation system

Safety Plan for fully automated operation (NHTSA Level 4) to support Vision Zero. Regulators and product designers should bar the use of partially automated vehicles (NHTSA Level 3) on any roadway without controlled
access, like city streets. Maximum operating speed in a city street environment should not exceed 25 miles per hour.

Federally and state supported research on automated vehicles should focus on city street operations of shared, automated, electric vehicles.

Increased Federal and State funding for city operation of automated vehicles. Research should address any needs for on-street infrastructure in the city environment and how to cover those costs.

The future of transit vehicles and their unique needs in terms of automation should be investigated to ensure transit can benefit from advances in technology. Adjust and standardize lower travel time costs beginning with model year 2020.

Beginning as soon as model year 2020, per-minute travel time costs could be an estimated 80 percent lower.  To support this change in modeling, a metropolitan modelling exercise for North America similar to the Lisbon
model released by the International Transport Forum in 2015 would be beneficial in understanding how this shift in transportation costs may affect overall travel patterns.

Automakers are bouncing back quickly from the tech world’s sneak attack. But can the industry use the AV transition to overcome its anti-urban legacy?

In the 20th century, car companies worked for decades to turn the public against cities and public transportation. Those mistakes are mostly in the past, and many automakers are trying to find a new role in sustainable urban
mobility that’s less focused on selling cars to consumers.

AVs are a key to making this shift. Reeling from the tech industry’s media blitz in 2015 and 2016, automakers have retooled and expanded long-simmering AV research and development programs. In the coming decade, these
programs are expected to push a wide variety of partially and fully autonomous vehicles to market.

Existing car companies will have a builtin scale advantage to help fend off the tech industry’s challenge. Today, Volkswagen and Toyota each produce nearly 10 million passenger cars each year. Tesla would have to double its meager production—just over 75,000 cars in 2016—every year for a decade to reach an equivalent level of production. (FIG. 26)

A Flurry of Investments

Big automakers have gone on a spending spree to boost their tech credentials. A growing flurry of investments, acquisitions, and alliances has gained momentum in the last year. (FIG. 27) US automakers are most exposed to the technology giants’ advances. General Motors invested $500 million in Lyft in 2016, a deal the automaker described as an “alliance.” The move sets up GM as the ride-share pioneer’s future supplier of AV taxis. Ford, an early
investor in Lyft, doubled the size of its Silicon Valley office.

Outside the United States, European and Asian carmakers are also heavily investing in AVs. Volvo, will begin the world’s first full-scale consumer pilot in Gothenburg in 2017. Fiat Chrysler flirted with Apple before partnering
with Google in the Waymo spin-off.

Japan: A Sleeping Giant Awakes in 2020? 

Japanese automakers have lagged badly behind their US and European counterparts in the development of AV technology. According to Bloomberg Technology, most do not expect to market AVs until 2025, considerably later
than US and European counterparts. The 2020 Toyko Summer Games could be a landmark event for urban AVs, situated in the world’s largest megacity in the most robot-friendly country. Japan has long been the world leader in robotics and industrial automation—with more industrial robots in use than the US and Germany combined. The Japanese government is pushing hard to exploit the opportunity, hoping to pull off a major demonstration of AV taxis to transport athletes. The event could prove catalytic for Japanese industry and have impacts that shape the urban AV market as profoundly as the Toyota Prius and Nissan Leaf did for hybrid and electric vehicles.

Heavy Merge Ahead

With a decade or more before AV sales really take off, automakers still have time to retool. Signs point towards a more symbiotic relationship between tech giants and AVs could become so cheap and convenient that they compete head-on with walking and cycling for short-distance trips

Major Simulation Studies All Point to Big Benefits from Shared AVs

New York City 2017 MIT, Computer Science and Artificial Intelligence Laboratory:  In a simulation based on 3 million yellow cab trip logs, researchers found 3,000 shared AVs carrying up to 4 passengers could meet 98% of existing taxi demand within an average wait time of 2.8 minutes and an average additional trip delay of 3.5 minutes.

Lisbon 2015 OECD, International Transport Forum With existing transit left in place, a shared AV taxi system would result in just a 6 percent increase in total miles travelled, but eliminate 90 percent of vehicles citywide. Reclaimed roads and parking could provide significant environmental benefits.  Depends on complete shift from private to fleet vehicles and efficiencies from common dispatch for all.

Austin 2015 University of Texas, Center for Transportation Research. This advanced simulation (based on the leading MATSim agent-based engine) focused on a shared AV network in a 12-mile by 24-mile area in Austin’s regional core, finding that private ownership could be replaced with a fleet just 10 percent the size. An 8 percent increase in vehicle travel, however, would result from repositioning of AVs.

Singapore 2014 Singapore-MIT Alliance in Research and Technology.  This study modeled a city-wide shared AV taxi system replacing all private autos and public transit, and found that all current mobility needs could be fulfilled with one third the number of vehicles currently operating in Singapore.

New York City 2013 Columbia University, The Earth Institute.  Using aggregate taxi system data, this model concluded that a conventional non-shared AV taxi system could provide on-demand rides at a cost of 90 percent below current fares, reduce taxi fleet size from more than 13,000 to 9,000 vehicles, and cut wait times from an average of 5 minutes to under 1 minute.

Engine of Growth: Could AVs also create jobs?

The question of direct employment related to the rollout of AVs has been almost entirely neglected. However, Toronto’s 2015 study identified three sectors that could experience potential employment gains of up to 15
percent—construction related to conversion of parking facilities, expansion of highways and roads, and IT products and services directly related to AV rollout.

Indirect positive impacts on employment and overall economic health seem likely, however, as businesses will see the same dramatic drops in transport costs as consumers. AVs will enable the creation of new consumer
products and services, contributing to economic growth. And AVs are likely to allow businesses to restructure and reorganize, increasing productivity.

Shifting Fortunes for Transit, as Paratransit is Eating Transit System Budgets

AVs could provide new ways to manage the fast-growing costs of demand-responsive services (or paratransit), which serve those unable to use conventional buses and trains. These systems are costly to operate, while fares are fixed, resulting in large operating deficits. (FIG. 38) Some cities are partnering with ridesource providers to serve these riders— only about 25 percent of which require specially equipped vehicles to handle wheelchairs. There is considerable resistance among disabled advocates.

Funding and finance

It is unclear how AVs would affect funding like property taxes and long-term debt. But since fuel taxes are mostly collected by higher levels of government, cities may find themselves hard-pressed to hold on to these new funds.  Recently, analysts at McKinsey modeled the impacts of various AV deployment scenarios in London, New York, Los Angeles, and Delhi, forecasting 20 to 65 percent drops in energy
tax revenues by 2030. (FIG. 37)

To more fairly price access to urban streets, the use of congestion tolls has spread to many cities in recent years. AVs provide a critical opportunity where barriers to introducing these fees more widely may be significantly reduced. Research shows that motorists are much less sensitive to electronic toll payment than cash. Electrification will demand another innovation, distance-based road charges, to replace fuel tax revenues. Properly used, these revenues could directly subsidize transit and shared AV systems (which might be exempted). This could reduce the burden on other sources of mitigate these factors, and they could actually
increase costs—as drivers are eliminated, personnel would still be required to assist with boarding and medical emergencies.

Long-term Issues

The long-term fiscal impacts of widespread use of AVs are highly uncertain. Property tax revenues could expand if AVs, especially shared taxis, unlock large amounts of land that can be rebuilt at higher densities. Selfdriving
sprawl on the other hand, could exacerbate capital flight and further erode the tax base of less desirable districts. Finally, long-term financing tools like revenue bonds would need to be restructured as traffic volumes shift. But AV-road pricing could help expand and drive innovation in these approaches.

The transition to AVs threatens city revenue streams, such as parking. But it will also allow the creation of targeted taxes and fees that more effectively advance policy goals. The elimination of driving won’t mean big cutbacks in the city government workforce.

Most cities employ a relatively small number of people whose sole function is driving. But AVs will still have widespread financial impacts.

Pricing Roads 

The biggest financial opportunity AVs present cities is in restructuring how road systems are financed. In most countries, motorists pay high taxes on fuel but almost nothing for actual use of roads. This scheme is breaking down all over the world as roads become more congested and electrification looms.

FIG. 37 Potential Impact of Vehicle Electrification on Energy Tax Revenues Source: McKinsey



Bloomberg New Energy Finance, and McKinsey & Company (2016). “An Integrated Perspective on the Future of Mobility.” http://bit.ly/FutureMob_B.

Boston Consulting Group (2015). “Revolution in the Driver’s Seat: The Road to Autonomous Vehicles.” http://bit.ly/RevDriver

Brookings Institution (2016). “Moving Forward: Self-Driving Vehicles in China, Europe, Japan, Korea, and the United States.” By Darrell M. West. http://bit.ly/MvgFwd

Computer History Museum (2014). “Where to? A History of Autonomous Vehicles.” By Marc Weber. http://bit.ly/AVhistory

The Earth Institute, Columbia University (2013). “Transforming Personal Mobility.” By Lawrence D. Burns, William C. Jordan, and Bonnie A. Scarborough. http://bit.ly/EarthInst

European Parliament, Directorate-General for Internal Policies of the Union (2016). Research for TRAN Committee: Self-Piloted Cars: The Future of Road Transport? http://bit.ly/SelfPilot

Greenblatt, Jeffery B., and Samveg Saxena (2015). “Autonomous Taxis Could Greatly Reduce Greenhouse-Gas Emissions of US Light-Duty Vehicles.” Nature Climate Change 5 (9): 860–63.

Guerra, Erick (2016). “Planning for Cars That Drive Themselves: Metropolitan Planning Organizations, Regional Transportation Plans, and Autonomous Vehicles.” Journal of Planning Education and Research 36 (2): 210–24.  http://bit.ly/CarsDri

Harper, Corey D., Chris T. Hendrickson, and Constantine Samaras (2016). “Cost and Benefit Estimates of Partially-Automated Vehicle Collision Avoidance Technologies.” Accident Analysis & Prevention 95: 104–15.

IHS Automotive (2016). “Autonomous vehicle sales forecast to reach 21 mil. globally in 2035, according to IHS Automotive.” http://bit.ly/IHSforecast

Lipson, Hod and Melba Kurman (2016). Driverless: Intelligent Cars and the Road Ahead. (Cambridge, Massachusetts: The MIT Press) London School of Economics (2016). “Autonomous Vehicles: Negotiating a
Place on the Road.” By Chris Tennant, et. al. http://bit.ly/LSE_avs

City of Los Angeles (2016). Urban Mobility in A Digital Age: A Transportation Technology Strategy for Los Angeles. http://www.urbanmobilityla.com/strategy

Maurer, Markus J. et. al., eds. (2016). Autonomous Driving: Technical, Legal and Social Aspects. (Berlin Heidelberg: Springer Open)

Mitchell, William J., Chris Borroni-Bird, and Lawrence D. Burns (2010). Reinventing the Automobile: Personal Urban Mobility for the 21st Century. Cambridge, Mass: MIT. Morgan Stanley (2016). “Shared Mobility on the Road of the Future.” http://mgstn.ly/2k3c6v6

National Association of City Transportation Officials (2016). NACTO Policy Statement on Automated Vehicles. http://bit.ly/NactoAV

National Center for Transit Research (2016). “Evaluation of Automated Vehicle Technology for Transit: 2016 Update.” By Brian Pessaro. http://bit.ly/NCTR_avs.

National Highway Traffic Safety Administration (2016). “Federal Automated Vehicles Policy: Accelerating the Next Revolution in Roadway Safety.” http://bit.ly/NHSTA_avs

National League of Cities (2015). “City of the Future: Technology & Mobility.” By Nicole DuPuis, Cooper Martin, and Brooks Rainwater. http://bit.ly/NLCavs

New York University Rudin Center for Transportation Policy and Management (2014). “Re-Programming Mobility: The Digital Transformation of Transportation in the United States.” NYU Wagner Rudin Center.

OECD International Transport Forum (2015). “Urban Mobility System Upgrade: How Shared Self-Driving Cars Could Change City Traffic.” http://bit.ly/ITF_avs

RAND Corporation (2016). Autonomous Vehicle Technology: A Guide for Policymakers. By James Anderson et al. http://bit.ly/RANDavs

Rocky Mountain Institute (2016). “Peak Car Ownership: The Market Opportunity of Electric Automated Mobility Services.” By Charlie Johnson and Jonathan Walker. http://bit.ly/RMI_avs.

This horizon scan draws on a wide variety of publications by academic researchers, government agencies, industry analysts, and policy and planning think tanks. The diversity of topics, forecasts, and insights they contain reflects the inherent scope and uncertainty of the AV transition.

Ruderman Family Foundation and Securing America’s Future Energy (2017). “Self-Driving Cars: The Impact on People with Disabilities.” By Henry Claypool, Amitai Bin-Nun, Ph.D., and Jeffery Gerlach. http://bit.ly/RFF_AVs.

Singapore University of Technology and Design (2016). On the Move in 2040: A Foresight Study on Urban Mobility in Singapore. http://bit.ly/singapore2040

Smith, Bryant Walker (2016). “How Governments Can Promote Automated Driving.” SSRN Electronic Journal. National Academy of Sciences, Transportation Research Board (2015). “Road Transport as a Public-Private
Enterprise’, by Steven E. Shladover and Richard Bishop. https://www.nap.edu/read/22087/chapter/9

Thompson, Clive (2016). “No Parking Here.” Mother Jones. http://bit.ly/noparkinghere

UITP / International Association of Public Transport (2017). “Autonomous Vehicles: A Potential Game Changer for Urban Mobility.” http://bit.ly/UITPavs

University College London Transport Institute (2017). “Social and behavioural questions associated with automated vehicles,” by Tom Cohen, Peter Jones and Clémence Cavoli, http://bit.ly/ukAVsociety

Vanderbilt, Tom (2012). “Autonomous Cars Through the Ages.” WIRED. http://bit.ly/AVsthruAges

Victorian Transport Policy Institute (2017). “Autonomous Vehicle Implementation Predictions: Implications for Transport Planning.” By Todd Litman. http://bit.ly/AV_VTI