A large driving force for the adoption of autonomous technology is the impact AVs will have on society. From convenience to emission reductions, AVs will undoubtedly revolutionize the impact that transportation has on our day-to-day lives.



One of the most important benefits of self-driving vehicles is that they will provide safe mobility to the elderly and the disabled, but also to people who are too young to drive. Fully automated vehicles (AVs) also facilitate personal comfort and independence while traveling safely.

The U.S. population is aging, with the demographic of people over the age of 65 growing at a faster rate than other demographics (Fagnant and Kockelman, 2015, pg. 5). Elderly people are more likely to develop disabilities that make it less safe or near impossible to drive.

One of the most important benefits of self-driving vehicles is that they will provide safe mobility to the elderly and the disabled, but also to people who are too young to drive. AVs also facilitate personal comfort and independence while traveling safely. All of these factors will further increase automobile travel demands, and eventually VMT and congestion. However, by providing safe mobility for everyone, there is an increase in productivity and efficiency. Generally, many of the expected benefits depend on the shift to mobility services more so than on automation. However, automation can make mobility services more competitive by reducing its price significantly.


Screen Shot 2018-05-22 at 4.32.17 PM.png
Source: Litman, 2018, pg. 6

Mobility as a Service (MaaS)

The future is likely to see trends that move toward the usage of vehicles as an on-demand service as opposed to an owned product. SAVs can act as an on-demand service for passengers from origin to destination. This can be implemented either by private agencies or as part of public transit.

Potential Impacts

  1. AVs are designed to improve mobility, both by making it safer and more comfortable. Since this will potentially offer greater benefits to high income areas, we can expect an automation-induced sprawl in developing areas. This will most likely be similar to the traditional automobile driven sprawls.
  2. Parking needs may be severely affected. There likely will not be as much demand for expensive paid on-street parking as it becomes easier for an AV to self-park in cheaper areas, to send the vehicle back home (creating empty VMT), or to use SAVs that are generally in a loop providing services.
  3. As one study notes, “car and ride-sharing programs could expand, as AVs serve multiple persons on demand; and the trucking industry may realize better fuel savings via road-trains, or even one day go driver-less.” (Fagnant and Kockelman, 2015, pg. 5)
  4. Introduction of AVs may also result in a shift of destination choice. People may tend to travel to places of further distances due to the decrease of the value of travel time and increase mobility. Perrine et al. predict a 9.6% increase in long-distance travel for personal vehicles (Perrine et al., 2018).

The wide benefits of AVs on mobility are also highlighted in the Shared (fully) automated Vehicles page.


Fully Automated vehicles (AVs) are likely to change emissions, but the manner in which it change is still uncertain. These changes may also be influenced by other changes in vehicle operations, vehicle design, or transportation system design.


In 2015, 27% of greenhouse gas emissions in the US were due to transportation (EPA, 2017). Fully automated vehicles (AVs) are likely to change this contribution to emissions, but how it will change is still uncertain. These will follow other changes in vehicle operations, vehicle design, or transportation system design, which may be facilitated by automation (Wadud et al., 2016). Many aspects of AVs will bring about positive energy impacts, while others will likely cause negative energy impacts. These factors are summarized in the graphic below (Gonder et al., 2016, pg. 3). While increases in travel demand would largely increase emissions, AVs have the potential to reduce this effect overall.ea1.png

Vehicle-Miles Traveled

AVs will likely increase vehicle miles traveled (VMT) as mobility increases and encourages travel. VMT will also increase with the rise of shared (fully) automated vehicles (SAVs) due to extra empty driving between stops and to and from charging stations, in the case of electric vehicles. The graphic below summarizes the estimates for this VMT increase based on different models and parameters.

Study Method Estimate Sources of increase or decrease in VMT
Brown et al. (2014) Additional miles if all people over 16 had VMT of highest demographic Upper bound annual VMT: +40% New demand from underserved populations (youth, disabled, and elderly)
Childress et al. (2015) Activity-Based Model Daily VMT: -35% to 20% Changes in value of travel time, road capacity, parking costs and per mile driving costs.
Fagnant and Kockelman (2014) Agent-Based Model Daily VMT: +11% Relocation of unoccupied automated taxis
Fagnant and Kockelman (2015) Assumptions based on published literature VMT per AV: +10% to +20% Induced Demand
Schoettle and Sivak (2015) Developed trip overlap and household requirements in an AV environment Upper Bound VMT per AV: +75% Reductions in household vehicle ownership
Wadud et al. (2016) Assumptions based on natural declines in travel due to age Upper Bound Annual VMT: +10% New demand from new user groups
Harper et al. (2016) Demand Wedges Upper Bound Annual VMT: +14% New demand from underserved populations
Liu et al. (2017) Activity-Based Agent-Based Model Daily VMT: +9.8% to 15.7% Empty SAV repositioning
Perrine et al. (2018) Trip-Based Model Daily VMT: +9.6% Destination choice due to the change of value of travel time

Source: Harper et al., 2016, pg. 26

In order to minimize this VMT increase, a certain policy can be put in place. The Autonomous Vehicle Policy Framework Summit Report suggests that the best practice for tackling the emissions issue is encouraging zero emission AVs with clean energy sources, to reduce impacts on air quality (Snyder et al., 2017, pg. 8). See more policy recommendations on the Policy page.

Hybrid and Electric AVs

If the benefits of electrification are realized, they would far outweigh the worst case impacts from increased VMT. However, it may be difficult and costly to fully implement electric automated vehicles (EVs) to the extent that we would see this benefit. Fully electric vehicles would require constant and convenient access to charging stations, which may contribute to congestion and increases in VMT. Alternatively, hybrid automated vehicles (HAVs),  prove to provide a significant benefit, and are less costly overall, “the main advantage of EV HAVs will be the environmental benefits they bring, in particular, emissions at the point of use (IMInnovation, 2017).” According to Loeb and Kockelman, EVs, unlike Hybrid-EVs (HEV) or AVs, “have the potential to provide zero-carbon transportation with a renewable power grid,” (Loeb and Kockelman, 2017, pg. 1). However, at this time, it appears that HEVs would be more easily integrated because they allow for more leeway in the availability of charging stations. The choice between the types of HAVs will “depend on the progress in battery technology over the next few years, and specifically in power density, charging time and effective cycles. These improvements will help to remove concerns about the range they can travel to, without having to pause to re-charge” (IMInnovation, 2017). According to Loeb et al., “as the price of EV batteries continues to fall, charging facilities become more convenient, and renewable energy sources grow in market share, EVs will become more economically and environmentally competitive with conventionally fueled vehicles” (Loeb et al., 2018, pg. 1).

Shared AVs

While shared (fully) automated vehicles (SAVs) are likely to increase VMT, they may still bring energy and fuel savings if implemented carefully. According to Fagnant and Kockelman, “SAVs may be purpose-built as a fleet of passenger cars, replacing many current, heavier vehicles with higher emissions rates,” (Fagnant et al., 2015, pg. 15). Simulation-based estimations suggest that, “SAV fleets could deliver an energy savings of 12%, along with a 5.6% reduction in greenhouse gas (GHG) emissions, relative to privately owned and operated human-driven vehicles,” (Kockelman et al., 2016, pg. 211). Electrification is relevant to the future of SAVs as well. EVs are an attractive option for SAVs because of energy efficiency and reduced emissions, but a simulation-based study in Austin, TX found that “for all metrics studied, a gasoline hybrid-electric (HEV) fleet performed better than EV fleets, while remaining more profitable, providing response times of 4.5 minutes” (Loeb and Kockelman, 2017, pg. 1).

Engine Load Smoothing

Engine Load Smoothing is the term that refers to the removal of fuel-consuming human driving tendencies like speed fluctuations, long reaction times, hard braking, and rapid acceleration. The graphs below from Liu and Kockelman illustrate the potential impacts of removing the human element from the vehicle as fat as the driving cycles are concerned (Liu and Kockelman, 2018, pg. 7).

cycle smoothing.png

As noted in a 2016 study sponsored by the Texas Department of Transportation on connected and autonomous transportation systems, “hard braking and rapid acceleration events are associated with increased emissions, so, by smoothing HVs’ existing driving cycles, this work anticipates the emission benefits of CAVs.” More efficient driving habits, as well as more efficient engines, will lead to decreased emissions per VMT due to more gradual acceleration and deceleration in driving cycles (Kockelman et al., 2016, pg. 229). However, safety and human capabilities generally limit highway speeds , while AVs may be capable of driving at higher speeds more safely in the future, which could increase energy consumption per mile (Wadud et al, 2016).

Other Environmental Benefits of AVs

In addition to the incorporation of hybrid and electric vehicles, SAVs, and cycle smoothing, there are other minor environmental benefits associated with AVs.


AVs are anticipated to be lighter in many cases, as passengers are less likely to need to keep their belongings in their vehicles and as SAVs become more widespread. AVs may also become lighter in the long term due to reduced safety equipment, if crash rates are greatly reduced.

Truck platooning will also likely decrease emissions through the utilization of vehicle spacing, cruising speed, vehicle loading, and engine loading (Gonder, 2016, pg. 7).

While more frequent travel through mobility service of SAVs, it will also decrease the amount of cold starts. Cold-start emissions are much higher than after a vehicle’s catalytic converter has warmed up (Loeb et al., 2018, pg. 5).

There is potential to move freight trips or others to nighttime travel or other off-peak times to reduce congestion and emissions (due to cooler nighttime weather) (Snyder et al., 2017, pg. 20).

Connectivity of vehicles between their environment and each other through cooperative intersection coordination and coordinated adaptive cruise control will also help to decrease emissions due to more efficient travel and fewer start-stop requirements of the vehicle.



AVs have innate potential to encourage additional travel as compared to human-driven vehicles, both by those who currently drive privately owned vehicles and those who do not. This is due to the many added benefits that occur as a result of removing or limiting the driving task.


Both short and long distance travel will become more convenient for everyone following AV implementation, as Kockelman et al. notes, “vehicle-miles traveled (VMT) is predicted to rise, as travelers experience falling values of travel time (easier travel),” (Kockelman et al., 2017, pg. 138). This is likely to increase congestion on our roadways unless some changes are made, including increasing vehicle occupancy.

Shared (Fully) Automated Vehicles (SAV) and Dynamic Ride-sharing (DRS)

Ride-Sharing Today


Ride-sourcing services, such as Uber and Lyft, have grown their presence in the world of transportation in the recent years, and through uberPOOL and LyftLine these services have popularized the concept of ride-sharing as we know it. Ride-sharing offers a number of advantages as, “these shared services typically charge less than regular ride-sourcing offerings and allow for dynamic changing of routes as passengers request pickups in real time,” (Stocker and Shaheen, 2017, pg. 11). According to a study conducted by Paolo Santi, adding a shareability network to millions of New York City taxi trips, “cumulative trip length can be cut by 40% or more. This benefit comes with reductions in service cost, emissions, and with split fares, hinting toward a wide passenger acceptance of such a shared service,” (Santi et al., 2014). As AVs are implemented, ride-sharing is expected to grow even more. SAVs “are a fleet of AVs that provide low-cost service to travelers, possibly replacing the need for personal vehicles,” (Levin et al., 2017, pg. 2). SAVs are made more efficient through a system called dynamic ride-sharing (DRS).

What is DRS?

Dynamic ride-sharing is a form of ride-sharing in which people hail a vehicle using their smartphones and they are grouped with a carpool of strangers whose destinations are located on or near a common path. A study using cellphone GPS data in Orlando, Florida found that ,“nearly 60% of the single-person trips could be shared with other individuals traveling solo and with less than 5 minutes added travel time (to arrive at their destinations), and this value climbs to 80% for 15 to 30 minutes of added wait or travel time,” (Gurumurthy and Kockelman, 2018, pg. 1). Brownell and Kornhauser (2014) predict that automated taxis will use road space more efficiently and with a lower incident rate than automobiles, suggesting reduced trip times. DRS should be convenient and reliable enough to be used on a daily basis.

The illustration below shows the differences between a system with and without DRS (Gurumurthy and Kockelman, 2018, pg. 7).

A simulation-based study found that SAV systems that used DRS outperformed those without DRS in nearly every aspect as shown in the chart below (Fagnant and Kockelman, 2016, pg. 9).

Austin Network-Based Model Results across Various Scenarios (serving 56,324 person-trips over 24 hours)

Measure With DRS Without DRS + 30% DRS travel time +40% DRS travel time
# SAVs 1715 1715 1643 1601
Vehicle replacement rate 10.77 10.77 11.24 11.53
Extra VMT 4.49% 8.68% 2.67% 1.52%
Avg. waiting time (min.) 1.18 1.87 1.27 1.37
Avg. PM peak waiting time (min.) 4.49 8.96 4.82 4.99
Avg. total service (min.) 14.71 14.97 15.20 15.69
% Waiting ≥ 10 min. 1.45% 5.65% 1.71% 1.90%
% Waiting ≥ 15 min. 0.22% 2.08% 0.27% 0.43%
# Shared trips 6151 0 9233 11,723
% Shared miles 4.83% 0.00% 8.32% 11.20%

As you can see, systems using DRS had reduced passenger wait time and total service time.

Effects on Vehicle Occupancy

According to a report by Dr. Daniel Fagnant and Dr. Kara Kockelman, SAVs offering DRS can increase average vehicle occupancy and reduce regional VMT (Fagnant and Kockelman, 2016). By maximizing the amount of time that an SAV is in use by more than one person at a time, SAVs become a crucial component in increasing vehicle capacity. The continued use of transit, especially automated buses and trains, will also become important in terms of increasing shared trips. The impact of incorporating sharing into the current transportation system can be illustrated in the infographic below:

Source: ITDP, 2017

The Institute for Transportation Development and Policy predicts that the number of vehicles on the road by 2050 will decrease by 1.6 billion vehicles through maximizing shared vehicle trips.


Through policy changes, shared trips can be encouraged and single occupancy travel can be discouraged. A few measures recommended by ITDP are (ITDP, 2017):

  • discourage or restrict private ownership of AVs
  • added fees for vehicular travel
  • travel subsidies for high occupancy vehicles

While these measures are likely to increase AVO, special attention should be taken to decrease empty driving as much as possible in order to further reduce congestion and VMT associated with AVs.


One of the main concerns and unknowns associated with AVs is the possibility that they will worsen congestion due to empty driving, among other factors. Measures will need to be taken to increase vehicle occupancy in order to achieve maximum benefits.

Causes of Empty Driving

When driverless cars become a reality, people are likely to take advantage of empty driving capabilities.

Empty driving will likely occur in the following scenarios:

  • shared AVs driving to pickup location
  • personal AVs driving to parking after drop-off
  • personal AVs driving home after drop-off
  • shared electric AVs (SAEVs) driving to and from charging stations
  • personal AVs sent to run errands

SAVs and Empty Driving

While SAVs will inevitably lead to a certain amount of empty driving. A study led by Dr. Jun Liu involving a large-scale micro-simulation of SAVs found that, “empty vehicle miles traveled by the fleet of SAVs ranged from 7.8% to 14.2%.” The concern is that this level of empty vehicle travel, while seemingly harmless from an environmental standpoint, “could still turn already-congested cities into far more congested cities, due to the convex nature of link performance functions,” (Liu et al., 2016, pg. 2).

Another study led by Dr. Donna Chen yielded similar results, “SAEV fleets are predicted to generate an additional 7.1 to 14.0% of travel miles,” due to empty charging and passenger pickup (Chen et al., 2016). While empty travel could be reduced by strategic charging, “model results also reveal the inherent tradeoffs between reduction of induced ’empty’ travel and improvement of user experience (as measured by wait times and percent of trips served).” In order to enhance the user experience while still reducing empty travel, a dynamic pricing scheme can be utilized, “which penalizes trips that incur more relocation miles” and “incentivizes trips that coincide with strategic relocation,” (Chen et al., 2016, pg. 16).

Similarly, IMInnovation discusses the justification of a Road User Charge (RUC) to help manage and minimize the use of empty HAVs, “the charge is likely to be set to reflect the externalities, congestion and emissions, of empty movements.” This would encourage carpooling as well as the use of transit systems (IMInnovation, 2017).


Schaller consulting (2018) released a report discussing the TNC using SAV in the future. In recent years, there exist companies that could operate in passenger service with AVs. The author of the report, Bruce Schaller, who is Deputy Commissioner for Traffic and Planning at the New York City Department of Transportation and Policy Director at the NYC Taxi and Limousine Commission, proposed “heaven and hell”, two outcomes of the SAV: “In the “heaven” scenario, people rely on SAVs and expanded public transit; electric vehicles replace gasoline power thus reducing greenhouse gas emissions; and acres of surface parking are replaced with parks, affordable housing and other active land uses. In the “hell” scenario, AVs induce sprawl as people are less concerned about long commutes; miles driven and traffic congestion increase in both cities and suburbs; empty cars cruise city streets instead of paying for parking; and public support for bus and rail service erodes, leaving lower-income people stranded. ” Whether AVs lead to heaven or hell is to depends in large part on whether people want to use shared autonomous services.

However, Schaller (2018) believes that shared, door-to-door services become a major component of urban transportation systems in the autonomous future is unlikely, since TNC experience has proven the appeal of private ride TNC service, and the service model of six-seat, on-demand, door-to-door shared rides does not appear viable. Therefore, the report notes that “What seems far more likely is the continued centrality of two time-honored modes: door-to-door private ride taxis, and fixed-route transit. Both modes can be enhanced by technologies now in use by TNCs and microtransit to provide greater transparency and manage operations in real-time, and by autonomous technologies that promise to dramatically improve safety and reduce costs.” (Schaller, 2018)