Autonomous Vehicle Implementation Predictions: Implications For Transport Planning

www.vtpi.
org
Info@vtpi.org
Phone 250-360-1560
Autonomous Vehicle Implementation Predictions

Implications for Transport Planning
27 October 2019
By
Todd Litman
Victoria Transport Policy Institute
Waymo’s self-driving taxis are a well-publicized example of autonomous vehicles.
Abstract
This report explores autonomous (also called self-driving, driverless or robotic) vehicle benefits
and costs, and implications for various planning issues. It investigates how quickly self-driving
vehicles are likely to be developed and deployed based on experience with previous vehicle
technologies, their benefits and costs, and how they are likely to affect travel demands and
planning decisions such as optimal road, parking and public transit supply. This analysis
indicates that some benefits, such as more independent mobility for affluent non-drivers, may
begin in the 2020s or 2030s, but most impacts, including reduced traffic and parking congestion
(and therefore infrastructure savings), independent mobility for low-income people (and
therefore reduced need for public transit), increased safety, energy conservation and pollution
reductions, will only be significant when autonomous vehicles become common and affordable,
probably in the 2040s to 2050s, and some benefits may require prohibiting human-driven
vehicles on certain roadways, which could take even longer.
Todd Alexander Litman © 2013-2019

You are welcome and encouraged to copy, distribute, share and excerpt this document and its ideas, provided the author is
given attribution. Please send your corrections, comments and suggestions for improvement.
Autonomous Vehicle Implementation Predictions: Implications for Transport Planning
Table of Contents
Introduction .................................................................................................................................... 3
Autonomous Vehicle Operational Models ..................................................................................... 4
Benefits and Costs........................................................................................................................... 5
Reduced Stress, Improved Productivity and Mobility .............................................................................. 5
Ownership and Operating Costs ............................................................................................................... 7
Traffic Safety ........................................................................................................................................... 10
External Cost ........................................................................................................................................... 11
Benefit and Cost Summary ..................................................................................................................... 13
Development and Deployment..................................................................................................... 14
Experience with Previous Vehicle Technology Deployment .................................................................. 16
Deployment Predictions ......................................................................................................................... 18
Travel Impacts ............................................................................................................................... 20
Potential Conflicts and Solutions .................................................................................................. 26
Planning Implications .................................................................................................................... 27
Conclusions ................................................................................................................................... 31
References .................................................................................................................................... 34
2
Introduction
The future is ultimately unknowable, but planning requires predictions of impending conditions
and needs (Shaheen, Totte and Stocker 2018). Many decision-makers and practitioners
(planners, engineers and analysts) wonder how autonomous (also called self-driving or robotic)
vehicles will affect travel and land use development patterns; road, parking and public transit
demands; traffic problems; and whether public policies should encourage or restrict their use
(APA 2016; Grush and Niles 2018; Guerra 2015; Kockelman and Boyles 2018; Levinson 2015;
Milakis, van Arem and van Wee 2017; Sperling 2017).
There is considerable uncertainty about these issues. Optimists predict that by 2030,
autonomous vehicles will be sufficiently reliable and affordable to replace most human driving,
providing independent mobility to non-drivers, reducing driver stress and tedium, and be a
panacea for congestion, accident and pollution problems (Johnston and Walker 2017; Keeney
2017; Kok, et al. 2017). However, there are good reasons to be skeptical of such claims.
Most optimistic predictions are based on experience with electronic innovations such as digital
cameras, smart phones and the Internet. Their analysis often overlooks significant obstacles
and costs. Although vehicles can now operate autonomously under certain conditions, many
technical problems must be solved before they can operate autonomously in all conditions –
including extreme weather, unpaved roads and during wireless service disruptions – and those
vehicles must be tested, approved for general commercial sale, affordable to most travellers,
and attractive to consumers. Motor vehicles last much longer and cost much more than
personal computers, cameras or telephones, so new technologies generally require many years
to penetrate vehicle fleets. A camera, telephone or Internet failure can be frustrating but is
seldom fatal; motor vehicles system failures can be frustrating and deadly to occupants and
other road users. Autonomous driving can induce additional vehicle travel which can increase
traffic problems. As a result, autonomous vehicles will probably take longer to develop and
provide smaller net benefits than optimists predict.
These factors have significant transport policy and planning implications (Papa and Ferreira
2018; Speck 2017). Vehicles rely on public infrastructure and impose external costs, and so
require more public planning and investment than most other technologies. For example,
autonomous vehicles can be programed based on user preferences (maximizing traffic speeds
and occupant safety) or community goals (limiting speeds and protecting other road users), and
many predicted autonomous vehicle benefits, including congestion and pollution reductions,
require dedicated lanes to allow platooning (numerous vehicles driving close together at
relatively high speeds). Policy makers must decide how to regulate and price autonomous
driving, and when potential benefits justify dedicating traffic lanes to their exclusive use.
This report explores these issues. It investigates, based on experience with previous vehicle
technologies, how quickly self-driving vehicles are likely to be developed and deployed,
critically evaluates their likely benefits and costs, and discusses their likely travel impacts and
their implications for planning decisions such as optimal road, parking and public transit supply.
3
Autonomous Vehicle Operational Models

The Society of Automobile Engineers (SAE) defined five levels of autonomous driving, as
summarized in Exhibit 1. Levels 1-3 require a licensed driver, but levels 4 and 5 allow driverless
operation, which is necessary for many predicted benefits.
Exhibit 1 Automated Driving Levels (SAE J3016 2014)
The SAE defines five vehicle

automation levels. Most
predicted benefits require
levels 4 or 5.
Exhibit 2 summarizes three autonomous vehicle operational models.
Table 2 Autonomous Vehicle Operational Models Compared

Advantages Disadvantages Appropriate Users
High convenience. Available High costs. Does not allow People who travel a lot,
Personal autonomous without delay. Items, such users to choose different reside in sprawled areas,
vehicles - Motorists own as equipment, tools and vehicles for different trips, want a particular vehicle,
or lease their own self- snacks, can be left in such as cars for commuting or or leave items in their
driving vehicles. vehicles. trucks for errands. vehicles.
Users must wait for vehicles.
Shared autonomous Limited service (no driver to
vehicles - Self-driving taxis Users can choose vehicles help passengers carry luggage
transport individuals and that best meet their needs. safely reach their door). Lower-annual-mileage
groups to destinations. Door to door service. Vehicles may be dirty. users.
Shared autonomous rides
- Self-driving vans (micro- Least convenience, comfort
transit) take passengers to and speed, particularly in Lower-income urban
or near destinations. Lowest costs. sprawled areas. residents.
Autonomous vehicles can be personal or shared. Each model has advantages and disadvantages.
4
Benefits and Costs

Autonomous vehicles can provide various benefits and impose various costs.
Reduced Stress, Improved Productivity and Mobility

Autonomous vehicles can reduce driver stress and tedium. Self-driving cars can be mobile
bedrooms, playrooms and offices, as illustrated below, allowing passengers to rest or be
productive while travelling (WSJ 2017). This can reduce travel time unit costs.
Exhibit 3 Productivity and Relaxation While Travelling
Self-driving cars can be mobile bedrooms, playrooms or offices, allowing travellers to rest and work.
On the other hand, self-driving vehicles can introduce new stresses and discomforts. To
minimize cleaning and vandalism costs, self-driving taxis and buses will have “hardened”
interiors (vinyl seats and stainless steel surfaces), minimal accessories, and security cameras.
Demand response ridesharing (vehicles with flexible routes to pick up and drop off passengers
at or near their destinations) will reduce security (passengers may need to share space with
strangers), and reduce travel speed and reliability since each additional pick-up or drop-off will
impose a few minutes of delay to other passengers, particularly in sprawled areas with dead-
end streets. Grush (2016) suggests that travellers will experience “access anxiety,” if they fear
that their vehicle cannot reach a desired destination.
Autonomous vehicles can provide independent mobility for non-drivers, including people with
disabilities, adolescents, and others or who for any reason cannot or should not drive. This
directly benefits those travellers, reduces chauffeuring burdens on their family members and
friends, and improves their access to education and employment opportunities, increasing their
economic productivity. Some affluent non-drivers living in sprawled areas may purchase
personal autonomous vehicles, and urban non-drivers are likely to use autonomous taxies.
Optimistic predictions of autonomous vehicle benefits may cause some communities to reduce
support for public transit services which may reduce mobility options for non-drivers.
Dedicating highway lanes for autonomous vehicle platooning may reduce capacity for human-
operated traffic, making travellers in human-operated vehicles worse off.
5
The Autonomous Vehicle Travel Experience

Autonomous vehicles are often illustrated (see below) with happy, well-dressed passengers lounging or
working in tidy self-driving cars that look like science fiction spaceships. However, the actual experience
will probably be less idyllic.
Self-driving vehicles will allow all vehicle occupants to rest, read, work and watch television (rather than
only listen to audio), but for safety sake they should wear seatbelts, and like any confined space, vehicle
interiors can become cluttered and dirty. Manufactures will probably produce vehicles with seats that turn
into beds and mobile offices (NYT 2017). For the foreseeable future autonomous vehicles are likely to be
and unable to operate in heavy rain and snow, on unpaved roads, or where GPS service or special maps
are unavailable, and they may be relatively slow and unreliable in mixed urban traffic.
Self-driving taxi and “micro-transit” (van) services will be cheaper than human-operated taxis, but offer
minimal service quality. To minimize cleaning and vandalism costs most surfaces will be stainless steel and
plastic, and passengers will be monitored by security cameras, yet passengers may still encounter previous
occupants’ garbage, stains and odors (Broussard 2018). There will be no drivers to help carry packages or
ensure passenger safety.
Like other public transportation, autonomous micro-transit will require passengers to share interior space
with strangers, who are mostly friendly and responsible but occasionally unpleasant and freightening.
Each additional passenger will add pickup and drop-off delays, particularly for passengers with special
needs, such as packages, children or disabilities, who need extra time, and in more sprawled areas with
dead-end streets where an additional stop can add several minutes. Because of these limitations,
autonomous taxi and micro-transit will most suited to local urban trips, and many travellers will choose to
own their own vehicle, or have a human operator, despite the extra cost.
Once the novelty wears off, autonomous vehicle travel will be considered utilitarian and tedious, a useful
but not particularly enjoyable or glamourous mobility option, more like an elevator than a spaceship.
6
Ownership and Operating Costs

Autonomous vehicles will require various equipment and services summarized in the box
below. Such technologies can add thousands of dollars to vehicle purchase prices and hundreds
of dollars of annual fees. For example, a package of optional electronic features such as remote
starting, high beam assist, active lane assist, adaptive cruise control and top view camera
typically increases new vehicle prices by more than $5,000, and navigation and security
services, such as OnStar and TomTom, cost $200-600 per year. Since failures could be deadly,
autonomous driving systems will need robust, redundant and abuse-resistant components
maintained by specialists, similar to aviation service standards, further increasing costs. To
monitor passenger behavior, autonomous vehicles will also require in-vehicle security cameras
and enforceable behavior rules, plus frequent interior cleaning and repairs (Broussard 2018).
Exhibit 4 Autonomous Vehicle Equipment and Service Requirements

All Autonomous Vehicles Shared Autonomous Vehicles
 Sensors (optical, infrared, radar, laser, etc.).
 Frequent cleaning and repairs.
 Automated controls (steering, braking, signals, etc.)
 Dispatching and fleet management.
 Software, servers and power supplies.
 Business administration and insurance.
 Wireless networks. Short range vehicle-to-vehicle
communications and long-range access to maps,  Business profits.
software upgrades and road reports.
 Security.
 Navigation. GPS systems and special high quality maps.
 Delays and empty vehicle-miles for
 Critical component testing and maintenance. passenger pick-up and drop-off.
Autonomous vehicles, particularly those that are shared, will incur additional costs.
This suggests that Level 4 and 5 autonomous driving capabilities will probably increase vehicle
purchase prices by several thousands of dollars and require hundreds of dollars in additional
annual services and maintenance costs, adding a few thousand dollars per vehicle-year for the
foreseeable future (one to three decades). Experience with previous vehicle innovations, such
as automatic transmissions and airbags, discussed later in this report, suggests that
autonomous driving capability will initially be available only on higher priced models, and will
take one to three decades to be incorporated into middle- and lower-priced models.
Advocates argue that these additional costs will be offset by insurance and fuel cost savings,
but that seems unlikely. For example, if autonomous driving cuts insurance costs in half, the
$300-500 annual savings is just 10-20% of estimated additional costs. Additional equipment and
larger vehicles to serve as mobile offices and bedrooms are likely to increase rather than reduce
energy consumption. Electric vehicles have low fuel costs, in part because they currently pay no
road user fees comparable to motor vehicle fuel taxes; cost-recovery road-user fees would
increase electric vehicle operating costs 5-10¢ per vehicle-mile (FHWA 2015).
7
Cleaning – An Often Overlooked Cost

Although most autonomous taxi passengers are likely to be courteous and responsible, some will
probably be messy and a few vandalous. To discourage abuse, autonomous taxis will have surveillance
cameras, and their interiors will be plastic and chrome, which may reduce but cannot eliminate these
problems, so vehicles will occasionally contain garbage and unpleasant odors, or be damaged.
Autonomous taxis will therefore require frequent inspections and cleaning. They will probably need a
quick cleaning that includes an inspection, garbage pick-up, vacuuming and surface wipe-downs
approximately every five to ten pick-ups, more comprehensive interior and exterior cleaning each day,
plus occasional repairs. Assuming $5-10 per cleaning this will add $0.50-1.00 per trip, or 5-10¢ per
vehicle-mile, plus travel time and costs for driving to cleaning stations.
This indicates that for the foreseeable future (one to three decades) autonomous vehicle costs
will probably average (total annual costs divided by annual mileage) $0.80-$1.20 per vehicle-
mile, which may eventually decline to $0.60-$1.00 per mile, which is somewhat more expensive
than human-operated vehicles’ $0.40-$0.60 per mile average costs (Stephens, et al. 2016).
Johnson and Walker (2017) predict that shared, electric, autonomous taxis cost will decline
from about 85¢ per vehicle-mile in 2018 to 35¢ per mile by 2035 (Exhibit 5), but they overlook
some previously-mentioned costs such as cleaning and roadway user fees, and so are probably
underestimates. Shared autonomous rides (self-driving public transit) will probably cost $0.20-
0.40 per passenger-mile, assuming that they average 3-6 passengers (Bösch, et al. 2017).
Exhibit 5 Automated Versus Personal Car Costs
Johnston and Walker (2017)

predict that Automated
Mobility Services using
electric sedan will decline
from 85¢ per mile in 2018 to
about 35¢ per mile by 2035,
less than half of a typical
Personal Sedan’s Total Cost
of Ownership (TCO), and
slightly less than their
operating expenses.
Some studies estimate lower costs (Keeney 2017). For example, Kok, et al. (2017) predict that
shared, electric autonomous vehicles operating costs will be less than 10¢ per mile, making
their use so inexpensive that it will often be funded through advertising, but such estimates
ignore significant costs such as vehicle maintenance and cleaning, business profits, empty
vehicle-travel, insurance (based on optimistic assumptions of autonomous vehicle safety), and
roadway costs (they assume that electric vehicles should continue to pay no road user fees),
and so are probably underestimates.
8
Automobiles currently have about $3,600 in fixed expenses (financing, depreciation, insurance,
registration fees, residential parking and scheduled maintenance) and $2,400 in variable
expenses (fuel, oil, tire wear and paid parking), and are driven about 12,000 annual miles,
which averages about 50¢ per mile, of which about 20¢ per mile is operating expenses (AAA
2017; Litman 2009). Human-operated taxis generally cost $2.00-$3.00 per mile, ride-hailing
(also called ridesourcing and Transportation Network Companies, such as Uber and Lyft) about
$1.50-2.50 per mile, and conventional public transit 20-40¢ per mile.
The following figure compares these costs. Average costs are what travellers consider when
deciding whether to purchase a vehicle; operation costs are what vehicle owners consider when
deciding how to make a particular trip.
Exhibit 6 Cost Comparison
$2.50
Autonomous Vehicles (AV)
$2.00 Human Driven (HD)
Dollars Per Mile
$1.50
$1.00
$0.50
$0.00
HD car AV Public AV HD car AV AV Ride- HD
operation rideshare transit operation average Taxi average hailing Taxi
Autonomous vehicles (AVs) are predicted to cost less than human-driven taxis and ride-hailing services,
but more than human-driven personal vehicles (HVs) and public transit services.
This indicates that in the future personal autonomous vehicles will continue to cost more than
human-operated vehicles, but shared autonomous vehicles will be cheaper than human-
operated ride-hailing and taxi services. Since most vehicle costs are fixed, owners of personal
autonomous vehicles will have little financial incentive to use shared vehicles. However, the
availability of shared autonomous vehicles may encourage more households to reduce their
vehicle ownership, and so reduce their annual vehicle travel, as discussed later in this report.
Autonomous vehicles can provide large savings for commercial vehicles, such as freight trucks
and buses, where driver wages and benefits are a major portion of total costs, although many
delivery vehicles require an operator to unload goods.
9
Traffic Safety and Security

Optimists claim that, because human error contributes to 90% of crashes, autonomous vehicles
will reduce crash rates and insurance costs by 90% (Fagnant and Kockelman 2013; Kok, et al.
2017; McKinsey 2016), but this overlooks additional risks these technologies can introduce (Hsu
2017; ITF 2018; Kockelman, et al. 2016; Koopman and Wagner 2017; Ohnsman 2014):
 Hardware and software failures. Complex electronic systems often fail, and even small vehicle
operating system failures - a false sensor, distorted signal, or software error - can have
catastrophic results. Self-driving vehicles will certainly have failures that contribute to crashes;
the question is their frequently compared with human drivers.
 Malicious hacking. Self-driving technologies can be manipulated for amusement or crime.
 Increased risk-taking. When travellers feel safer they often take additional risks, called offsetting
behavior or risk compensation. For example, if autonomous vehicles are considered very safe,
passengers may reduce seatbelt use and other road users may be less cautious (Millard-Ball
2016), described as “over-trusting” technology (Ackerman 2017).
 Platooning risks. Many potential benefits, such as reduced congestion and pollution emissions,
require platooning (vehicles operating close together at high speeds on dedicated lanes), which
can introduce new risks, such as human drivers joining platoons and increased crashes severity.
 Increased total vehicle travel. By improving convenience and comfort autonomous vehicles may
increase total vehicle travel and therefore crash exposure (Trommer, et al. 2016; WSJ 2017).
 Additional risks to non-auto travellers. Autonomous vehicles may have difficulty detecting,
communicating with and accommodating pedestrians, bicyclists and motorcycles (PBIC 2017).
 Reduced investment in conventional safety strategies. The prospect of autonomous vehicles may
reduce future efforts to improve driver safety (Lawson 2018).
Reports by eight companies operating autonomous test vehicles in 2017 indicate that
disengagements (when human drivers override automated systems) exceeded one per 5,600
miles (Edelstein 2018). Common problems included failing to recognize a “no right turn on red
signal,” cars that planned to merge into traffic with insufficient space, failing to brake enough at
a stop, difficulty detecting vehicles approaching in opposite lanes, problems maintaining GPS
location signals, software crashes, inability to recognize construction cones, confusion over
unexpected behavior by other drivers, plus other hardware and software problems.
These new risks will probably cause crashes, so net safety impacts are likely to be smaller than
the 90% crash reductions that advocates claim. Sivak and Schoettle (2015a) conclude that
autonomous vehicles may be no safer per mile than an average driver, and may increase total
crashes when self- and human-driven vehicles mix. Groves and Kalra (2017) argue that
autonomous vehicle deployment is justified even if they only reduce crash rates 10%, but their
analysis indicates that net safety gains are significantly reduced if this technology increases
total vehicle travel. For example, if autonomous vehicles reduce per-mile crash rates 10% but
increase vehicle travel 12%, total crashes, including risks to other road users, will increase.
10
Shared autonomous vehicles can reduce crashes by providing more affordable alternatives to
higher-risk drivers. Efforts to reduce higher-risk driving, such as graduated driver’s licenses,
special testing for senior drivers, and anti-impaired driver campaigns, can be more effective and
publicly acceptable if affected groups have convenient and affordable mobility options.
Many factors will affect these safety impacts, including how vehicles are programmed, and how
they affect total vehicle travel. For example, to maximize mobility autonomous vehicles can be
programmed to drive faster, take more risks in unpredictable situations, and platoon; to
maximize safety they can be programmed to drive slower, be more cautious, for example,
stopping for human instructions in any unexpected situation, and public policies, such as high
efficient road pricing, can encourage vehicle travel reductions.
External Cost
Advocates claim that autonomous driving will reduce external costs including traffic congestion,
energy consumption, pollution emissions, roadway and parking facility costs, although those
benefits are uncertain (Eddy and Falconer 2017; TRB 2019). To be more space and energy
efficient autonomous vehicles require dedicated lanes for platooning (Exhibit 7). This is only
feasible on grade separated highways.
Exhibit 7 Driverless Car “Platooning” (Chuen, et al. 2013)
Many proposed
autonomous vehicle
benefits, including
congestion and
emission reductions,
require platooning:
multiple electrically
connected vehicles
travelling close together
at relatively high
speeds, preferably lead
by a large truck. This
requires dedicated
highway lanes.
11
Under many circumstances, autonomous operation may increase congestion, energy, pollution
and roadway costs. Optimists assume that autonomous vehicles will reduce pollution because
they will be all electric and mostly shared, but as discussed previously, many users will probably
choose personal autonomous vehicles, unless widely applied public policies, such as high fossil
fuel taxes and high occupancy vehicle lanes on congested roadway, favor electric and shared
vehicles. Self-driving technologies requires additional equipment, and vehicle manufactures are
likely to market seats that turn into beds and mobile offices, which can increase total energy
consumption and pollution emissions.
Autonomous vehicles may require higher roadway maintenance standards, such as clearer line
painting and special traffic signals (Lawson 2018). Autonomous operation can reduce parking
costs by allowing vehicles to park further from destinations, but most users will probably want
their vehicles available within five or ten minutes, and so must park within a mile or two. Their
impacts on overall congestion, energy, emissions and crash costs will depend on how self-
driving technologies affect total travel and urban development patterns. By proving a
comfortable alternative to public transit travel, they may increase total urban vehicle traffic. To
avoid paying for parking, autonomous vehicles may circle city blocks, increasing traffic
congestion. If they strictly follow traffic laws and maximize caution, such as speed limits and
optimal spacing between vehicles, they will reduce traffic speeds and increase delays. To
maximize comfort, so passengers can rest or work, users may program their vehicle to minimize
acceleration and deceleration rates, reducing traffic speeds (Le Vine, Zolfaghari and Polak
2015). If programmed for maximum caution in unexpected conditions, they may frequently
stop to wait for human instructions.
Over the long term, autonomous vehicles may stimulate sprawled, automobile-dependent
development patterns, increasing sprawl-related costs and total vehicle travel, and by reducing
public transit demand they can reduce transit system revenues and efficiencies.
12
Benefit and Cost Summary

This review indicates that autonomous vehicles are likely to have both benefits and costs
(Milakis, van Arem and van Wee 2017). Some of these impacts, including increased vehicle
costs, reduced driver stress and productivity gains, directly affect users. Others, such as changes
in roadway costs, congestion, accident risk, pollution, and mobility options for non-drivers, are
external impacts. The total magnitude of these impacts will depend on how the technologies
affect vehicle travel: improved convenience and productivity cause vehicle travel to increase,
traffic problems such as congestion, accidents, pollution emissions and roadway costs may
increase, as ride-hailing services have done in large cites (Schaller 2017), but if more affordable
shared mobility options allow households to reduce their vehicle ownership, and therefore
leverage reductions in total vehicle travel, traffic problems should decline. The next section of
this report explores these impacts. Exhibit 8 summarizes these benefits and costs.
Exhibit 8 Autonomous Vehicle Potential Benefits and Costs

Benefits Costs/Problems
Internal (user Impacts)
Reduced drivers’ stress and increased Increased vehicle costs. Requires additional vehicle
productivity. Motorists can rest, play and work equipment, services and fees.
while travelling.
Additional user risks. Additional crashes caused by system
Mobility for non-drivers. More independent failures, platooning, higher traffic speeds, additional risk-
mobility for non-drivers can reduce motorists’ taking, and increased total vehicle travel.
chauffeuring burdens and transit subsidy needs.
Reduced security and privacy. May be vulnerable to
Reduced paid driver costs. Reduces costs for taxis information abuse (hacking), and features such as location
and commercial transport drivers. tracking and data sharing may reduce privacy.
Additional risks. May increase risks to other road users and

Increased safety. May reduce crash risks and may be used for criminal activities.
External (Impacts on others)
insurance costs. May reduce high-risk driving.

Increased traffic problems. Increased vehicle travel may
Increased road capacity and reduced costs. More increase congestion, pollution and sprawl-related costs.
efficient vehicle traffic may reduce congestion
and roadway costs. Social equity concerns. May reduce affordable mobility
options including walking, bicycling and transit services.
Reduced parking costs. Reduces demand for
parking at destinations. Reduced employment. Jobs for drivers may decline.
Reduced energy consumption and pollution. May Increased infrastructure costs. May require higher roadway
increase fuel efficiency and reduce emissions. design and maintenance standards.
Supports vehicle sharing. Could facilitate Reduced support for other solutions. Optimistic predictions
carsharing and ridesharing, reducing total vehicle of autonomous driving may discourage other transport
ownership and travel, and associated costs. improvements and management strategies.
Autonomous vehicles can provide various benefits and costs, including external impacts on other people.
13
Development and Deployment Predictions

Many current vehicles have Level 1 and 2 technologies such as cruise control, hazard warning
and automated parallel parking. Tesla’s Autopilot offers automated steering and acceleration in
limited conditions, but deployment was delayed after it caused a fatal crash in 2016 (Hawkins
2017). Several companies are now implementing Level 4 pilot projects, which means that
vehicles can drive themselves under certain conditions. For example, Waymo and Uber
announced plans to begin testing driverless taxi services (Bergen 2017; Lee 2017). Despite this
progress, significant technical progress is needed before vehicles can drive themselves under all
normal conditions (Simonite 2016). For the foreseeable future autonomous vehicles are
unlikely to operate reliably in heavy rain and snow, on unpaved roads, or where GPS service or
special maps are unavailable, and they may be slow and unreliable in mixed urban traffic.
Operating a vehicle on public roads is complex due to the frequency of interactions with other,
often-unpredictable objects including vehicles, pedestrians, cyclists, animals and potholes.
Because of these interactions, autonomous vehicles will require orders of magnitude more
complex software then aircraft (Exhibit 9). Producing such software is challenging and costly,
and ensuring that it never fails is virtually impossible. There will almost certainly be system
failures, including some that cause severe accidents.
Exhibit 9 Aircraft and Automobile Software Code Compared (GAO 2016)

Vehicles need more
complex computer
systems than aircraft,
due to the complexity
of roadway
interactions.
Consider one challenge. For safety sake motorists are advised to drive defensively, which means
anticipating potential risks such as wild animals and playful children. To do this, autonomous
vehicles will need a database that categorizes, for example, fire hydrants as low-risk, pets on
leashes as medium risk, and wild animals, such as kangaroos, as high risk. In addition, children
sometimes dress in animal costumes, and adolescents in zombie variations. Most drivers can
understand such risks. If I warn, "Watch out for teenagers in zombie kangaroo costumes," you
could probably understand the threat since you too were once a playful youth, but a computer
would be flummoxed: such an unusual situation is unlikely be in a standard database, so the
vehicle would either miss-categorize the risk, perhaps treating costumed fun-seekers as injured
crash victims or a riotous mob, or stop and wait for human instructions. These systems can self-
learn, and so could eventually recognize new costumes and behaviors, but this will require new
software coding that may interact unpredictably with other instructions. This is not to suggest
that autonomous driving is impossible or inherently harmful, it simply illustrates one of many
problems they face: new risks leading to solutions that further increase system complexity and
therefore potential failures.
14
Even after Level 5 technology is fully functional and reliable, additional time will be required for
independent testing and regulatory approval. Because vehicles can impose significant external
costs, including accident risks and delays to other road users, vehicle technologies require
higher testing and regulation standards than most other technological innovations such as
personal computers and mobile phones. Under optimistic conditions testing and approval will
only require a few years, but the technology may prove unreliable and dangerous, for example,
if it causes high-profile crashes, which could require several more years (Bhuiyan 2017). It is
likely that different jurisdictions will impose different testing, approval and regulations,
resulting in varying rates of deployment.
Most experts acknowledge that significant progress is needed before Level 5 automation is
reliable, tested and approved (Mervis 2017). For example, Michigan Mobility Transformation
Center director Huei Peng said that, “it may be decades before a vehicle can drive itself safely at
any speed on any road in any weather” (Truett 2016). Similarly, Toyota Research Institute CEO,
Gill Pratt stated that autonomous driving, “is a wonderful goal but none of us in the automobile
or IT industries are close to achieving true Level 5 autonomy” (Ackerman 2017). Uber self-
driving vehicle lab director Raquel Urtasun said that, "Having self-driving cars at a smaller scale,
on a small set of roads, we are fairly close. To see at an Uber scale we are far…Nobody has a
solution to self-driving cars that is reliable and safe enough to work everywhere" (Marowits
2017).
Artificial intelligence expert Yoshua Bengio said that, "I think people underestimate how much
basic science still needs to be done before these cars or such systems will be able to anticipate
the kinds of unusual, dangerous situations that can happen on the road" (Marowits 2017).
Similarly, Heilbronn University artificial intelligence expert Professor Nicolaj Stache said, “The
vision that drives us is to replicate the human car driver – only without replicating human
mistakes. In other words, we are aiming to substitute the human brain through artificial
intelligence. That’s still a long way away, but we are working on it” (Ebert 2016).
In contrast to these cautious predictions by experts, most optimistic predictions are made by
people with financial interests in autonomous vehicle industries, based on experience with
other types of technology. For example, the widely-cited report, “Rethinking Transportation
2020-2030: The Disruption of Transportation and the Collapse of the Internal-Combustion
Vehicle and Oil Industries” was written by ReThink, “an independent think tank that analyzes
and forecasts the speed and scale of technology-driven disruption and its implications across
society.” Mobility-As-A-Service: Why Self-Driving Cars Could Change Everything, was published
by ARK Investment Management and written by an analyst who “covers autonomous cars,
additive manufacturing, infrastructure development, and innovative materials,” with little
apparent experience with transportation innovation. Automotive Revolution – Perspective
Towards 2030: How the Convergence of Disruptive Technology-Driven Trends Could Transform
the Auto Industry, was published by the McKinsey Corporation, a business management firm.
To their credit, such predictions are often qualified – autonomous vehicles “could” or “might”
change everything – but their conclusions are repeated with certitude.
15
Such reports are primarily oriented toward investors and so focus on the autonomous vehicle
sales potential, but policy makers and planners are interested in their fleet penetration and
travel impacts. Motor vehicles are durable and expensive; consumers seldom purchase new
vehicles simply to obtain a new technology, so innovations generally take decades to fully
penetrate vehicle markets. As a result, even if autonomous driving technologies penetrate new
vehicle markets in the 2020s, it will be the 2040s or 2050s before most vehicles are capable of
autonomous driving. Optimists argue that benefits will be large enough to justify premature
scrapping of most vehicle that lack autonomous driving capability, but that seems unlikely
under realistic assumptions of their benefits and costs.
In addition to technological progress, market deployment depends on consumer demand:

travellers’ willingness to pay for autonomous mobility. Surveys indicate significant concerns
about autonomous vehicle privacy and safety (Schoettle and Sivak 2014), and until they are
proven reliable in all conditions, many travellers will have “access anxiety,” they will fear that
their vehicle cannot reach desired destinations (Grush 2017). Although current technologies
allow autonomous vehicle operation in approximately 90% of conditions, achieving 99%
operability (vehicles cannot reach about 1% of desired destinations, or about 10 times a year
for a typical motorist) will be exponentially more difficult, and achieving 99.9% of conditions
(vehicles are unable to make 0.1% of trips, or about once a year), a reasonable target for
regulators and customers, will be exponentially more difficult again (Wharton 2017).
Experience with Previous Vehicle Technology Deployment

Previous vehicle technologies can help predict autonomous vehicle deployment:
 Automatic Transmissions (Healey 2012). First developed in the 1930s, it took until the 1980s to
become reliable, and affordable. When optional they typically cost $1,000 to $2,000. Their
current new vehicle market share is about 90% in North America and 50% in Europe and Asia.
 Air Bags (Dirksen 1997). First introduced in 1973. Initially an expensive and sometimes
dangerous option (they could cause injuries and deaths), they became cheaper and safer, were
standard on some models starting in 1988, and mandated by U.S. federal regulation in 1998.
 Hybrid Vehicles (Berman 2011). Became commercially available in 1997, but were initially
unreliable and expensive. Their performance has improved, but typically adds about $5,000 to
vehicle prices. In 2012 they represented about 3.3% of total vehicle sales.
 Subscription Vehicle Services. Navigation, remote lock/unlock, diagnostics and emergency
services. OnStar became available in 1997, TomTom in 2002. They typically cost $200-400
annually. About 2% of U.S. motorists subscribe to the largest service, OnStar.
 Vehicle Navigation Systems (Lendion 2012). Vehicle navigation systems became available as
expensive accessories in the mid-1980s. In the mid-1990s factory-installed systems became
available on some models, for about $2,000. Performance and usability have since improved,
and prices have declined to about $500 for factory-installed systems, and under $200 for
portable systems. They are standard in many higher-priced models.
16
Exhibit 11 summarizes their deployment. Most required decades from initial commercial
availability to market saturation, and some never became universal. Lavasani and Jin (2016)
conclude that, because autonomous vehicle technologies are more revolutionary than those in
Exhibit 11, they will probably have slower initial market acceptance and penetration.
Exhibit 11 Vehicle Technology Deployment Summary

Technology Deployment Cycle Typical Cost Premium Market Saturation Share
Automatic transmissions 50 years (1940s-90s) $1,500 90% U.S., 50% worldwide
Air bags 25 years (1973-98) A few hundred dollars 100%, due to federal mandate
Hybrid vehicles 25+ years (1990s-2015+) $5,000 Uncertain. Currently about 4%.
Subscription services 15 years $250 annual 2-5%
Navigation systems 30+ years (1985-2015+) $500 and rapidly declining Uncertain; probably over 80%.
New technologies usually require several decades between commercial availability to market saturation.
The first affordable car, Ford’s Model T, began production in 1908, leading to mass automobile
ownership, but the transportation system continued to be mixed for several decades, with most
travellers relying on walking, bicycling and public transit in addition to cars. Only after the 1980s
did motorization approach saturation, with most potential drivers having a personal vehicle.
Exhibit 10 U.S. Population and Automobile Ownership (FHWA 2016)

0.9
300 Population
Although mass
Population and Vehilces
0.8
250
Vehicles automobile production
Vehicles Per Capita
0.7
Vehicles Per Capita started in 1908, during
(Millions)
200 0.6 the next half century the

0.5 transportation system
150 was mixed. Only after
0.4 the 1960s did most
100 0.3 potential drivers own a
personal vehicle, and
0.2
50 only after 1980 did
0.1 ownership approach
0 0 saturation.
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
New vehicles are becoming much more durable, which reduces fleet turnover. As a result, new
vehicle technologies normally require three to five decades to penetrate 90% of vehicle fleets.
Deployment may be faster in developing countries where fleets are expanding, and in areas
with strict vehicle inspection requirements, such as Japan’s shaken system. Annual mileage
tends to decline with vehicle age: vehicles average approximately 15,000 miles their first year,
10,000 miles their 10th year, and 5,000 miles their 15th year, so vehicles over ten years
represent about 50% of vehicle fleets but only 20% of mileage (ORNL 2012, Table 3.8).
17
Innovation Deployment Patterns

Innovations generally follow a predictable S-curve deployment pattern, as illustrated in Exhibit
13. An initial concept requires development, testing, approval, commercial release, product
improvement, market expansion, differentiation, maturation, and eventually saturation and
decline. Autonomous vehicle technology will probably follow this pattern.
Exhibit 12 Innovation S-Curve

Decline
100%
Saturation
80% Maturation
Potential Market
Market diffusion
60%
Expansion
40% Product
improvement
20% Commercial release

Testing and approval
Development
0%
Concept
Most innovations follow a predictable deployment pattern, often called in innovation S-curve.
Autonomous vehicle technologies are currently in development, testing and approval stages.
There are several stages and therefore many years, before they are widely commercially
available, become reliable and affordable, and therefore start to saturate the vehicle fleet.
Deployment Predictions
Exhibit 13 uses the previous analysis to predict autonomous vehicle sales, fleet and travel
market penetration, assuming that Level 5 vehicles become commercially available in the 2020s
but are initially expensive and limited in performance. Due to these limitations, during their first
decade only a minority of new vehicle are likely to be fully autonomous, with market shares
increasing as their prices decline, performance improves, and consumers gain confidence. In
the 2040s approximately half of vehicles sold and 40% of vehicle travel could be autonomous.
Without mandates, market saturation will probably take several decades, and a portion of
motorists may continue to choose human operated vehicles due to costs and preferences.
These results are approximately consistent with other estimates by researchers (Cathers 2014;
Grush 2016; Lavasani and Jin 2016; Simonite 2016), although slower than the optimistic
predictions by some industry experts (Kok, et al. 2017; McKinsey 2016).
18
Exhibit 13 Autonomous Vehicle Market Penetration Projections

Stage Decade New Sales Fleet Travel
Available with large price premium 2020s 2-5% 1-2% 1-4%
Available with moderate price premium 2030s 20-40% 10-20% 10-30%
Available with minimal price premium 2040s 40-60% 20-40% 30-50%
Standard feature included on most new vehicles 2050s 80-100% 40-60% 50-80%
Saturation (everybody who wants it has it) 2060s ? ? ?
Required for all new and operating vehicles ??? 100% 100% 100%
Autonomous vehicle will probably take several decades to penetrate new vehicle sales, fleets and total
vehicle travel.
Exhibit 14 illustrates these deployment rates.
Exhibit 14 Autonomous Vehicle Sales, Fleet and Travel Projections (Based on Exhibit 13)
100%
Sales - Optimistic
Sales - Pessimistic
80% Travel - Optimistic
Travel - Pessimistic
Fleet - Optimistic
60%
Fleet - Pessimistic
40%
20%
0%
2020 2030 2040 2050 2060 2070
If autonomous vehicles follow previous vehicle technologies, it will take one to three decades for them to
dominate new vehicle sales, and one or two more decades to dominate vehicle travel, and even at
saturation a portion of vehicle travel may continue to be human operated, indicated by dashed lines.
Because bus and truck drivers earn relatively high wages, they are likely to become automated
most quickly, particularly for long-haul trips. However, professional drivers provide various
services – passenger security and assistance, systems monitoring and minor repairs – that will
be lost with fully automated vehicles.
Significantly faster implementation will require more rapid development, deployment and fleet
turnover than previous vehicle technologies. For example, for most vehicle travel to be
autonomous by 2035, almost all vehicles produced after 2025 would need to be autonomous,
and new vehicle purchase rates would need to triple so fleet turnover that normally takes three
decades can occur in one. This would require significant vehicle spending increases, at least in
the short-run, and scraping many otherwise functional vehicles that lack self-driving capability.
19
Emerging shared mobility services, such as carsharing and ride-hailing are reducing vehicle
ownership and parking demand in some situations (DeLuca 2018), and could accelerate
autonomous vehicle travel, but there are significant obstacles. As previously described, outside
dense urban areas autonomous taxis and micro-transit are relatively inconvenient and
inefficient, and so are unlikely to replace most private vehicle travel in suburban and rural areas
where most Americans live.
Autonomous vehicle implementation could be slower and less complete than optimistic
predictions. Technical challenges may prevent reliable and affordable autonomous vehicles
from be commercially available until the 2030s or 2040s. Their costs may be higher and benefits
smaller than expected. Consumer acceptance may be reduced by fears, privacy concerns, or
preferences, resulting in a significant portion of vehicle travel remaining human-driven even
after market saturation, indicated by dashed lines in Exhibit 15.
Travel Impacts
Many costs and benefits will depend on how autonomous vehicles affect total vehicle travel.
Exhibit 15 summarizes ways that autonomous vehicles may increase or reduce vehicle travel.
Miller and Kang (2019) offer guidance for modelling these effects.
Exhibit 15 Autonomous Vehicle Impacts on Total Vehicle Travel

Increases Vehicle Travel Reduces Vehicle Travel
More convenient shared vehicle services allow
Increased vehicle travel by non-drivers. households to reduce vehicle ownership and use.
Increased convenience and productivity increases travel. Shared autonomous vehicles reduce vehicle ownership.
Empty vehicle travel to drop off and pick up passengers Self-driving buses can improve transit services.
Encourage sprawled development. Reduced traffic risk and parking facilities can make urban
living more attractive.
Reduces traffic congestion and vehicle operating costs,
which induces additional vehicle travel. Reduce some vehicle travel, such as cruising for parking.
Self-driving vehicles can affect total vehicle travel (VTM) in various ways.
Autonomous vehicles are likely to increase vehicle travel by non-drivers, such as people with
disabilities and adolescents. They typically represent 10-30% of community residents but tend
to have relatively low vehicle travel demands, and are now often chauffeured by family
members or friends, so self-driving vehicles may cause little net increase in their vehicle travel.
Autonomous driving increases driver convenience and productivity, which can stimulate vehicle
travel, for example, encouraging users to choose longer commute and errand trips, and more
sprawled locations (Fleming and Singer 2019; Stephens, et al. 2016). Autonomous vehicles can
also stimulate empty vehicle travel, for example, when picking up or dropping off passengers,
or when waiting to be summoned; it will often be cheaper for a car to drive around than to pay
parking fees. With current policies these factors are likely to increase total vehicle travel.
20
Fleming and Singer (2019) surveyed 1,000 U.S. adults concerning their preferences and
responses to autonomous vehicles. Although many respondents indicated that autonomous
driving technology would not change their annual vehicle travel, those that do anticipate
changes are far more likely to travel more rather than less. Sivak and Schoettle (2015) estimate
that accommodating non-drivers’ latent travel demands could increase total vehicle up to 11%.
Trommer, et al. (2016) predict that autonomous vehicles will increase total vehicle travel 3-9%
by 2035. Keeney (2017) predicts a three-fold traffic increase but provides no supporting
evidence.
Taiebat, Stolper and Xu (2019) developed a microeconomic model to estimate vehicle travel
elasticities with respect to fuel and time costs. Their central estimate of the combined elasticity
of VMT demand is −0.4. They find that most households are more sensitive to time than to fuel
costs, and that wealthier households have more elastic demand. They use these estimates to
predict the VMT and energy use impacts of various connected and autonomous vehicle
adoption scenarios, and forecast a 2–47% increase in travel demand for an average household.
This indicates that a net rise in energy use is possible, especially in higher income groups.
Affordable self-driving taxis and micro-transit may cause households to reduce vehicle
ownership and rely more on shared vehicles and trips. This can significantly affect total vehicle
travel because owned and shared vehicles have very different cost profiles: owned vehicles
have high fixed (typically $4,000 annual) and low variable costs (typically 20¢ per mile), which
gives owners an incentive to maximize their driving in order to “get their money’s worth,” while
shared vehicles, such as carsharing and taxis, have minimal fixed costs and high variable costs
(typically $0.50-2.50 per mile), giving users incentives to minimize vehicle travel. As a result,
households tend to significantly reduce their vehicle travel, typically by 25-75%, when they shift
from owning to sharing vehicles (Lovejoy, Handy and Boarnet 2013).
On the other hand, taxi services require significant amounts of deadheading (vehicle travel with
no passenger to relocate vehicles). Analysis by Henao and Marshall (2018) estimates that at
least 41% of current ride-hailing vehicle travel is deadheading, resulting in a 0.8 average
passenger occupancy rate, accounting for deadheading. As autonomous taxi services expand,
deadheading distances may decline, but cannot disappear, and will probably be significant in
suburban and rural areas where demand is dispersed.
Advocates predict that convenient and affordable autonomous taxis will quickly displace private
vehicle (ITF 2014; Keeney 2017). Kok, et al (2017), predict that, “By 2030, within 10 years of
regulatory approval of fully autonomous vehicles, 95% of all U.S. passenger miles will be served
by transport-as-a-service (TaaS) providers who will own and operate fleets of autonomous
electric vehicles providing passengers with higher levels of service, faster rides and vastly
increased safety at a cost up to 10 times cheaper than today’s individually owned (IO) vehicles.”
21
However, these predictions are based on optimistic assumptions of shared vehicle convenience
and affordability. Many travellers will have good reasons to own personal vehicles:
 Convenience. Motorists often keep items in their vehicles, including car seats, tools, sports
equipment and emergency supplies.
 Speed and Reliability. Under optimal conditions taxis can arrive in less than five minutes of a
summons, but often take much longer, particularly during busy periods, for special vehicle types
(such as a van to carry multiple passengers or a wheelchair), and in suburban and rural areas.
 Costs. Vehicle sharing is generally cost effective for motorists who drive less than about 6,000
annual miles. People who live in suburban and rural areas, who usually commute by car, or who
for other reasons drive high annual miles will probably choose to own a personal vehicle.
 Status. Many people take pride in their vehicles and their driving ability, and so may prefer to
own private vehicles, and have the option of driving.
Exhibit 16 summarizes the travellers and trips most suitable for personal or shared vehicle
travel. In many cases, shared autonomous vehicles will allow households to reduce but not
eliminate personal vehicles, for example, owning one rather than two vehicles.
Exhibit 16 Personal Versus Shared Vehicles

Personal Vehicles Shared Vehicles
Travellers who place a high value on comfort or status.
Motorists who drive more than 6,000 annual miles, including Trips currently made by taxi or carshare vehicles.
most suburban and rural residents, and commercial travellers.
Utilitarian trips currently made by a private
Motorists who require special accessories in their vehicles. vehicle driven less than 6,000 annual miles.
Motorists who carry equipment, tools or dirty loads. Urban residents.
Travellers who place high values on privacy. People who want to save money more than time.
Some travellers are most suitable for personal vehicles, other for shared vehicles.
These scenarios illustrate how autonomous vehicles could impact various users’ travel:
Jake is an affluent man with degenerating vision. In 2026 he gives up driving and purchases an
autonomous vehicle. Impacts: An autonomous vehicle allows Jake to maintain independent mobility,
which increases his vehicle ownership and travel, residential parking demand, and external costs
(congestion, roadway costs, parking subsidies, and pollution emissions).
Bonnie lives and works in a suburb. She can bike to most destinations but occasionally needs a car. In
a city she could rely on taxis, but in suburbs they are slow and expensive. Starting in 2030 a local
company started offering convenient and affordable autonomous taxi services. Impacts:
Autonomous vehicles allow Bonnie to rely on bicycling and shared vehicles rather than a personal
car, which reduces her total vehicle travel, residential parking demand, and external costs.
Melisa and Johnny have two children. Melisa works at a downtown office. After their second child
was born in 2035, they shopped for a larger home. With conventional cars they would need a house
22
within a 30-minute commute of the city center, but the availability of new autonomous vehicles let
them consider more distant homes with commutes up to 60-minutes, during which Melisa could rest
and work. Impacts: Affordable new autonomous vehicles allow Melisa and Johnny to choose an
exurban home which increased their total vehicle travel and associated costs.
Garry is a responsible driver when sober but dangerous when drunk. By 2040 he had accumulated
several impaired citations and at fault accidents. With conventional cars Garry would continue
driving impaired until he lost his drivers’ license or caused a severe crash, but affordable used self-
driving vehicles allow lower-income motorists like Garry to avoid such problems. Impacts: Affordable
used autonomous vehicles allow Garry to avoid impaired driving, accidents and revoked driving
privileges, which reduces crash risks but increases his vehicle ownership and travel, and external
costs compared with what would otherwise occur.
Exhibit 17 summarizes the resulting impacts of these various scenarios. In most of these cases
autonomous vehicles increase total vehicle mileage.
Exhibit 17 Autonomous Vehicle Scenario Summary

User Benefits Travel Impacts Infrastructure Impacts
Jake (affluent and Independent mobility for Increased vehicle travel and Increased residential
visually impaired) non-drivers external costs parking and roadway costs
Bonnie (multi- Reduced vehicle ownership Reduced residential parking
modal traveller) Vehicle cost savings and travel and roadway costs
Melisa and Johnny Improved home location Increased vehicle ownership Increased residential
(suburban parents) options and travel parking and roadway costs
Garry (high-risk Avoids driving drunk and Less high-risk driving, more Increased residential
driver) associated consequences total vehicle travel parking and roadway costs
Autonomous vehicle availability can have various direct and indirect impacts.
23
These impacts will vary by travel demands, that is, trip types, as summarized in Exhibit 18.
Exhibit 18 Autonomous Vehicle Impacts on Various Travel Demands

Travel Type Autonomous Vehicle Impacts Portion of Travel
Particularly suitable for long-haul fright travel, due to its high labor
Freight trucks costs and limited routes, mostly on grade-separated highways. 10%
Trades (plumbers, computer technicians, etc.) carry equipment in
Small commercial their vehicles, and deliveries often require a person to unload, and so
(trades and deliveries) are likely to use owned autonomous vehicles with no travel change. 5%
Particularly suitable for public transit, due to its high labor costs. Currently 2%, but
Public transport Allows micro-transit with frequent and demand-response services. could increase.
Longer-distance (> 50 Particularly suitable for longer-distance personal trips, due to tedium. Currently 20%, but
mile) personal trips May increase longer-distance travel. could increase.
Affluent suburban and rural residents are likely to purchase private
autonomous vehicles and increase total vehicle travel. Lower-income
Local suburban and residents are likely to continue driving personal vehicles or use shared
rural autonomous vehicles, which could reduce their total vehicle travel. 50%
Many are likely to shift from personal cars to shared autonomous
Local urban trips mobility services, which is likely to reduce their total vehicle travel. 20%
Particularly suitable for non-drivers. Many are likely to increase their Currently 2-4%, but
Non-drivers vehicle travel. could increase.
Autonomous vehicle travel impacts will vary by types of trips.
24
A detailed review of these effects, Travel Effects and Associated Greenhouse Gas Emissions of
Automated Vehicles (Rodier 2018b), identified various ways that autonomous vehicle
technologies can affect total vehicle travel, as summarized in Exhibit 19.
Exhibit 19 Summary of Findings and Quality of Evidence (based on Rodier 2018b)

Mechanism Summary of Findings Quality of Evidence
Reduced headways could almost double or Limited research largely uses microsimulation traffic
triple roadway capacity. Elasticity of VMT with models. More measured data needed. The body of
respect to road capacity increase is 0.3 to 0.6 literature on the effect of expanded road capacity
Road Capacity (short run) and 0.6 to 1.0 (long run). and VMT is relatively strong.
Studies largely extrapolate from car and rail travel
Vary widely, but 75% to 82% of current driver experiences, which may not be consistent with
Time Cost values of time may be reasonable. automated vehicle travelers’ actual experience.
Insurance and fuel costs may decline. Avoided The magnitude of cost reductions is largely
labor costs enable low-cost AV taxis and micro- speculative. The body of literature on the effect of
transit. Elasticity of VMT with respect to gas gas prices on VMT is relatively strong. Gas price is
price is -0.03 to -0.10 (short run) and - 0.13 to - the largest component of the variable cost of driving
Monetary 0.30 (long run). Elasticity of taxi trips with a conventional owned vehicle. Only one study in
Cost respect to fares is -0.22. New York City estimates taxi fare elasticity.
Available research suggests that AVs would Limited research confirms expected direction
reduce transit and non-motorized and increase change, but magnitude is highly uncertain due to
Mode Choice car mode shares study quality.
Fully AV taxis may reduce parking demand by Only one U.S. study that uses observed travel data.
about 90%. However, reduced parking may Two other studies are in European cities. All studies
Parking increase relocation travel. use simulation models.
Empty relocation travel is positively correlated
Empty with distance from the urban core, parking
Relocation prices, and inversely correlated with vehicle Limited research confirms expected direction
Travel operating costs and transit fares. This could change, but the magnitude is uncertain. Few studies
(deadheading) significantly increase vehicle travel. fully represent induced travel effects.
Most studies estimate a 10% to 14% increase in Extrapolations from National Household Travel
New Travelers VMT. Survey data. Analysis based on study assumptions.
Many factors can affect autonomous vehicle travel impacts.
This suggests that with current policies, autonomous vehicles are likely to significantly increase
total vehicle miles travelled (VMT) and pollution emissions, probably by 10-30%, and more on
some travel corridors. This is likely to increase urban traffic congestion and sprawled
development unless road use is more efficiently priced (Miller and Kang 2019). Electrifying the
vehicle fleet could counter emission growth, but unless electric vehicle operation is efficiently
priced this will reduce vehicle operating costs which will further increase vehicle travel and
traffic impacts. Shared autonomous taxis could significantly reduce vehicle travel and
emissions, but only if policies favor their use over personal cars.
25
Potential Conflicts and Solutions

There are potential conflicts between user and community goals in autonomous vehicle design
and programming. For example, if programmed to maximize sleeping passengers’ comfort they
may reduce traffic speeds, and if programmed to protect occupants they may increase crash
risk to other road users. Some benefits (reduced congestion and possibly pollution emissions)
require that autonomous vehicles have dedicated lanes. This will raise debates about the
fairness, pricing, regulations and enforcement of these requirements.
There are also potential transportation planning conflicts. By increasing vehicle travel demand
and traffic speeds, and displacing public transit, autonomous vehicles could exacerbate traffic
congestion, sprawl-related costs, and mobility inequity. For example, if parking is priced but
roads are not, autonomous vehicles may cruise urban streets to avoid paying for parking,
exacerbating congestion and pollution problems. Some advocates claim that autonomous
vehicles eliminate the need for conventional public transit services, but high capacity transit will
still be needed on major travel corridors, and autonomous technologies can support transit by
reducing operating costs and improving access to stops and stations (ITF 2014; TRB 2017).
Shared vehicles reduce parking demand but increase the need for convenient pick-up and drop-
off options, which requires better curb management to minimize conflicts and risks (OECD/ITF
2018). Various public interest organizations have developed guidelines for optimizing the
benefits of emerging mobility technologies and services (Fulton, Mason and Meroux 2017;
Kaohsiung EcoMobility Festival 2017). The box below summarizes one example.
Shared Mobility Principles for Livable Cities (www.sharedmobilityprinciples.org)

1. Plan our cities and their mobility together. 6. Lead the transition towards clean and renewable energy.
2. Prioritize people over vehicles. 7. Support fair user fees across all modes.
3. Support the shared and efficient use of 8. Aim for public benefits via open data.
vehicles, lanes, curbs, and land. 9. Work towards integration and seamless connectivity.
4. Engage with stakeholders. 10. In dense urban areas autonomous vehicles should only
5. Promote equity. operate in shared fleets.
The following policies can help maximize benefits (Schlossberg, et al. 2018; TRB 2017):
 Test and regulate new technologies for safety and efficiency.
 Require autonomous vehicles to be programed based on ethical and community goals.
 Efficiently regulate and price roads and curb space to minimize conflicts, congestion and risks.
 Favor shared and higher-occupant vehicles over lower-occupant vehicles on public roads.
 Support high capacity public transit on major travel corridors.
 Reduce parking requirements to take advantage of shared vehicles.
 Efficiently price development to prevent inefficient sprawl.
 Use vehicle traffic reductions to redesign streets and improve urban livability.
26
Planning Implications
Autonomous vehicles raise many policy and planning issues (Taeihagh and Lim 2018). Their
development is just one of many trends that will affect future transport demands and planning
needs, as illustrated in Exhibit 20. Changes in demographics, consumer preferences, prices,
information technologies, mobility options, and other planning innovations will also influence
how people want to travel. These may have greater impacts than autonomous vehicles for the
foreseeable future.
Exhibit 20 Factors Affecting Transport Demands and Costs
Demographic Trends Price Changes Improved Travel Options

Aging population Rising fuel prices Better walking and cycling
More working at home Efficient road & parking pricing Improved public transit
Reduced youth drivers’ license Telework and delivery services
Carsharing
Changing User Preferences Intelligent Transport Systems (ITS)

Less driving Planning Innovations
Improved user information/navigation Expanded objectives
Shared rather than personal vehicles
Electronic pricing Systems operations
More walking & cycling
Autonomous vehicles Demand management
More urban living
Autonomous vehicles are one of many factors affecting future transport demands.
Some autonomous vehicle benefits, such as reduced driver stress, can occur with Level 2-3
automation, but other benefits, including independent mobility for non-drivers and increased
occupant safety require Level 4-5, and most external benefits (reduced traffic congestion, crash
risk, pollution, and infrastructure costs imposed on others) can only occur when autonomous
vehicles are common, and some require that highway lanes be dedicated to autonomous
vehicle platoons. The following matrix summarized the benefits provided by various AV levels.
Exhibit 21 Autonomous Vehicle Benefits

Autonomous Vehicle Mobility for Reduced User Occupant External
Levels Non-drivers Driver Stress Savings Safety Benefits
Level 1-3 personal vehicles  ?
Level 4 + 5 personal vehicles    
Shared autonomous vehicles   
Shared autonomous rides   
Dedicated AV lanes  ?
Autonomous vehicles benefit users by improving their mobility options, reducing stress, saving money
and increasing safety. External benefits (reduced crash risk, congestion delay, emissions and parking
costs imposed on others) primarily result from shared vehicles and rides that reduce total vehicle travel.
27
Exhibit 22 summarizes key autonomous vehicle planning issues.
Exhibit 22 Key Autonomous Vehicle Planning Issues (based on Papa and Ferreira 2018)
Issues Optimistic Outcome Pessimistic Outcome
Sharing Policies encourage autonomous vehicle sharing. AVs are promoted as private luxury goods.
Social Policies designed to maximize AV affordability and AVs are only affordable and available by privileged
exclusion accessibility ensure that they are widely available. (affluent) users.
Environmental AV policies give little consideration of to
sustainability AV policies support environmental goals. environmental concerns.
Operated AV operating systems are programmed based on AV operating systems are programmed based on
cooperation cooperative, altruistic and ethical principles. competitive, aggressive and defensive principles.
Public Public policies support public transport, providing Public policies focus too much on AVs and fail to
transport funding and favoring shared vehicles in traffic. support public transport.
Intermodal AV policies and programming respect human life. Public policies and programming favor AV
traffic They minimize crash risks and protect vulnerable occupants over other road users, and so will favor
regulations road users (e.g., through lower speeds). affluent over more vulnerable groups.
Network Data networks are designed make more Data networks are designed to maximize profits so
information sustainable and efficient decisions regarding route critical information is only available to affluent
systems choice and parking at a fleet level. users.
Sensitive data Personal data are carefully managed based on Data are used for commercial purposes. AVs collect
management general public interest. an abundance of sensitive private information.
Policies facilitate the conversion of parking Parking policies remain as they are, so parking
facilities into recreational, green, and building continues to consume valuable land that could be
Parking areas, or into active transport infrastructure. used for more sustainable or social purposes.
Curb access is efficiently managed to serve shared Curb space is congested and dangerous, and other
Curb Access vehicle passengers along with other uses. others (pedestrian and bicyclists) are harmed.
Land use Urban areas become more attractive places to Urban land is managed to accommodate AV travel,
policies live. Transport policies promote quality of life. to the detriment of other social groups.
Transport Transport networks are designed to be safe for all. Transport networks are restructured to
network Urban transport planning favors sustainable accommodate AVs’ needs. Other modes see no
design transport modes. comparable protection or investment.
Autonomous vehicles raise many policy and planning issues.
The Pedestrian and Bicycle Information Center identifies ten special risks that autonomous
vehicles can impose on pedestrians and cyclists, and how these can be minimized (PBIC 2017).
Appleyard and Riggs (2018) identify planning principles to ensure that autonomous vehicles
support community livability goals by improving driving behavior (slower speeds, and enhanced
ability to yield and stop), improving walking and bicycling conditions, and reducing parking
supply, but these will only occur if supported by suitable public policies.
There is much that policy makers and planners can do to maximize the benefits and minimize
the costs of autonomous vehicle implementation (Henaghan 2018; Largo, et al. 2018). As the
technology develops, transportation professionals should help establish performance
28
standards, analyze impacts, and support policies to minimize their costs and maximize their
benefits. Exhibit 23 identifies various planning implications of various planning needs and
requirements for autonomous vehicle development.
Exhibit 23 Autonomous Vehicle Planning Needs and Requirements

Impact Needs Requirements Time Period
Define performance, testing and data
Demonstrated functionality collection requirements for automated
Become legal and safety driving on public roads. 2018-25
Develop efficient curb and roadway
Develop policies to address management policies, such as curb
Address new conflicts increased curb and road regulations, congestion pricing and high-
and risks congestion risks. occupant vehicle priority policies. 2020-2040
Increase traffic Road lanes dedicated to Evaluate impacts. Define requirements.
density by vehicle vehicles with coordinated Identify lanes to be dedicated to vehicles
coordination platooning capability capable of coordinated operation. 2020-40
Independent mobility Fully autonomous vehicles Allows affluent non-drivers to enjoy
for non-drivers available for sale independent mobility. 2020-30s
Automated Moderate price premium. May provide demand response services
carsharing/taxi Successful business model. in affluent areas. Supports carsharing. 2030-40s
Independent mobility Affordable autonomous Reduced need for conventional public
for lower-income vehicles for sale transit services in some areas. 2040-50s
Reduced parking Major share of vehicles are
demand autonomous Reduced parking requirements. 2040-50s
Reduced traffic Major share of urban peak
congestion vehicle travel is autonomous. Reduced road supply. 2050-60s
Major share of vehicle travel Reduced traffic risk. Possibly increased
Increased safety is autonomous walking and cycling activity. 2040-60s
Energy conservation Major share of vehicle travel
and emission is autonomous. Walking and Supports energy conservation and
reductions cycling become safer. emission reduction efforts. 2040-60s
Improved vehicle Most or all vehicles are Allows narrower lanes and interactive
control autonomous traffic controls. 2050-70s
Need to plan for Major share of vehicles are More complex traffic. May justify
mixed traffic autonomous. restrictions on human-driven vehicles. 2040-60s
Most vehicles are
Mandated autonomous and large
autonomous vehicles benefits are proven. Allows advanced traffic management. 2060-80s
Autonomous vehicles will have various impacts on transportation planning.
29
The timeline below summarizes autonomous vehicle planning requirements.
Exhibit 24 Autonomous Vehicle Planning Requirement Time-Line
Develop performance Study, and where appropriate If autonomous vehicles prove

and data collection support, autonomous vehicle overall beneficial and are the
requirements for implementation for specific majority of vehicles, it may be
autonomous vehicles applications such as taxi, possible to change roadway
operating on public carsharing and demand design and management
roadways. response services. practices.
2018 2020s 2030s 2040s 2050s 2060s+ 
Support large-scale If autonomous vehicles If autonomous vehicles

autonomous vehicle testing. prove to be effective and prove to be very
Evaluate their benefits and common, consider beneficial, it may be
costs under actual operating dedicating some highway appropriate to restrict
conditions. lanes to their use. human-driving.
This timeline summarizes how autonomous vehicles are likely to impact transport planning.
Autonomous Taxi Service Impacts

In 2017 Waymo and Uber announced plans to start testing driverless taxis in the Phoenix, Arizona region.
(Bergen 2017; Lee 2017). Within a few months a pedestrian death put the program on hold, but it will
probably continue eventually. How soon and how much will these services affect overall travel?
Phoenix was chosen because it has mild climate, wide streets and relatively few pedestrians. The vehicles
are relatively slow. Further development and testing is required before the technology can expand to
cities with extreme weather or congestion, and its expansion will depend on the service’s profitability,
which will require high consumer confidence and satisfaction, and cost reductions. As a result, it will
probably take several years before commercial autonomous taxi services are widely available.
Taxis primarily serve local urban trips when travellers lack a personal vehicle, which represents a minor
portion of total travel. To significantly reduce vehicle travel and associated costs, autonomous taxis must
become inexpensive, ubiquitous and integrated with other mobility options so households can reduce
their vehicle ownership and rely on shared vehicle. This can be accelerated by public policies that
discourage private vehicle ownership and encourage sharing, such as reduced parking supply, High
Occupancy Vehicle Lanes, and convenient drop off/pick up areas.
This is consistent with predictions that during the 2020s, autonomous vehicles will have limited availability
and performance. If the technology improves and become affordable and reliable, so self-driving taxi
services to become profitable, they can expand to serve more areas and trip types. However, until most
households shift from owning vehicles to relying on shared mobility services, and until a greater share of
households live in compact and multimodal neighborhoods, the new generation of autonomous taxis will
affect only a small portion of total travel and provide modest community benefits.
30
Conclusions
Recent announcements that autonomous vehicles will soon be commercially available raise
hopes that these technologies will quickly solve many transportation problems. Some
advocates predict that by 2030 such vehicles will be sufficiently reliable and affordable to
displace most human-operated vehicles, providing many benefits to users and society overall.
However, there are good reasons to be skeptical.
There is considerable uncertainty concerning autonomous vehicle benefits, costs, travel

impacts, deployment speed and consumer demand. Driving a vehicle on public roads is
complicated due to the frequency of interactions with other, often-unpredictable objects
including vehicles, pedestrians, cyclists and animals. Most objective experts acknowledge that
significant progress is needed before autonomous vehicles can operate reliably under all
normal conditions, including mixed urban traffic, heavy rain and snow, unpaved and unmapped
roads, and poor Internet connections, plus time needed for regulatory approval. Autonomous
operation will probably add significant equipment, maintenance and mapping costs.
Most optimistic predictions are speculative and exaggerated, often made by people with
financial interests in the industry, based on experience with other disruptive technologies such
as personal computers, digital cameras and smart phones. Advocates often ignore significant
costs and risks, rebound effects (increased vehicle travel caused by faster travel or reduced
operating costs), and potential harms to non-users. Benefits are often double-counted, for
example, by summing increased safety, traffic speeds and facility savings, although these often
involve trade-offs. Vehicles typically last an order of magnitude longer, cost two orders of
magnitude more, impose greater external costs, and rely more on public infrastructure than
other technologies. As a result, vehicle innovations tend to take longer and involve more
regulation than most other new technologies.
These performance limitations and additional costs are likely to limit sales. Most motorists will
be reluctant to pay thousands of dollars extra for an autonomous vehicle that will sometimes
respond, “That is not a feasible destination,” due to poor weather conditions or unmapped
roads. If they follow previous vehicle technology deployment patterns, autonomous vehicles
will initially be costly and imperfect. During the 2020s and perhaps the 2030s, autonomous
vehicles will be expensive novelties, unable to operate in conditions such as heavy rain and
snow, unpaved roads, where GPS services and special maps are unavailable, and in mixed urban
traffic. They will be purchased by affluent non-drivers and people who frequently drive long
distance, but many travellers will not consider the extra costs justified. It will probably be the
late 2030s or 2040s before they become affordable to middle-income households, and later
before they are affordable to lower-income motorists.
During the 2020 and 2030s, self-driving taxi and “micro-transit” (van) services may become
available in many urban areas. They should be cheaper than human-operated taxis, costing
about $0.60-1.00 per mile for a self-driving taxi and 30-60¢ per mile for micro-transit, but offer
low service quality: to minimize cleaning and repair costs their interiors will be metal and
plastic, and occupants will be monitored by cameras, yet passengers will probably still find
31
previous occupants’ garbage, stains and odors. No drivers will be available to assist passengers
and provide security. Additional passengers will add pickup and drop-off delays, particularly in
lower-density areas. Because of these limitations, autonomous taxi and micro-transit will only
be suitable for a portion of travel, mainly local urban trips. Once the novelty wears off,
autonomous taxi use will seem as tedious as commercial airline travel.
Exhibit 25 illustrates a prediction of market penetration and benefits. This indicates that it will
be at least 2040 before half of new vehicles are autonomous, 2050 before half of the vehicle
fleet is autonomous, and possibly longer due to technical challenges or consumer preferences.
Exhibit 25 Autonomous Vehicle Sales, Fleet, Travel and Benefit Projections
Based on previous vehicle technology deployment patterns, this analysis predicts that it will be at least
2040 before half of all new vehicles are autonomous, 2050 before half of the vehicle fleet is autonomous,
and possibly longer due to technical challenges or consumer preferences. Significantly faster deployment
will require scraping many otherwise functional vehicles because they lack self-driving capability. Some
user benefits can occur when autonomous vehicles are relatively costly and rare, but many benefits, such
as independent mobility for moderate-income non-drivers, can only be significant if they become very
reliable and affordable, and some benefits, such as reduced traffic congestion and emissions, require
dedicated lanes to allow autonomous vehicle platooning.
A critical question is whether autonomous vehicles increase or reduce total vehicle travel and
associated external costs. It could go either way, depending on public policies. By allowing
vehicle travel by non-drivers, increasing travel convenience and comfort, and allowing vehicles
to drive in circles rather than pay for a parking space, they could increase total vehicle mileage
and traffic problems. Alternatively, they may also facilitate vehicle sharing, which allows
households to reduce vehicle ownership and therefore total driving.
32
Many motorists may prefer to own personal vehicles for prestige and convenience sake, or use
a combination of modes (walking, cycling, conventional public transit and taxi). As a result,
shared autonomous vehicles are likely to reduce vehicle travel mostly in compact urban areas,
and increase travel in suburban and rural areas. Since most North Americans live in suburban
areas, total vehicle travel will probably increase unless discouraged by public policies.
Another critical issue is the degree potential benefits can be achieved when only a portion of
vehicle travel is autonomous. Some benefits, such as improved mobility for affluent non-
drivers, may occur when autonomous vehicles are uncommon and costly, but many potential
benefits, such as reduced congestion and emission rates, reduced traffic signals and lane
widths, require that most or all vehicles on a road operate autonomously.
A key public policy issue is the degree that this technology may harm people who do not use
such vehicles, for example, increased traffic volumes and speeds that degrade walking and
cycling conditions, reduced investments in public transit, or requiring restrictions on human-
operated vehicles. Platooning benefits require dedicated autonomous vehicle lanes. These
issues will probably generate considerable debate over their merit and fairness.
To minimize problems and maximize benefits many experts recommend that public policies
protect pedestrians and bicyclists from new conflicts and risks, encourage shared and electric
autonomous vehicles, and limit total vehicle traffic particularly in denser urban areas. This can
be done with High Occupancy Vehicle (HOV) priority lanes which favors shared over single-
occupancy vehicles on congested roads, increased fossil fuel taxes, efficient road pricing,
convenient passenger pick-up and drop-off facilities, reduced parking requirements and
efficient parking pricing. Experts also recommend various technical requirements and
regulations to ensure that autonomous vehicles are programed to minimize risks and delay to
other people, particularly pedestrians, bicyclists and public transit users.
Autonomous vehicle implementation is just one of many trends likely to affect future transport
demands and costs, and therefore planning decisions, and not necessarily the most important.
Its ultimate impacts depend on how it interacts with other trends, such as shifts from personal
to shared vehicles. It is probably not a “game changer” during most of our professional lives,
and is only a “paradigm shift” to the degree that this technology supports shifts to more
efficient and multimodal transport planning.
Transportation professionals (planners, engineers and policy analysts) have important roles to
play in autonomous vehicle development and deployment. We can help define the
performance standards they must meet to legally operate on public roads. We should evaluate
the risks and opportunities they present, and develop policies to ensure that their deployment
supports strategic community goals including congestion reduction, public safety and health,
and improved opportunity for disadvantaged people. Once they become more common they
may affect road, parking and public transit planning decisions.
33
References
AAA (2017), Your Driving Costs, American Automobile Association (http://newsroom.aaa.com); at

https://publicaffairsresources.aaa.biz/YDC/html5/index.html
Evan Ackerman (2017), “Toyota's Gill Pratt on Self-Driving Cars and the Reality of Full Autonomy,”
Spectrum, International Institute of Electrical Engineers (www.ieee.org); at https://bit.ly/2FJYJax.
James M. Anderson, et al. (2014), Autonomous Vehicle Technology: A Guide for Policymakers, RAND
Corporation (www.rand.org); at https://bit.ly/1F9o70b.
APA (2016), Autonomous Vehicles: Knowledgebase Collection, American Planning Association

(www.planning.org); at www.planning.org/knowledgebase/autonomousvehicles.
Bruce Appleyard and William Riggs (2018), 10 Principles Toward More Sharing and Less Sprawl: A Manifesto
for Street Livability, Health, and Humanity in the Era of Driverless Cars, Planetizen (www.planetizen.com); at
www.planetizen.com/node/96769.
Jaâfar Berrada and Fabien Leurent (2017), “Modeling Transportation Systems involving Autonomous
Vehicles: A State of the Art,” Transportation Research Procedia, Vol 27, pp. 215-221
(https://doi.org/10.1016/j.trpro.2017.12.077).
Mark Bergen (2017), “Alphabet Launches the First Taxi Service with No Human Drivers,” Bloomberg
Technology (www.bloomberg.com); at https://bloom.bg/2Ea2dml.
Brad Berman (2011), History of Hybrid Vehicles, Hybrid Cars (www.hybridcars.com); at

https://bit.ly/2z9aSEg.
Patrick Bösch, Felix Becker, Henrik Becker and Kay W. Axhausen (2017), Cost-based Analysis of Autonomous
Mobility Services, Working Paper 1225, Institute for Transport Planning and Systems (www.ivt.ethz.ch),
Swiss Federal Institute of Technology; at www.ivt.ethz.ch/institut/vpl/publikationen/papers/1225.html.
Meredith Broussard (2018), “The Dirty Truth Coming for Self-Driving Cars: Trash. Odors. Bodily Fluids. Will
Autonomous Rideshares be Ready for our Mess?” Slate (https://slate.com); at https://slate.me/2Ls9IrI.
Tristan Cathers (2014), When Will You Be Able To Buy A Driverless Car?, Mojo Motors
(www.mojomotors.com); at www.mojomotors.com/blog/when-will-you-be-able-to-buy-a-driverless-car.
Angela DeLuca (2018), Ending the Search for Parking, Urbanism Next
(https://urbanismnext.uoregon.edu); at https://bit.ly/2q6uTWD.
Sarah E. DeWitt (2015), Driverless Cars Pose No Immediate Threat to Personal Auto Insurers, North America
Equity Research, J.P. Morgan Securities LLC (www.jpmorgan.com).
Stephen Dirksen (1997), Air Bags: History of American Technology, Bryant University Community Web
(http://web.bryant.edu/~ehu/h364proj/sprg_97/dirksen/airbags.html).
34
Julian Ebert (2016), Reinventing the Human Brain: How A.I. Will Revolutionize Driverless Cars, 2025 AD
(www.2025ad.com); at www.2025ad.com/latest/driverless-cars-and-artificial-intelligence.
John Eddy and Ryan Falconer (2017), A Civil Debate: Are Driverless Cars Good for Cities?, Doggerel
(http://doggerel.arup.com); http://doggerel.arup.com/a-civil-debate-are-driverless-cars-good-for-cities.
Stephen Edelstein (2018), Reports Highlight Autonomous Cars' Shortcomings: Self-Driving Cars are Still
Having Trouble Dealing with Real-World Traffic, The Drive (www.thedrive.com); at
www.thedrive.com/tech/20561.
Daniel J. Fagnant and Kara M. Kockelman (2013), Preparing a Nation for Autonomous Vehicles:
Opportunities, Barriers and Policy Recommendations, Eno Foundation (www.enotrans.org); at
www.enotrans.org/wp-content/uploads/wpsc/downloadables/AV-paper.pdf.
Fehr & Peers (2014), Effects of Next-Generation Vehicles on Travel Demand & Highway Capacity, Fehr and
Peers (www.fehrandpeers.com); at www.fehrandpeers.com/fpthink/nextgenerationvehicles.
FHWA (2015), Highway Statistics, Federal Highway Administration (www.fhwa.dot.gov); at

www.fhwa.dot.gov/policyinformation/statistics.cfm.
Kelly Fleming and Mark Singer (2019), Energy Implications of Current Travel and the Adoption of Automated
Vehicles, National Renewable Energy Laboratory; at www.nrel.gov/docs/fy19osti/72675.pdf.
Lew Fulton, Jacob Mason and Dominique Meroux (2017), Three Revolutions in Urban Transportation, UC
Davis and ITDP (www.itdp.org); at www.itdp.org/publication/3rs-in-urban-transport.
GAO (2016), Vehicle Cybersecurity: DOT and Industry Have Efforts Under Way, but DOT Needs to Define Its
Role in Responding to a Real-world Attack, General Accounting Office (www.gao.gov); at
https://bit.ly/1ruZi09.
GM (2013), OnStar Services (www.onstar.com)
Lee Gomes (2014), “Hidden Obstacles for Google’s Self-Driving Cars,” MIT Technological Review,
(www.technologyreview.com), 28 August 2014; at https://bit.ly/2B6BUxx.
David G. Groves and Nidhi Kalra (2017), Enemy of Good: Autonomous Vehicle Safety Scenario Explorer, Rand
Corporation www.rand.org); at www.rand.org/pubs/tools/TL279.html.
Bern Grush (2016), Driverless Cars Ahead: Ontario Must Prepare for Vehicle Automation, Residential and
Civil Construction Alliance of Ontario (RCCAO); at https://bit.ly/2DFNNwy.
Bern Grush and John Niles (2018), The End of Driving: Transportation Systems and Public Policy Planning for
Autonomous Vehicles, Elsevier (www.elsevier.com/books/the-end-of-driving/niles/978-0-12-815451-9).
Erick Guerra (2015), “Planning for Cars That Drive Themselves: Metropolitan Planning Organizations,
Regional Transportation Plans, and Autonomous Vehicles,” Journal of Planning Education and Research, pp.
1–15 (DOI: 10.1177/0739456X15613591); at http://bit.ly/1RqcBaZ.
35
James R. Healey (2012), “Stick Shifts Popular Again, Despite Lower Gas Mileage,” USA Today, 30 April
(www.usatoday.com); at https://bit.ly/2FoFPdc.
Alexandro Henao and Wes Marshall (2018), “The Impact of Ride-hailing on Vehicle Miles Traveled,”
Transportation (https://doi.org/10.1007/s11116-018-9923-2).
Mark Harris (2014), “FBI Warns Driverless Cars Could Be Used As 'Lethal Weapons'” The Guardian
(www.theguardian.com), 16 July 2014; at https://bit.ly/1p7fQU6.
Andrew Hawkins (2017), “Tesla’s Autopilot is Supposed to Deliver Full Self-Driving, So Why Does it Feel
Stuck in the Past?,” The Verge (www.theverge.com); at https://bit.ly/2Fmtfed.
Jennifer Henaghan (2018), Preparing Communities for Autonomous Vehicles, American Planning Association
(www.planning.org); at https://bit.ly/2piKBhX.
Jeremy Hsu (2017), “When It Comes to Safety, Autonomous Cars Are Still ‘Teen Drivers,’” Scientific
American (www.scientificamerican.com); at http://bit.ly/2j9gFPT.
ITF (2014), Urban Mobility: System Upgrade, International Transport Forum

(www.internationaltransportforum.org) and Corporate Partnership Board; at https://bit.ly/2JJrWUo.
ITF (2018), Safer Roads with Automated Vehicles? International Transport Forum (www.itf-oecd.org); at
www.itf-oecd.org/sites/default/files/docs/safer-roads-automated-vehicles.pdf.
Tay Hong Chuen, et al. (2013), Autonomous Vehicles, MT5009; at https://bit.ly/2RTGdS7.
Charlie Johnston and Jonathan Walker (2017), Peak Car Ownership: The Market Opportunity for Electric
Automated Mobility Services, Rocky Mountain Institute (www.rmi.org); at http://bit.ly/2rhJRNi.
Kaohsiung EcoMobility Festival (2017), Kaohsiung Strategies for the Future of Urban Mobility,
EcoMobility Festival (www.ecomobilityfestival.org); at https://bit.ly/2OMwsDD.
Tasha Keeney (2017), Mobility-As-A-Service: Why Self-Driving Cars Could Change Everything, ARC
Investment Research (http://research.ark-invest.com); at http://bit.ly/2xz6PNV.
Irem Kok, et al. (2017), Rethinking Transportation 2020-2030: The Disruption of Transportation and the
Collapse of the Internal-Combustion Vehicle and Oil Industries, RethinkX (www.rethinkx.com); at
http://bit.ly/2pL0cZV.
Kara Kockelman, et al. (2016), Implications of Connected and Automated Vehicles on the Safety and
Operations of Roadway Networks, University of Texas Center for Transportation Research
(http://ctr.utexas.edu), for Texas DOT; at http://library.ctr.utexas.edu/ctr-publications/0-6849-1.pdf.
Kara M. Kockelman and Stephen D. Boyles (2018), Smart Transport for Cities & Nations: The Rise of Self-
Driving & Connected Vehicles, The University of Texas at Austin (www.caee.utexas.edu); at
www.caee.utexas.edu/prof/kockelman/public_html/CAV_Book2018.pdf.
36
Philip Koopman and Michael Wagner (2017), “Autonomous Vehicle Safety: An Interdisciplinary Challenge,”
IEEE Intelligent Transportation Systems, Vol. 9, No. 1; at http://ieeexplore.ieee.org/document/7823109.
Nico Larco, et al. (2018), AVs in the Pacific Northwest: Reducing Greenhouse Gas Emissions in a Time of
Automation, Urbanism Next Center (https://urbanismnext.uoregon.edu): at https://bit.ly/2MHIXix.
Steve Lawson (2018), Tackling the Transition to Automated Vehicles, Roads that Cars Can Read Report III,
European Road Assessment Association (www.eurorap.org); at https://bit.ly/2IrYTTQ.
Timothy Lee (2017), “Fully Driverless Cars Could be Months Away. Google's Self-driving Car Unit Prepares to
Launch a Taxi Service Near Phoenix,” ArsTechnica (https://arstechnica.com); at http://bit.ly/2DkP3FO.
David Levinson (2015), “Climbing Mount Next: The Effects of Autonomous Vehicles on Society,” Minnesota
Journal of Law Science and Technology, Vo. 16, No. 2, pp. 787-809; at https://bit.ly/2Dmql6J.
Scott Le Vine, Alireza Zolfaghari and John Polak (2015), “Autonomous Cars: The Tension Between
Occupant-Experience and Intersection Capacity,” Transportation Research Part C: Emerging Technologies
(www.journals.elsevier.com/transportation-research-part-c-emerging-technologies).
Jamie Lendino (2012), “The History of Car GPS Navigation,” PC Magazine (www.pcmag.com); 16 April; at
www.pcmag.com/article2/0,2817,2402755,00.asp.
Todd Litman (2009), Transportation Cost and Benefit Analysis, Victoria Transport Policy Institute
(www.vtpi.org/tca).
Todd Litman (2013), “The New Transportation Planning Paradigm,” ITE Journal (www.ite.org), Vo. 83, No. 6,
pp. 20-28, 2013; at www.vtpi.org/paradigm.pdf.
Todd Litman (2014), “Ready or Waiting,” Traffic Technology International

(www.traffictechnologytoday.com), January, pp. 36-42; at http://www.vtpi.org/AVIP_TTI_Jan2014.pdf.
Todd Litman (2017), Presentation to the Canadian Standing Senate Committee on Transport and
Communication Concerning Connected and Self-driving Vehicles (http://bit.ly/2u2Grhe). Hearing page:
https://sencanada.ca/en/Committees/TRCM/NoticeOfMeeting/449871/42-1.
Kristin Lovejoy, Susan Handy and Marlon G. Boarnet (2013), Technical Background Document on Impacts of
Carsharing, California Air Resources Board (www.arb.ca.gov); at https://bit.ly/2P3tlqM.
Jerome M. Lutin, Alain L. Kornhauser and Eva Lerner-Lam (2013), “The Revolutionary Development of Self-
Driving Vehicles and Implications for the Transportation Engineering Profession,” ITE Journal, July, pp. 21-
26; at https://bit.ly/2DkRXt4.
Mohammad Lavasani and Xia Jin (2016), “Market Penetration Model for Autonomous Vehicles on the Basis
of Earlier Technology Adoption Experience,” Transportation Research Record, No. 2597, pp. 67–74
(DOI: 10.3141/2597-09).
Marowits (2017), “Self-driving Ubers Could Still be Many Years Away, Says Research Head,” CTV News
(www.ctvnews.ca); at http://bit.ly/2Dl48Y7.
37
McKinsey (2016), Automotive Revolution – Perspective Towards 2030: How the Convergence of
Disruptive Technology-driven Trends Could Transform the Auto Industry (www.mckinsey.de); at
https://bit.ly/2zYBTfG.
Jeffrey Mervis (2017), “Are We Going Too Fast on Driverless Cars?,” Science Magazine
(www.sciencemag.org); at www.sciencemag.org/news/2017/12/are-we-going-too-fast-driverless-cars.
Dimitris Milakis, Bart van Arem and Bert van Wee (2017), “Policy and Society Related Implications of
Automated Driving: A Review of Literature and Directions for Future Research,” Journal of Intelligent
Transportation Systems, Vol. 21, No. 4, pp. 324–348; at https://bit.ly/2zSSOgZ.
Adam Millard-Ball (2016), “Pedestrians, Autonomous Vehicles, and Cities,” Journal of Planning Education
and Research, pp. 1-7 (DOI: 10.1177/0739456X16675674); at https://bit.ly/2hhYrxV.
John S. Miller and Di Kang (2019), Ways to Consider Driverless Vehicles in Virginia Long-Range Travel
Demand Models, Virginia Transportation Research Council (www.virginiadot.org/vtrc); at
www.virginiadot.org/vtrc/main/online_reports/pdf/19-R11.pdf.
Jim Motavalli (2012), “Self-Driving Cars Will Take Over by 2040,” Forbes Magazine, 25 Sept. 2012; at
www.forbes.com/sites/eco-nomics/2012/09/25/self-driving-cars-will-take-over-by-2040
NCHRP (2018), Impacts of Connected Vehicles and Automated Vehicles on State and Local Transportation
Agencies, National Cooperative Highway Research Program (http://trb.org); at https://bit.ly/2dwWddo.
NHTSA (2013), Preliminary Statement of Policy Concerning Automated Vehicles, National Highway Traffic
Safety Administration (www.nhtsa.gov).
NYT (2017), “How Will Sex, Death and Liability Change on the Road to the Driverless Revolution? New York
Times Magazine (www.nytimes.com); at https://nyti.ms/2DEvojT.
OECD/ITF (2018), The Shared-use City: Managing the Curb, Organization for Economic Cooperation and
Development and the International Transport Forum (www.itf-oecd.org); at https://bit.ly/2B8hLqG.
Alan Ohnsman (2014), “Automated Cars May Boost Fuel Use, Toyota Scientist Says,” Bloomberg Press, 16
July 2014 (www.bloomberg.com); at https://bloom.bg/2mDkmAu.
ORNL (2012), Transportation Energy Book, Oak Ridge National Lab. (www-cta.ornl.gov/data), USDOE.
Enrica Papa and António Ferreira (2018), “Sustainable Accessibility and the Implementation of Automated
Vehicles: Identifying Critical Decisions,” Urban Science, Vol. 2, No. 1 (doi:10.3390/urbansci2010005); at
https://bit.ly/2DHjZQz.
PBIC (2017), Automated and Connected Vehicles, Pedestrians, and Bicyclists, Pedestrian and Bicycle
Information Center (www.pedbikeinfo.org); at www.pedbikeinfo.org/AV.
Caroline Rodier (2018a), The Effects of Ride Hailing Services on Travel and Associated Greenhouse Gas
Emissions, UC Davis Institute for Transportation Studies (https://ncst.ucdavis.edu); at https://bit.ly/2qTLXja.
38
Caroline Rodier (2018b), Travel Effects and Associated Greenhouse Gas Emissions of Automated Vehicles,
UC Davis Institute for Transportation Studies (https://ncst.ucdavis.edu); at https://bit.ly/2w1rVsK.
Shelley Row (2013), “The Future of Transportation: Connected Vehicles to Driverless Vehicles…What Does It
Mean To Me?” ITE Journal (www.ite.org), Vol. 83, No. 10, pp. 24-25.
SAE (2014), Levels of Driving Automation Are Defined In New SAE International Standard J3016, Society of
Automotive Engineers (www.sae.org); at www.sae.org/misc/pdfs/automated_driving.pdf.
Bruce Schaller (2017), Empty Seats, Full Streets: Fixing Manhattan's Traffic Problem, Schaller Consulting
(http://schallerconsult.com); at http://schallerconsult.com/rideservices/emptyseats.pdf.
Susan Shaheen, Hannah Totte and Adam Stocker (2018), Future of Mobility White Paper, ITS Berkeley; at
https://escholarship.org/uc/item/68g2h1qv.
Shared Mobility Principles for Livable Cities (www.sharedmobilityprinciples.org). Principles developed by a

working group of international NGOs are designed to guide decision-makers and stakeholders toward the
best outcomes for all in the transition to new mobility options.
Marc Schlossberg, et al. (2018), Rethinking the Street in an Era of Driverless Cars, Urbanism Next Research
(www.urbanismnext.com); at www.urbanismnext.com/s/Rethinking_Streets_AVs_012618-27hcyr6.pdf.
Brandon Schoettle and Michael Sivak (2014), A Survey Of Public Opinion About Autonomous And Self-
Driving Vehicles In The U.S., The U.K., And Australia, Report UMTRI-2014-21, Transportation Research
Institute, University of Michigan (www.umich.edu/~umtriswt).
Brandon Schoettle and Michael Sivak (2015a), A Preliminary Analysis Of Real-World Crashes Involving Self-
Driving Vehicles, Report UMTRI-2015-34, Transportation Research Institute, University of Michigan
(www.umich.edu/~umtriswt); at http://umich.edu/~umtriswt/PDF/UMTRI-2015-34.pdf.
Brandon Schoettle and Michael Sivak (2016), Motorists' Preferences for Different Levels of Vehicle
Automation, Transportation Research Institute, University of Michigan (www.umich.edu/~umtriswt).
Ben Schonberger and Steve Gutmann (2013), A Self-Driving Future: At the Intersection of Driverless Cars
and Car Sharing, Sightline Institute (www.sightline.org); at http://bit.ly/2Bt9ue4.
Tom Simonite (2016), “Prepare to be Underwhelmed by 2021’s Autonomous Cars: Ford, Uber, and BMW
Promise Fully Self-Driving Cars in Five Years—But They Will Probably Only Work in Very Limited Areas,” MIT
Technology Review (www.technologyreview.com); at https://bit.ly/2PZxH6X.
Michael Sivak and Brandon Schoettle (2015a), Road Safety With Self-Driving Vehicles: General Limitations
And Road Sharing With Conventional Vehicles, Sustainable Worldwide Transportation Program
(www.umich.edu/~umtriswt), University of Michigan.
Michael Sivak and Brandon Schoettle (2015), Potential Impact of Self-Driving Vehicles on Household Vehicle
Demand and Usage, Sustainable Worldwide Transportation Program (www.umich.edu/~umtriswt),
39
University of Michigan. Also see, Influence of Current Nondrivers on the Amount of Travel and Trip Patterns
with Self-Driving Vehicles, at www.umich.edu/~umtriswt/PDF/UMTRI-2015-39_Abstract_English.pdf.
Michael Sivak and Brandon Schoettle (2015c), Influence of Current Nondrivers on the Amount of Travel and
Trip Patterns with Self-Driving Vehicles, Sustainable Worldwide Transportation Program
(www.umich.edu/~umtriswt), University of Michigan; at http://bit.ly/2BrEHxV.
Jeff Speck (2017), Autonomous Vehicles, United Conference of Mayors; at

www.youtube.com/watch?v=2kBEvg8bftE.
Daniel Sperling (2017), Three Revolutions: Steering Automated, Shared, and Electric Vehicles to a Better
Future, Island Press (https://islandpress.org).
T. Stephens, et al. (2016), Estimated Bounds and Important Factors for Fuel Use and Consumer Costs of
Connected and Automated Vehicles, Technical Report, National Renewable Energy Laboratory
(www.nrel.gov); at www.nrel.gov/docs/fy17osti/67216.pdf.
Araz Taeihagh and Hazel Si Min Lim (2018), “Governing Autonomous Vehicles: Emerging Responses for
Safety, Liability, Privacy, Cybersecurity, and Industry Risks,” Transport Reviews, at https://bit.ly/2DYQaLd.
Morteza Taiebat, Samuel Stolper and Ming Xu (2019), “Forecasting the Impact of Connected and
Automated Vehicles on Energy Use: A Microeconomic Study of Induced Travel and Energy Rebound,”
Applied Energy, Vol. 247, pp 297-308 (https://doi.org/10.1016/j.apenergy.2019.03.174).
Stefan Trommer, et al. (2016), Autonomous Driving: The Impact of Vehicle Automation on Mobility
Behaviour, Institute of Transport Research (www.ifmo.de); at http://bit.ly/2kIAOOQ.
TRB (2017), Strategies to Advance Automated and Connected Vehicles, Transportation Research Board
(www.trb.org); at www.nap.edu/download/24873.
TRB (2019), Socioeconomic Impacts of Automated and Connected Vehicles, European Commission and the
Transportation Research Board (www.trb.org); at www.trb.org/Publications/Blurbs/178576.aspx.
Wharton (2017), The Road Ahead for Connected Vehicles, Wharton School of Management
(http://wharton.upenn.edu); at http://whr.tn/2BqKIuT.
WSJ (2017), “Why Your Next Car May Look Like a Living Room,” Wall Street Journal (www.wsj.com); at
http://on.wsj.com/2tlCvYp.
www.vtpi.org/avip.pdf
40

Autonomous Vehicle Implementation Predictions: Implications For Transport Planning

Uploaded by

Copyright:

Available Formats

Autonomous Vehicle Implementation Predictions: Implications For Transport Planning

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Autonomous Vehicle Implementation Predictions: Implications For Transport Planning

Uploaded by

Copyright:

Available Formats

www.vtpi.

Autonomous Vehicle Implementation Predictions

Waymo’s self-driving taxis are a well-publicized example of autonomous vehicles.

Todd Alexander Litman © 2013-2019

Autonomous Vehicle Operational Models

Exhibit 1 Automated Driving Levels (SAE J3016 2014)

The SAE defines five vehicle

Exhibit 2 summarizes three autonomous vehicle operational models.

Table 2 Autonomous Vehicle Operational Models Compared

Benefits and Costs

Reduced Stress, Improved Productivity and Mobility

Exhibit 3 Productivity and Relaxation While Travelling

The Autonomous Vehicle Travel Experience

Ownership and Operating Costs

Exhibit 4 Autonomous Vehicle Equipment and Service Requirements

Cleaning – An Often Overlooked Cost

Exhibit 5 Automated Versus Personal Car Costs

Johnston and Walker (2017)

Exhibit 6 Cost Comparison

Traffic Safety and Security

Exhibit 7 Driverless Car “Platooning” (Chuen, et al. 2013)

Benefit and Cost Summary

Exhibit 8 Autonomous Vehicle Potential Benefits and Costs

Additional risks. May increase risks to other road users and

insurance costs. May reduce high-risk driving.

Development and Deployment Predictions

Exhibit 9 Aircraft and Automobile Software Code Compared (GAO 2016)

In addition to technological progress, market deployment depends on consumer demand:

Experience with Previous Vehicle Technology Deployment

Exhibit 11 Vehicle Technology Deployment Summary

Exhibit 10 U.S. Population and Automobile Ownership (FHWA 2016)

200 0.6 the next half century the

Innovation Deployment Patterns

Exhibit 12 Innovation S-Curve

20% Commercial release

Exhibit 13 Autonomous Vehicle Market Penetration Projections

Exhibit 14 illustrates these deployment rates.

Exhibit 15 Autonomous Vehicle Impacts on Total Vehicle Travel

Exhibit 16 Personal Versus Shared Vehicles

Exhibit 17 Autonomous Vehicle Scenario Summary

Exhibit 18 Autonomous Vehicle Impacts on Various Travel Demands

Exhibit 19 Summary of Findings and Quality of Evidence (based on Rodier 2018b)

Potential Conflicts and Solutions

Shared Mobility Principles for Livable Cities (www.sharedmobilityprinciples.org)

Exhibit 20 Factors Affecting Transport Demands and Costs

Demographic Trends Price Changes Improved Travel Options

Changing User Preferences Intelligent Transport Systems (ITS)

Exhibit 21 Autonomous Vehicle Benefits

Level 1-3 personal vehicles  ?

Level 4 + 5 personal vehicles    

Shared autonomous vehicles   

Shared autonomous rides   

Exhibit 22 summarizes key autonomous vehicle planning issues.

Exhibit 23 Autonomous Vehicle Planning Needs and Requirements

The timeline below summarizes autonomous vehicle planning requirements.

Exhibit 24 Autonomous Vehicle Planning Requirement Time-Line

Develop performance Study, and where appropriate If autonomous vehicles prove

2018 2020s 2030s 2040s 2050s 2060s+ 

Support large-scale If autonomous vehicles If autonomous vehicles