The Transportation Futures
Project: Planning for Technology
Change
Final Report
Prepared by:
David Levinson
Adam Boies
Department of Civil,
Environmental and Geo- Engineering University of Minnesota
Jason Cao
Yingling Fan
Humphrey School of Public
Affairs University of Minnesota
January 2016
Published by:
Minnesota Department of Transportation
Research Services & Library
395 John Ireland Boulevard, MS 330
St. Paul, Minnesota 55155-1899
This report represents the results of research conducted by
the authors and does not necessarily represent the views or policies of the
Minnesota Department of Transportation or the University of Minnesota. This
report does not contain a standard or specified technique.
The authors, the Minnesota
Department of Transportation and the University of Minnesota do not endorse
products or manufacturers. Trade or manufacturers’ names appear herein solely
because they are considered essential to this report.
Executive Summary
The next two decades will see more change in the
transportation sector than have been seen in 100 years. The introduction of
autonomous vehicles, from a few cars initially, to all new cars, to eventually
all cars will radically change how transportation is used. The concomitant
electrification of vehicles will provide further opportunities to better
optimize the use of transportation systems. Finally, continuing advances in
information and mobile communications technology will up-end the way people
think about transportation systems. This report explores in eight chapters the
changes that are coming.
Fully autonomous test vehicles from automakers and new
entrants like Google have traveled in general traffic over 1 million miles
collectively. Semi-autonomous vehicles
are already here. Tesla auto-pilot (available in about 100,000 cars), for
instance, can both keep lanes and follow the car in front, in addition to
automatically changing lanes with direction from the driver. Tesla cars drive
over 1 million miles per day nationally, though the amount of that in
semi-autonomous mode is proprietary. The transition from human driven vehicles
to fully autonomous vehicles is tricky. Some automakers believe that
incremental transition is viable. Others note the danger of semiautonomous
vehicles that require periodic human intervention, and argue instead for
step-jump to fully autonomous vehicles. The impacts described below are
associated with fully autonomous vehicles (no driver control).
We anticipate the following
timeline for the deployment of fully autonomous
vehicles (chapter
1):
• 2020
market availability,
• 2030
regulatory requirement for all new cars,
• 2040
prohibition of non-autonomous vehicles from public roads at most times.
The consequences of fully autonomous vehicles are
numerous. Some of the more important ones are listed below and discussed more fully
in the report.
• Increased
safety overall as driverless cars don’t get tired and have better sensors and
algorithms than humans. If driverless cars are not significantly safer, they
will not be permitted. Total fatalities may drop over 90% with driverless
vehicles as human error is eliminated. No system can be perfectly safe, but it
will be significantly safer. Road designs and sight-lines will be far less
relevant than design criteria as a result.
• An
explosion of vehicle forms, including new, narrow, single-passenger vehicles,
which will be safer than motorcycles given their automated drivers and new
structural designs enabled by electrification. People will feel more
comfortable in small vehicles mixed with large vehicles if all are automated.
• Increased
capacity from existing pavements as cars can follow with a shorter headway and
can occupy narrower lanes. This implies far more capacity in existing lanes,
and less need to expand roads.
• Higher
speeds on limited access roadways, as driver comfort with car-following and
speed is no longer determinative of the maximum speed of travel.
• Lower
speeds on local streets as automated vehicles better obey traffic laws and slow
down to avoid collisions with other road users like pedestrians and bicyclists.
• Vehicles
moving without people. After dropping off passengers, vehicles will redeploy to
park or to pick up other passengers, meaning there will be many unoccupied
vehicles on the road. Freight and delivery vehicles may similarly be
unoccupied. Unoccupied vehicles have less need for speed than vehicles carrying
people, creating opportunities for differentiating network speeds.
• Mobility
for everyone. Children, disabled persons, and others who today cannot drive
will be able to achieve the same level of mobility as others with the full
deployment of automated vehicles, especially mobility as a service.
• Lowered
vehicle costs (for all vehicles as all user-facing vehicle control equipment is
eliminated -- saving money -- even as
new vehicle sensors are added)
• Lowered
vehicle insurance costs (as crash insurance is offered by vehicle
manufacturers)
• Lowered
vehicle repair costs (as crashes, particularly small property damage crashes
are reduced and vehicles are simplified with electrification)
• Lowered
labor costs (for transit, taxis, freight) as all vehicle types are automated.
This implies these modes will be more price competitive than presently.
• Retrofitting
rights-of-way so that small lightweight neighborhood electric vehicles don’t
need to mix with heavyweight trucks and large cars.
• Roadspace
reallocation so that lanes no longer needed for moving or storing cars can be
used for other purposes (bike lanes, exclusive transit lanes, linear parks).
• Increased
ability to use time for non-driving tasks (see Chapter 2), which implies both
bigger and smaller vehicles
• Increased
willingness to travel longer distances. In-vehicle time becomes more useful,
and therefore less likely to be avoided. The saved travel time and the
increased utility of travel are likely to encourage visits to more distant but
more attractive destinations.
• Increased
gender equality as household chores like shopping and pick-up/drop-off services
are increasingly automated.
• Increased
willingness to live farther out. People will be more likely to make housing
location choices based on their residential preferences (such as school
quality, neighborhood security, neighborhood cohesion, etc.) than spatial
accessibility
The ownership structure of
automobiles will also change in coming years as Mobility-as-a-Service
(MaaS)
(Chapter 4) becomes more prominent. Sharing implies a reduction in auto
ownership (increased mobility-as-a-service) in cities as car-sharing (Car2Go,
Zipcar, Hourcar) and ridesharing (Uber, Lyft, taxi) converge into a single
driverless service that provides the right-size car for a given trip on a
per-trip basis. While the degree to which people will give up the on-demand
convenience of owning a car is unclear, it is far more likely in large cities
where people rent apartments and car ownership is a hassle, than in rural
areas, where response times of car rental will be larger. MaaS has a number of
implications:
• The
average age of the car will be younger, as shared vehicles are utilized more
hours per day and turn-over more quickly. Cars become more like phones and less
like long-lasting durable goods.
• The
average size of car will be smaller as firms can right-size the fleet for
demand, in contrast with privately owned cars, which are typically sized for
extreme or unusual uses, rather than the daily one- or two-person trips.
• MaaS
customers will travel less frequently than those who own cars, as they will pay
outof-pocket for capital costs each trip, while those who own cars forget about
the sunk cost of ownership, which is paid for independent of the number of
trips made.
• Streets
will need to be redesigned to favor loading and unloading passengers, rather
than on-street parking.
• Sharing
implies an increased willingness to live in cities, which will be cleaner,
safer, and more accessible with electric, automated, and shared vehicles
respectively.
Information and
communications technologies (Chapter 3) are changing travel demand
patterns. Work at home, now at 4.4 percent, is rising, and while unlikely to
replace all or even most work outside the home in the next two to three decades
(when still fewer than 10 percent of workers are likely to work at home), it
can certainly substitute in significant ways for many information economy jobs,
and for the information-rich components of traditional jobs. Part-time
telecommuting can reduce peak travel, both by shifting the time-of-day when
commutes occur and avoiding it on select days altogether. Online shopping
continues to grow, and is now about 8% of retail sales, and it could continue
to rise to upward of 50% of retail activity, leading to a substitution of
delivery for many more shopping tasks. The rise of virtual connectivity has
occurred at the same time that the amount of in-person interaction has fallen
in the past decade.
Yet, information and communication
technologies (ICT) not only reduce travel and but also induce new travel. For
telecommuting, the key findings include the following:
• Telecommuting
reduces commute travel during both peak and non-peak hours;
• Telecommuting
enables commuters to move farther away from their employment location and
become even more auto-dependent;
• Telecommuting
increases non-work travel, which takes place mostly close to home;
• Telecommuting
reduces vehicle miles traveled (VMT) slightly, but it helps mitigate the growth
of congestion on freeways;
For e-shopping, the literature
shows that
• Online
searching is positively associated with store shopping and people who buy
online also buy in person more;
• Studies
are mixed on whether e-shopping reduces travel to stores and other leisure
activities in the short term;
• E-shopping
for now digital products (books, records, videos) has already changed retail
patterns and shopping travel behavior;
• Online
buying increases delivery traffic and freight transportation;
• Existing
studies are based on the small share of e-shopping in retail industries. If its share is large enough to change the
distribution of commercial land uses in the region, e-shopping will have a
profound effect on shopping-related travel.
ICT are often promoted as a virtue alternative to
physical travel, but transportation planners should be realistic about the
relationships between ICT and travel: Although the short-term effect of ICT on travel
may be substitution, in the long term, travel demand has historically grown as
ICT demand increases.
New sensors
(Chapter 5) attached to the vehicle, person, and roadway will create increasing
streams of information about real-time conditions on all transportation
systems. This should have numerous applications, for instance, enabling
transportation agencies to improve traffic signal timing, and better matching
of supply to demand. Connected vehicles are coming independent of automated
vehicles. Whether the infrastructure providers add intelligence to their road
and signal systems (for instance, telling vehicles when the light is about to
change) is an open question.
The potential transition away from gasoline is another
important change confronting the transportation sector (Energy -Chapter 6). The timeline for electrification is similar but
slower than that for automation. Though automated vehicles need not be
electric, and electric vehicles need not be automated, we expect these systems
to track and both see increasing deployment. If current trends hold, electric
vehicles (EVs) may make up 68% of new car sales by 2050. This number is highly
dependent on gasoline prices and environmental regulations. Minnesota will
likely lag the US as the cold weather is less conducive to EVs than the US as a
whole.
Electricity generation costs are dropping, as are
battery storage costs. There are new opportunities for in-roadway electric
charging (dynamic wireless power transfer), probably beginning with buses at
bus stops, that should be explored by transportation agencies. The advantage of
such charging systems are a reduction in on-board battery storage weight
required, which greatly improves vehicle efficiency (since energy is not
consumed moving around stored energy).
Gasoline remains the fuel to beat, and if gasoline costs remain low,
electric vehicle deployment will be slower. Other fuels like methanol have an
opportunity to become more significant,
especially for truck fleets, for which electrification is much less efficient.
Urban fleets with a lot of stop-start activity may see hybrid electric
vehicles.
We anticipate a reduction in energy consumption overall
per distance traveled with reductions in vehicle weight for passenger cars and
more efficient use of trucks (which are likely to get heavier, as they carry
larger loads).
Biofuel use for surface
transportation is likely to plateau near existing use levels; however, it may
increasingly be used in the electricity sector (and thus indirectly for an
increasingly electrified transportation sector).
Importantly, a reduction in gasoline consumption has
large implications for transport financing. The lack of user fees for electric
vehicles is a growing inequity that creates opportunities to move toward road
pricing, as discussed below
Pricing
(Chapter 7) transportation proportionate to use has been a holy grail for
transportation economists for decades. Pricing can be used to reduce or
eliminate congestion by managing demand so that it does not exceed available
supply. However, to date, it has been technologically and politically difficult
to implement such a system. The advent of electronic toll collection (ETC) in
the 1990s has resulted in a small resurgence in the number of toll roads, but
there is no evidence that individual toll roads will expand to be a significant
share of all roads anytime soon.
Cities like Singapore, London, and Stockholm have
established congestion charging zones. However, urban congestion charges have
yet to be deployed in any large US city, and are unlikely to come to Minnesota
before playing on the more congested New York, Los Angeles, San Francisco, and
Chicago stages.
High occupancy toll lanes, such as
the MnPASS lanes in Minnesota, are being deployed at a more rapid rate. The
additional merit of these lanes is the opportunity to have this converted to
serve automated-only traffic much sooner than all roads can be, providing a
much higher throughput than general purpose lanes. This could occur as soon as
2025, and provide a decade of additional road capacity before human-driven cars
are driven-off the freeway for the last time.
Notably, EVs do not pay gas tax. (And hybrid electric
vehicles pay much less per mile in gas tax than traditional internal combustion
engine vehicles). As EVs gain market share, if the user-pays principle is to be
maintained and reinforced, a new financing system needs to be found for these
vehicles. This provides an opportunity to implement mileage charges with
off-peak discounts, helping spread the peak and better-use road capacity.
Phasing in road pricing one electric vehicle at a time seems the most promising
strategy to deploy pricing on roads without the risks of a new large-scale
system deployment.
Logistics
(Chapter 8) identifies a number of potential changes affecting the freight
sector and how goods are delivered. Automation will affect deliveries as it has
changed passenger transportation. A variety of automated delivery systems are
likely to trialed in the coming decade, as distributors and retailers aim to
connect directly to customers.
On the logistics side, there are a number of changes
enabled by information technologies. Supply chain network pooling and the
physical Internet for long-distance shipments may become increasingly common as
a means of getting better capacity utilization out of vehicles and drivers or
vehicle controllers. Similarly efficiencies can be garnered through
consolidated home delivery. All of these mean that fewer, but heavier trucks
will be using Minnesota roads. Same day delivery in business-to-business, and
more significantly, in business-to-consumer sectors is also likely to become
more common, reducing shopping trips, and making online purchasing even more
spontaneous, but in the net not affecting road usage much in terms of amount,
but perhaps more in terms of additional traffic in evening and weekend periods.
The overall conclusions are complex, but they suggest
significant changes in the transportation sector over the coming few decades.
Business-as-usual practices will need to change consistent with changing
technologies and their effect on both supply and demand.
Chapter 1: Autonomous Vehicles
In March 2004, DARPA1 hosted the first Grand
Challenge on vehicle automation. Set in the Mojave Desert (crossing the Nevada
- California border), with $1 million going to the winner, the objective was
for driverless cars to complete a 150 mile (240 km) route. Carnegie Mellon
University's robot vehicle finished first, by completing almost 5 percent of
the route, but was not awarded the prize. A second Grand Challenge was held in
October 2005. Five vehicles completed the course, and Stanford University's
team won with a time of just under 7 hours.2 In little over 18
months, vehicle automation technology rapidly improved.
Two years later, in November 2007, DARPA established the Urban
Challenge on a closed course at George Air Force Base. The 60 mile (96 km)
route resembled an urban obstacle course. Carnegie Mellon took first,
completing the run in just over 4 hours. Stanford secured second at just under
four and a half hours. Unlike the Grand Challenges, cars had to have more
sophisticated and intelligent sensors. Though road quality was better (paved
rather than off-road), the challenge was far more challenging.
A more important outcome (perhaps) is that Google hired many
of the leaders of the Stanford and Carnegie Mellon teams,3 including
Sebastian Thrun of Stanford and Chris Urmson of CMU, for their own internal
secret project, which they announced in 2010. Google Cars had at that time
driven 1,000 miles (1,600 km) without human intervention and 140,000 miles with
limited control around the San Francisco Bay Area, see Figure 1.2.4 Google cars have now had more than a dozen
crashes, and at least one of which resulted in injury. Google's official position is that none of
the safety events were caused by Google’s vehicles or vehicle technology. There
is some concern that automated vehicles have different driving styles than
following human drivers may be used to, causing potential conflicts.
To date, Google's cars are very map-dependent, running where
the roads have been mapped out in detail, so that they can compare what they
see with what they expect to see.5 That strategy has strengths and
weaknesses. The strength is a reduction in computation costs and better
understanding
1 DARPA stands for
Defense
Advanced
Research
Projects
Agency; it is a unit of the
Department of Defense, as driverless cars have obvious military application.
2 Carnegie
Mellon teams took second and third place. The Gray Insurance Company from New
Orleans and Oshkosh Trucks also completed the course.
3
Markoff, John
(2010) Google Cars Drive Themselves,
in Traffic. New York Times
http://www.nytimes.com/2010/10/10/science/10google.html?_r=2&src=sch&pagewanted=all
Erico Guizzo (2011-10-18)
How Google's
Self-Driving Car Works
IEEE Spectrum
http://spectrum.ieee.org/automaton/robotics/artificial-intelligence/how-google-self-driving-car-works 4 Source: Data
on Google Cars from
140,000 -
http://googleblog.blogspot.com/2010/10/what-were-driving-at.html
300,000
http://googleblog.blogspot.com/2012/08/the-self-driving-car-logs-more-miles-on.html
500,000
http://www.businessinsider.com/google-self-driving-car-problems-2013-3?op=1
Nearly a million http://googleblog.blogspot.com/2015/05/self-driving-vehicle-prototypes-on-road.html
5 Alexis
Madrigal @ The Atlantic: How Google Builds Its Maps—and What It Means for the Future of Everything –
Technology :
http://www.theatlantic.com/technology/print/2012/09/how-google-builds-its-maps-and-what-it-meansfor-the-future-of-everything/261913/
and anticipation of the environment. Weaknesses include that
(1) not everywhere is necessarily mapped, (there may be “Google deserts”, for
instances some places are private property, and (2) the world changes, the map
cannot be updated instantaneously. The first Google style AV to pass the
unmapped or incorrectly mapped area will update the map as it passes, but it
will of course need the capability of traveling with unmapped, incompletely
mapped, or incorrectly mapped instances. And if it can do that autonomously,
does it really need the map to proceed? It cannot do that autonomously, there
remain issues with autonomous to human control interfaces.
Box 1: NHTSA (2013) Policy on Automated Vehicle Development
No-Automation (Level
0): The driver is in complete and sole control of the primary vehicle
controls – brake, steering, throttle, and motive power – at all times.
Function-specific
Automation (Level 1): Automation at this level involves one or more
specific control functions. Examples include electronic stability control or
pre-charged brakes, where the vehicle automatically assists with braking to
enable the driver to regain control of the vehicle or stop faster than possible
by acting alone.
Combined Function
Automation (Level 2): This level involves automation of at least two
primary control functions designed to work in unison to relieve the driver of
control of those functions. An example of combined functions enabling a Level 2
system is adaptive cruise control in combination with lane centering.
Limited Self-Driving
Automation (Level 3): Vehicles at this level of automation enable the
driver to cede full control of all safety-critical functions under certain
traffic or environmental conditions and in those conditions to rely heavily on
the vehicle to monitor for changes in those conditions requiring transition
back to driver control. The driver is expected to be available for occasional
control, but with sufficiently comfortable transition time. Google’s converted test
vehicles are an example of limited self-driving automation.
Full Self-Driving
Automation (Level 4): The vehicle is designed to perform all safetycritical
driving functions and monitor roadway conditions for an entire trip. Such a
design anticipates that the driver will provide destination or navigation
input, but is not expected to be available for control at any time during the
trip. This includes both occupied and unoccupied vehicles. Google’s new
“bug-like” car design without a steering wheel or brakes is an example.
In Fall of 2015, the electric vehicle automaker Tesla remotely
upgraded its most recent model year cars (about 50,000 vehicles) with
“auto-pilot”, making them semi-autonomous (late Level 2, early Level 3).[1]
Elon Musk, the CEO of Tesla, says he expects fully autonomous vehicles within 3
years (i.e. by 2018).
A test ride by this report’s first author indicates that
upgraded Teslas are able to function in handsoff mode some of the time. They
use adaptive cruise control to follow the vehicle in front at a desired speed
constrained by a fixed following distance and use lane markings to stay in the
travel lane. They change lanes automatically at the request of the driver (who
must hit the turn signal).
As of fall 2015, none of these functions can be safely
performed in a Tesla running “Auto-pilot” without driver observation and
monitoring. In fact, the vehicle requires the driver to periodically return
hands to the steering wheel. The vehicles do not yet automatically stop at
traffic lights or stop signs, though it is assumed that engineers are working
on and testing those functionalities. Ambiguities in lane markings (for
instance at freeway merges and diverges, or as a result of road construction or
restriping) still create difficulties for the vehicle in Auto-Pilot mode. First
person observations are that vehicles still over-react on curves (following the
average of the inside and outside curve, rather than a fixed distance from the
inside curve). Additionally, the give-way game between merging vehicles and an
on-road Tesla cannot yet be safely conducted in the absence of driver
intervention. At this stage, there is no obviously linkage between satellite
navigation and mapping and the control function. Teslas appear to be
map-independent, and controls are through on-vehicle sensors.
These are not the first serious attempts at autonomous cars.[2]
At Demo '97 Automated Highway Systems were successfully demonstrated. [3]
Cars could travel at high speeds, without driver intervention when closely (1
meter) following, on an isolated test track. It has long been envisioned,
dating back to the 1930s and the GM Futurama exhibition designed by Norman Bel
Geddes at the
While a technological success, the Automated Highway Systems
programs were a political failure, and the program was cancelled. The reason
for this is clear in retrospect: there was no deployment path. No one would
build limited-access roads for a very few specialized cars. No one would buy
cars that could be fully used only on selected lanes. Autonomous vehicles
running in mixed traffic solve this chicken and egg problem, since they will be
useful without special infrastructure, at the cost of much higher complexity.
However, after a critical mass of autonomous vehicles hits the
road, and once all the bugs are worked out, there will be potential gains
(closer following, narrower lanes) for them to travel in autonomousonly lanes
rather than mixed traffic. Existing managed lanes can be dedicated to AVs. Even
general purpose lanes can be designated and redesigned to AV-only traffic in
order to increase total system throughput. We may get special AV lanes on
highways as an interim step before all lanes on all highways are for AVs only,
and before non-AV cars are prohibited.
New players like Google, Tesla, Uber, Apple and others are
making serious investments in taking the driver out-of-the-loop for vehicle
control. Further, after six decades of technological dormancy, the traditional
automakers are responding to the DARPA Urban Challenge. For instance, Delphi,
an auto parts manufacturer spun-off from General Motors, drove an automated
Audi 3,500 miles (5,600 km) cross-country in March of 2015, with hands-off
control 99 percent of the time.[4]
In fact, Delphi’s forerunner (GM Subsidiary) Delco sponsored a similar trip in
1995 by Carnegie Mellon scientists, where the computer navigated 98.7% of the
time.[5]
Advance still requires clarity about what 'automation' really
means. As shown in Box 1, the National Highway Traffic Safety Administration
(NHTSA) has a series of levels describing degree of autonomy (from Level 0 -
"no autonomy" to Level 4 "full self-driving
automation").
More early versions of autonomous cars are anticipated to be
sold on the market for the 2017 Model Year (for instance Cadillacs with
“SuperCruise”, which like Tesla vehicles with “Auto-pilot” may be described as
somewhere between Level 2 "combined function automation" and Level 3
"limited self-driving automation"). Many of the necessary features
like lane-keeping, adaptive cruise control, and automated braking technologies
are already standard on high-end cars, as is automated parking assistance. None
of the automakers advises hands-off driving at this point.
Google itself [6]
is aiming for Level 4, and believes that incremental changes like Level 3 are
dangerous as once drivers cease paying attention it will be hard for them to
reassert attention in a timely manner. In other words Google favors a large
phase shift to AV for new vehicles. In
contrast, traditional automakers are instead rolling out particular packages
and features that still require the driver to attend to the driving task. The
market and regulators will determine which of these two technological
deployment paths actually occurs. The in-vehicle transition from autonomous
mode to make the driver assume control is likely to be dangerous (though on the
net, less dangerous than letting humans drive in the first place, otherwise the
technology would not be permitted).
1.1 Connected vs. Autonomous Vehicles
Some discussions conflate autonomous vehicle technology with
"connected vehicle" technology. These are distinct, however, as the
later allows individual vehicles to communicate with other nearby vehicles
(vehicle to vehicle or V2V) and connected infrastructure (V2I) with Mobile Ad
Hoc Networks. If widely deployed, this not only improves safety for those in
the vehicle, it improves the safety and environment for pedestrians,
bicyclists, and other drivers. Connected vehicles should enable vehicles to
anticipate better and negotiate with each other for use of a particular bit of
road space at a discrete point in time. Both autonomous and connected vehicles
are coming. It is important to recognize that cars may be autonomous but not
connected or connected but not autonomous, or both (or as today, neither).
Connected vehicles and infrastructure in particular may be more vulnerable to
hacking, though autonomous vehicles will likely have live connections to their
manufacturer as well.
The effects of autonomous vehicles are much more profound than
connected vehicles, as connected vehicles are only especially useful in the
presence of other connected vehicles, while autonomous vehicles are valuable
through the transition period when most vehicles are not up-to-date.
1.2 Timeline
Cumulatively, the distances driven driverlessly are rising
every year. This will take some time to perfect, but one day, in the near
future, you may wake up, give a voice command to a car, and never again touch a
steering wheel, gears, accelerator, or brakes (which won't be there)— as will
everyone else. You will step into your car, tell it where to go, and not think
about traffic. The window in front of you will be a heads up display giving you
information and entertainment, while allowing you to see the road coming
up.
To give a rough timeline, we anticipate Level 3 ("limited
self-driving automation") autonomous vehicles will be on the market by
2020. We predict Level 4 will be available in 2025 and required in new US cars
by 2030, and required for all cars by 2040. In other words, human drivers will
eventually be prohibited on public roads. Consumer acceptance remains an
unknown, and depends on the quality of the product being offered.
Automated vehicles are probably already legal in most US
states (New York requires hands on the wheel),12 so the burden of
proof is on those who want to slow them down. Several states, Nevada, Florida,
California, Michigan, Virginia, and the District of Columbia, have passed
special enabling legislation enabling testing of fully autonomous vehicles on
public roads.
1.3 Environments
The Cadillac SuperCruise entry into the
"semi-autonomous" vehicle market implies the first market for
autonomous vehicles would be the relatively controlled environment of the
freeway.13
However, entry into the relatively controlled environment of
low-speed places makes sense as well. These are two different types of vehicles
(high speed freeway vs. low speed neighborhood), and though they may converge,
there is no guarantee they will, and perhaps today's converged multipurpose
vehicle will instead diverge. There has long been discussion of Neighborhood
Electric Vehicles, ranging from golf carts to something larger, which are in
use in some communities, particularly southwestern US retirement complexes. In
Sun City, Arizona, for instance, people use the golf cart not just for golfing,
but for going to the clubhouse or local stores (usually as the household's
second or third car, but occasionally as the primary vehicle). They can do this
because local streets are controlled by low speed limits, and there are special
paths where golf carts are permitted and other vehicles aren’t.
Campuses, retirement communities, neighborhoods in some master
planned communities, and true parkways are almost ideal for these types of
“driverless carts”14 because these places don't have heavy traffic
and discourage high speeds.
12 See Smith, B.
W. (2012). Managing Autonomous Transportation Demand. Santa Clara Law Review,
52(4).
Weiner and Smith summarize the
outstanding questions that legal regimes will need to better address to cope
with driverless cars.
Gabriel Weiner
and Bryant
Walker Smith,
Automated Driving: Legislative and Regulatory
Action, cyberlaw.stanford.edu/wiki/index.php/Automated_Driving:_Legislative_and_Regulatory_Action
13 Nunez,
Alex
(2014-09-07) Cadillac will intro Super Cruise in 2017 flagship, V2V in
CTS.
Road and Track
http://www.roadandtrack.com/new-cars/future-cars/news/a8626/cadillac-announces-super-cruise-and-v2v-techintegration-for-2017-model-year/
14 We heard
the term attributed to Bryant Walker Smith
1.4 Consequences
Autonomous vehicles portend a series of consequences
affecting both the transport sector and society, such as status. We highlight
some below.
Safety. Autonomous
vehicles, powered by sensors, software, cartography, and computers can build a
real-time model of the dynamic world around them and react appropriately.
Unlike human drivers, they seldom get distracted or tired, have almost
instantaneous perception-reaction times, and know exactly how hard to brake or
when to swerve.
Cars would be much safer if only humans were not behind the
wheel. It is possible to plausibly imagine a reduction from tens of thousands
to hundreds of deaths per year in the US, upon full deployment.
Vehicle Form.
Autonomous vehicles promise a Cambrian explosion of new vehicle forms. Evidence
for this is already emerging. Google has
proposed and built prototypes of a new, light, low speed neighborhood vehicle
designed for slow speed (25 mph or 40 km/h) on campuses. The UK has four pilot programs starting.
Singapore is testing similar vehicles.
This has important implications. For example, cars can be
better designed for specific purposes, since, if they are rented on-demand or
shared, they don't need to be everything to their owner. Narrow and specialized
cars are more feasible in a world of autonomous vehicles. The fleet will have
greater variety, with the right size vehicle assigned to a particular job.
Today there is a car-size arms race, people buy larger cars, which are
perceived to be safer for the occupant even if more hazardous for those around
them, and taller cars, which allow the driver to see in front of the car
immediately in front of them. Both of these advantages are largely obviated
with autonomous vehicles. The car-size arms race ends.
The low mass of neighborhood and single-passenger vehicles
will save energy and reduce pavement wear, but also cause less damage when it
(inadvertently) hits something or someone. Combining the low mass with the
lower likelihood of a crash at low speed will magnify its safety advantage for
nonoccupants in this environment, compared with faster, heavier vehicles, which
privilege the safety of the vehicle occupants.
These savings will be passed on to consumers. Insurance
companies will recognize the lower risks and lower rates. This will help drive
adoption of autonomous vehicles. Alternatively, the autocompanies themselves
may choose to accept liability for autonomous vehicles in autonomous mode, as
some are already proposing.
Capacity. Because
they are safer, autonomous vehicles can follow other each other at a
significantly reduced distance.
Because they are safer and more precise and more predictable, autonomous
vehicles can stay within much narrower lanes with greater accuracy.[7]
Lateral distances can be closer. Lanes can be narrower. If skinny cars emerge
(designed for one-passenger, or several passengers in tandem) lanes can be
narrower still, or be shared with two such cars.
Thus, capacity at bottlenecks should increase, both in
throughput per lane and the number of lanes per unit road width. These cars
still need to go somewhere, so auto-mobility still requires some capacity on
city streets as well as freeways ubiquitous adoption of autonomous vehicles
would save space on lane width.
Parking: Autonomous
vehicles would save space on parking too. Cars can drop off passengers in front
of destinations and go elsewhere to park as needed. Subsequently they can pick them up at
origins. This requires reconfiguration of drop-off and pick-up areas to avoid
large queues. Parking stalls can be narrower and parking decks shorter if
people are not required to use them. Cars can be packed more tightly in such
parking facilities. Further those facilities will be farther from the high
value real estate locations. Parking is further discussed in Chapter 4.
Cars without people.
Autonomous cars can drive without people at all. They can be used for pickup
and delivery, in addition to the dead-heading from drop-off to parking, or from
drop-off of one passenger to pick-up of another, or for recharging or
refueling. All of this can increase total travel on the road.
Mobility for All.
Automated cars will enhance mobility for children and people with disabilities.
Parents, friends, and siblings need not shuttle children around; the vehicle
can do that by itself (assuming increasingly protective parents would allow
such). The child is securely identified with camera and biometrics, and parents
can even monitor their child with an in-vehicle video camera— yielding an
environment far more secure than the school buses and carpools children
currently ride. There likely will remain debate about how old a child must be
before she is placed alone in an autonomous car, but the consensus is likely to
be, if she is in kindergarten, she can ride alone, as with school buses. (This
is a similar argument with ridesharing services today that offer rides, but
that is to date a small phenomenon).
Human travel will be much more point-to-point, with far fewer
pick-up and drop-off passenger trips required as that can be off-loaded. Deadheading
autonomous vehicles, driving around without a passenger to pick up their next
family member may become common, though logistics and shared vehicles can
minimize the amount of this.
Costs. The capital
costs for autonomous vehicles are likely to be higher than traditional cars, at
least at first, until driver-facing technologies (like the steering wheel,
brake and accelerator pedals, and so on) can be removed for cost savings, as
the sensors and computers add some cost compared to existing systems. Those
additional costs decline over time, as learning curves, paying off R&D, and
mass production all lower expenses.
In contrast, fuel costs should be lower, as autonomous vehicles are likely
to be more efficient, both due to less congestion and more optimized driving
styles ranging from smoother acceleration to various hypermiling techniques
like drafting to reduce drag.
Labor is a significant share of costs in transport, for
vehicles such as taxis, buses, and trucks, which today require a driver. With
automation that labor cost vanishes. We imagine a transitional phase where
remote control drivers in a traffic center simultaneously monitor and manage
multiple vehicles for situations when autonomous vehicles are not fully trusted.
We expect those operators to be bored. This lower cost benefits taxis, buses,
and trucks which had held higher labor costs relative to their competitors.
There are additional labor costs associated with driving a private vehicle
which don’t show up in the economic statistics, but can be quite expensive,
particularly for high wage workers. This cost, too, will be reduced.
Delivery services with online purchasing will become even more
cost-competitive compared to traditional retail. Transit will either be more
cost effective than it is now, or be able to offer lower fares, or some
combination of the two.
Right-of-Way Retrofit. To
accommodate specialized low speed neighborhood or campus vehicles, most
non-ideal places will require retrofits so that places can be connected with
routes of low-speed. Retrofitting cities for transport has a long history as
cities and transport technologies co-evolve. Cities, which had originally
emerged with human and animal powered transport, were retrofitted first for streetcars,
and then for the automobile, and in some larger cities for subways. We have
also redesigned our taller buildings for escalators and elevators.
Some places where retrofits might be required and feasible
include cities laid out and built before the automobile. Much of the street
grid can be retrofitted ("calmed") to disallow high-speed traffic, in
much the same way bicycle boulevards are established. Similarly, retrofits are
technically feasible anywhere there is space to install a slow network in
parallel with the existing fast network, for instance, with barrier separated
lanes on wider suburban roads.
Vehicle diversity applies not only to a larger variety of
motorized vehicles of various sizes, but also to a greater variety of transport
using the existing streets, which today are highly segregated with cars (both
moving and parked) dominating the street and pedestrians the sidewalk. Slow
speed, light weight vehicles make shared spaces, which don't differentiate
between the road and the sidewalk much more palatable.
Roadspace Reallocation.
It follows that if transport systems require reduced lane width and have
adequate capacity, transport agencies can reduce paved area and still see
higher throughput. Today, most roadspace is not used most of the time, but road
agencies cannot just roll it up when it is not being used.
However, on freeways the space can be deployed more
dynamically to increase either safety (by increasing spacing) or capacity (by
reducing spacing), simultaneously adjusting speed and spacing accordingly.
Dynamically reversible lanes are possible once humans are out of the loop.
On local streets, roadspace no longer required for motor
vehicle movement can be reallocated to other uses (pedestrians, bicyclists, transit,
parks and so on). But for purposes of reliability and safety, bikes, bus-rapid
transit, and the newly emerging micro-transit modes benefit from priority
lanes.
Nomadism. For a
select few, driverless vehicles may bring back the recreational vehicle, as
some choose the fully nomadic lifestyle, spending much if not most of their
lives in motion, especially if energy costs are low.
Driving style. A
fundamental shift will occur once people no longer drive themselves. The
preference that is felt for fast acceleration will diminish, permitting much
more efficient engines (gasoline or otherwise).
Ownership.
Ownership of autonomous vehicles is a looming question. While most roads are
public; most cars are private and individually owned. Most transit also is
public, though services are sometimes performed under contract by private
firms. Some private roads are emerging, but these roads are impossible without
public approval and assistance. New technologies provide an opportunity to
revisit old arrangements. New forms of ownership and payment will inevitably be
a dynamic evolution. Customers would need to pay for services of any type
(either as a subscription or a per-use basis). Advertising could offset—though
not entirely cover—some costs. It is conceivable that stores might subsidize
transport, as might employers, as benefits for the customers or staff (as they
do today with parking).
As discussed in the chapter
on SHARING, Transport Network Companies such as Lyft and Uber compete with
taxis. But with their added labor, such services are too expensive for most
people for frequent mobility.
Alternative ownership currently exists to some extent with the
current carsharing companies (Car2go, Zipcar, etc.) which compete with rental
cars. But again the cost is too high for most people to use on a daily basis
for a primary mode of transportation, and unless they live in a place with many
other users, the distance to the vehicle may be high. However, with autonomous
vehicles, the cost of the driver can be skipped, the car can come to the
traveler.
In contrast, autonomous vehicles total costs will be
significantly lower, making it feasible that larger numbers of people replace
their personal car (which is parked 23 out of 24 hours) with one that comes
on-demand.
Status: Just as
owning a car was once a class signifier in the US, and remains so elsewhere in
the world, and as owning a particular model of car (like a Prius or a BMW)
persists as a signifier, we can expect that during the transition period owning
an autonomous car will be a class-signifier. It indicates at once that you are
wealthy enough to own a new car, and technologically sophisticated enough to
trust your life to it. While eventually we expect this to be uniform, early
adopters will have very different economic and social characteristics from the
population at large. During the long transition, those who cannot afford such
cars may come to be vilified as the cause of crashes.
1.5 The Future of Travel Demand and Where We Live: The Out Scenario
Autonomous vehicles will be faster. It will not escape the
astute reader that, as the economists say “all else equal” faster vehicles
increase demand. Will autonomous vehicles reverse the trends or plateauing or
dropping per capita vehicle travel that we have seen for the past decade with
existing technology? Will people make more trips? Will people make longer
trips? Will people relocate?
Each advance in mobility (the ability to go faster, either due
to new technologies or more connected networks) has heretofore increased the
size of metropolitan areas. People can reach more things in less time. Subways
drove the expansion of London, while streetcars did the same for many American
cities.[8] Historically the time saved from mobility gains
was used mostly in additional distance between home and workplace, maintaining
a stable commuting (home to work) time. Recent evidence of average reduction in
overall travel distances and time spent traveling presented is not as much
about a reduction in trip distances between home and work (which if anything
are still rising in the US) but a reduction in the number of work and other
trips being made across the whole population. In short: speed decentralizes.[9]
Autonomous vehicles will likely be faster, particularly on
freeways, especially after widespread deployment when all vehicles are
autonomous. This will occur either once human cars are prohibited from
freeways, or once a network of separate lanes are designated for autonomous
cars.
Coupling with just the faster speed, the fully autonomous
vehicle lowers the cognitive burden on the former driver/now passenger. Modes
with lower cognitive burden tend to have longer trip durations. Time is
important, of course. What you can do with that time (the quality of the
experience) also matters (See Chapter 2).
If you can work while traveling, the value of saving time is less than if you
must focus on the driving task. This may also explain the premium people are
willing to pay for high-quality transit and intercity rail service.[10]
If the time or money cost of traveling per trip declines, the
long-held theory of induced demand
predicts, all else equal: more trips, longer trips, and peakier trips (more
trips in the peak period). Privately owned autonomous vehicles lower the cost
of travel per trip.
Out: More vehicle
travel with increased exurbanization. Fast, driverless cars that allow
their passenger to do other things than steer and brake and find parking impose
fewer requirements on the traveler than actively driving the same distance.
Decreases in the cost of traveling (i.e., availability of multitasking) makes
travel easier. Easier travel means increases in accessibility and subsequently
increases in the spread of development and a greater separation between home
and work, (pejoratively, sprawl), just as commuter trains today enable exurban
living or living in a different city.19 This reinforces the
disconnected, dendritic suburban street grid and makes transit service that
much more difficult (as if low density suburbs weren’t hard enough). People
will live farther “Out”.
It has been estimated that a 1 percent increase in
accessibility leads to a 0.6 percent increase in travel.20 Couple
this increase with the new mode of cars deadheading without people, and perhaps
the doubling of capacities and speeds leads to a doubling of total travel,
assuming nothing else changes. (Compare with the scenario presented in Chapter 4: Mobility-as-a-Service).
Thus as the cost of travel decreases, people will be more
willing to live in cities far from where they work. It is not uncommon for the
Dutch to live on opposite sides of the country from where they work, relying on
the train network. The Northeast Corridor of the US has people living in one
city and commuting to another (for instance from Washington to Baltimore,
Philadelphia to New York). At speeds of 100 miles per hour (160 km/h), the
commuting range expands widely.
It is entirely likely that such new e-propelled forms of
transport, along with solar power, will greenwash a new generation of dispersed
development, which may in fact net to a much smaller environmental footprint
than today, if not a smaller footprint than future cities.
The interplay of autonomous vehicles and pricing is especially
important. While autonomous vehicle capacity eventually doubles or quadruples,
per capita demand will rise as well if traditional patterns of induced demand
hold, and people continue to work, shop, and play at today’s rates. It is quite
possible that sharing (described in Chapter 4) remains a niche while most
people choose to own their own cars — the “Out” scenario dominates. Thus exurbanization and cars driving around
without people make extensive use of the newly available capacity. To fully
mitigate these congestion effects, pricing is required.
1.6 Discussion
Traffic congestion as a problem has the potential to diminish
significantly if the capacity gains outweigh the increased demand they
generate. People won't be driving themselves, but passengerless cars will be
moving all over the place.
For personal travel, the question remains: rent or own? Will
people with regular everyday trips find it cheaper to rent than own? The cost
of ownership may turn out to be cheaper than the cost of rental in the cloud commuting model in many if not
most US markets (outside central cities). Personally owned cars are not
dead-heading everywhere for passengers, lowering energy and operational costs
and wear and tear. Owners have more motivation to care for their own car (in
subtle, non-detectable ways) than for a rental, so the car may last longer.
Cloud commuting will permit much higher fleet turnover, and thus increase
advanced technology penetration.
19 For more on this reasoning,
see Chapter 11 in Levinson, D. and Krizek, K (2008) Planning for Place and Plexus:
Metropolitan Land Use and Transport. Routledge.
20 Weis’
estimates give a demand elasticity of 0.6 to accessibility (Hansen style/log
sum term)
Weis, C.
(2012) Activity oriented modelling of
short- and long-term dynamics of travel behaviour, PhD Dissertation, IVT,
ETH Zürich, Zürich. Or a shorter version
Weis, C.
and K.W. Axhausen (2012) Assessing changes in travel behaviour induced by
modified travel times: A stated adaptation survey and modelling approach, disP, 48 (3) 40-53.
Continuing automation of vehicle technologies will lead to
widespread deployment of driverless cars in mixed or eventually fully automated
roadways. Eventually humans driving on
public roads will be banned or greatly restricted. This has numerous
implications. Chapter 4 considers
the ownership model, and it is on ownership which turns the question of the
travel demand and land use effects of Autonomous vehicles. Chapter 8 looks at issues of urban and long-haul freight
transportation and automation, as well as local delivery. Other implications
described in this chapter include those on for travel demand; safety; capacity;
mobility for children, elderly, and disabled; transit; parking; and land use,
population distribution and development.
The authors anticipate that autonomous vehicles will go from
their current status of 0% market share to an end state of 100% of all new car
sales (i.e. autonomous capability will be a requirement of new car purchases).
Further, older human-driven vehicles will be phased out except for special
purposes (car shows, races, parades). A rough guide to the anticipated timeline
is that NHTSA Level 4 cars (fully self-driving without human interaction) will
enter the market between the 2020 and 2025 model years, and be required in all
new cars by 2030, and by 2040 human-driven cars will be generally prohibited.
Self-driving cars in specific contexts (e.g. freeways or isolated campuses) are
expected enter the market before 2020. Current near-self-driving cars such as
SuperCruise and the Tesla Model S are approximately NHTSA 2.5.
While the end of state of
autonomous vehicles presents a variety of benefits for travelers and transportation
agencies alike, the adoption rate is unknown. It is far from clear the pace of
change of autonomous vehicles.
1.7 Consequences
1.
An increase in safety. This will reduce the
number and severity of crashes. Follow-on effects include a reduction in
non-recurring congestion, and ultimately less resources spent on emergency
response.
2.
An increase in capacity as cars will be able to
follow at closer headways.
3.
An increase in capacity due to better
lane-keeping. This will enable narrower lanes.
4.
More opportunities for Mobility-as-a-Service
[MAAS], and a potential change in ownership structure.
5.
A reduction in the effort (both physical and
cognitive) associated with driving, which should increase people’s willingness
to travel for longer periods of time.
6.
An increase in the speed of travel, which should
increase people’s willingness to travel longer distances.
7.
An increase in mobility for those who now cannot
drive (children, the disabled). This will also increase the total amount of
travel
8.
A reduction in the space required for parking
near the travelers destinations as cars can go and park themselves. Further
with Mobility-as-a-Service, cars may remain in use longer, and thus spend less
time parked.
9.
An increased ability to route vehicles in a system
optimal way, as the computers doing routing can choose alternative algorithms.
10.
Lowered labor costs for transit, trucks, and
other service vehicles (e.g. snow plows).
11.
Easier recharging for electric vehicles.
In the long run, this mostly
suggests that less road capacity will be required per person. While longer
trips, more trips, and increased mobility for the transportation disadvantaged
(assuming vehicle ownership remains high, as opposed to a cloud commuting
model) may offset some of the capacity gains, the capacity gains are likely to
be large in the end. Other societal changes also generally point in that
direction.
Chapter 2: Mobile Telecommunications and Activity In-Motion
2.1 Introduction
The rise of mobile technologies
have led to their increased use and increases in multitasking behavior while
in-motion (enabled by technologies such as 4G, and in the future 5G and 6G
phones and in-vehicle Wi-Fi). This research considers interactions including
evidence from time use while traveling (reading, listening to music, work).
Autonomous vehicles (see Chapter 1)
will likely make this phenomenon more widespread. Better future
telecommunication technology will also enable higher-bandwidth activities.
Theory predicts that increasing use of telecommunications in motion makes
travel time more useful, and this increases the willingness to travel. Note
that telecommunications in general may enable less passenger travel with
activities such as teleworking and e-shopping.
This report focuses on telecommunications in motion and the impact of
telecommunications on activities-in-motion.
With more willingness to travel, people will presumably
spend more time in motion. What types of activities can be carried out during
travel in self-driving vehicles with better mobile technologies? How these
activities differ from traditional activities that are conducted during travel
on public transit vehicles? Will the
self-driving and mobile technologies change spatial patterns of origins and
destinations? Will trip chaining behavior become more popular? Will the
behavior changes enabled by the self-driving and mobile technologies differ by
gender, age, and socio-economic status (SES)? If so, how? And, what would be
the societal impacts of these behavioral changes? These are the questions
explored in this report.
This chapter summarizes existing
empirical evidence in the literature on activities in-motion. Then, the
2011-2014 American Time Use Survey data are used to explore new activities
during travel, possibly enabled by self-driving and better mobile technologies.
It then discusses how the technologies may influence car preferences, perceived
utility of travel, spatial configurations of travel, including patterns of
origins and destinations as well as trip chaining. How the effects may differ
by personal socio-demographics and the societal implications of the effects are
also discussed.
2.2 Existing Empirical Evidence
Existing studies on mobile telecommunications and activity
in-motion have focused on travel time use by rail and bus passengers. Recent
research has shown that roughly 55% of rail passengers and 40% of bus
passengers engage in technology, and that 30-40% of rail passengers work on
board compared to roughly 10% report no use of travel time. Table 2.1
summarizes major time use activities reported as undertaken during travel in
existing studies.
Most studies listed in Table 2.1 are about time use during
public transit trips. Only Malokin et al. (2015) and Mokhtarian et al. (2014)
used samples of commuters including both transit commuters and commuters by
other modes. Consequently, Malokin et al. (2015) included activities that are
often considered as not suitable during transit trips, e.g., grooming and
exercising.
Table 2.1: Summary of activities undertaken
during travel in existing studies
Activities reported as undertaken during travel Study Dataset
|
Reading
for leisure; Window gazing/people watching; Working/studying; Lyons et al. Talking to other passengers;
Sleeping/snoozing; Listening to music/radio (2007)
|
Great Britain National Rail Passengers Survey
|
Communicating; Entertainment/recreation; Formal;
Household/personal Kenyon and Activity dairy survey at 6
Information
search; Shopping; Travel; Other/personal. Lyons
(2007) oflocations England in the south west
|
Window-gazing/people
watching; Sleeping/snoozing; Thinking about/planning personal matters;
Working/studying; Sent SMS/called by mobile telephone; Use mobile telephone
in other ways; Talked with other passengers; Took care of the children;
Reading for leisure; Listening to music/radio; Playing games (electronic);
Knitting, needlework; Other activities
|
Gripsrud
&
Hjorthol
(2012)
|
Rail passenger survey in Norway
|
|
sleeping/snoozing; reading
for leisure; working
(reading/writing/typing/thinking); talking to other passengers; window
gazing/people watching; playing games (electronic or otherwise); listening to
music/radio; text messages/phone calls—work; text messages/phone calls—personal;
eating/drinking; entertaining children; being bored/anxious
|
Lyons & Urry (2005)
|
Conceptual paper on time use during train trips
|
|
Idling/thinking/window watching;
Reading hard copy; Writing hard copy; Talking face to face; Personal
care (eating, drinking, baby); Listening; Talking on phone; Reading digital;
Game playing; Writing digital (texting, e-mailing, etc.)
|
Guo et al. (2015)
|
Survey and observation of bus passengers in Vancouver,
Canada
|
ICT use; Entertainment; Relaxation Study/work; Talk to
others Ettema et al. Survey of transit riders in
(2012) Sweden
|
Smartphone; Internet; Reading electronically; Gaming electronically; Malokin et al. Messaging; Watching scenery/
people; Daydreaming; Exercising; Writing (2015) electronically; Laptop/ tablet;
Thinking/ planning; Reading from paper; Sleeping/ resting; Talking to
strangers; Writing on paper; Talking to friends; Eating/ drinking; Audio;
Grooming; Talking on phone; Navigating; Watching video;
|
Paper and online surveys of workers in Northern
California
|
Talked with other people; Made phone call or sent text;
Listened to music Mokhtarian et 2007–2008
French National
or radio;
Looked at the landscape al. (2014) Travel
Survey
Figure 2.1 uses the activities
listed in Table 2.1 to generate a word cloud describing activities that are
most frequently studied as activities during travel. It is evident that
talking, watching, reading, writing, sleeping, listening, thinking, phone are
the most frequent activities being studied as activities in-motion.
Besides understanding how travel time is used, existing
studies on the subjects have suggested the following:
Activities
conducted during travel differ significantly by journey purpose and direction
of travel. Commuters are more likely to be engaged in work-related activities
during workrelated trips than leisure-related trips, and during morning commute
trip than return-home trips.
Activities
differ significantly by gender, age, and class. Compared to men, women are more
likely to spend travel time talking to other passengers and less likely to work
or study. Older passengers are less likely to use smart devices during
trips. Higher income commuters are more
likely to use travel time to work and study.
Activities
differ by trip duration and items individuals have at hand. People are much
more likely to do window gazing and people watching in trips of less than 15
min duration than in longer trips.
Activities differ by environmental
factors during the trip. Noise level significantly reduces the use of smart
devices such as smartphones and iPads. Seating significantly increases the use
of smart devices. Jerkiness reduces the likelihood of multitasking on bus.
Surveys on how air passengers value onboard Wi-Fi also
provide insights into understanding the impact of telecommunications in
motion. The 2014 Wireless Connectivity
Survey includes more than 1,000 adult flyers in the United States. The survey
found that having access to onboard Wi-Fi not only affecting passengers’ flying
experience, but also affecting passengers’ fight selections. In-flight Wi-Fi was found to influence flight
selection for 66% of travelers, and 22% of the respondents said they have paid
more for their ticket in order to fly on a Wi-Fi equipped aircraft. These
survey results indicate that telecommunications in motion could have a
significant impact on transportation mode choice, which is consistent with
recent studies that suggest broadband access on public transportation systems increases
ridership and encourages people to shift from cars to public transit modes
(Frei and Mahmassani, 2011; Dong et al., 2015).
Yet, it is important not to
overstate the impacts of telecommunications on mode choice. Regardless of the
age of the respondent, the two most important factors for choosing a public
transit mode are total travel time (walking/cycling/driving + waiting + riding
+ walking/cycling/driving) and service reliability (Transit Center, 2014). As shown in Figure 2.2 below, access to Wi-Fi
is a less important mode choice factor.
Figure 2.2: Potential Drivers of Transit Ridership by Age (Figure downloaded from Transit
Center, 2014)
No comments :
Post a Comment
Note: only a member of this blog may post a comment.