Literature Reviews

Increasing adoption of automated vehicles (AVs) is likely to impact municipal budgets through reduced revenues, increased expenditures, and possible reduction in operating expenditures by adopting automated vehicles for municipal services.

Revenue reductions are likely to be caused by reduced parking revenues and traffic fines. A discussion of the potential impact of autonomous vehicle adoption on government finances for eight Canadian governments suggests a reduction in municipal parking revenues [1]. A combination of accelerated adoption of electric vehicles through a transition to automated vehicles would reduce tax receipts from gasoline and diesel fuels as well as parking, traffic violations, and other revenues by a range of 3 - 51 percent across 7 combinations of AV/EV/Shared scenarios in five Oregon Cities [2].

Increases in expenditures may come due to a decision to subsidize mobility and from infrastructure improvements. A survey of US officials' perspectives and preparations for automated vehicles suggests that cities and transportation agencies may seek to subsidize automated mobility for low-income individuals [3]. An Australian study found that some investment in infrastructure will be needed to accommodate AVs, but there is still uncertainty about needed expenditures [4].

However, cities may also reduce operating costs by performing municipal services with automated vehicles. One study estimates that using automated vehicles for trash collection could reduce operating costs by 32 - 63 percent [5].

The introduction and potential proliferation of highly automated vehicles (AVs) present the classic challenge of balancing the freedom of private manufacturers to innovate with the government's responsibility to protect public health. AVs raise many public health issues beyond their potential to improve safety, ranging from concerns about more automobile use and less use of healthier alternatives like biking or walking, to concerns that focusing on autonomous vehicles may distract attention and divert funding from efforts to improve mass transit. There are, additionally, issues of access, especially for the poor, disabled, and those in rural environments [1].

As the classic Code of Ethics for Public Health recommends [2], public health advocates can advocate for the rights of individuals and their communities while protecting public health by helping to establish policies and priorities through “processes that ensure an opportunity for input from community members.” Public health thought leaders can ensure that communities have the information they need for informed decisions about whether and how autonomous vehicles will traverse their streets, and they can make sure that manufacturers who test and deploy autonomous vehicles obtain “the community’s consent for their implementation.” Finally, public health leaders can work for the empowerment of the disenfranchised, incorporating and respecting “diverse values, beliefs, and cultures in the community” and collaborating “in ways that build the public’s trust” [2].

Some researchers indicate that environmental impacts of automated vehicles (AVs) strongly depend on the connectivity and market penetration rates [1], [2], [3], [4]. For example, Mattas et al., [5] shows that with dense traffic, AVs that lack interconnectivity are likely to reduce speed in adherence to safety and comfort guidelines, consequently producing an additional 11 percent in emissions. Wadud et al. [6] developed an energy decomposition framework and quantified the potential percentage change of greenhouse gas (GHG) emissions from AVs depending on energy intensity effect, travel demand effects and net effects of automation. Wadud et al. [6] concluded that vehicle automation offers the potential to reduce light-duty energy consumption by nearly half, but this decrease is dependent on several factors including the degree to which energy-saving algorithms and design changes are implemented into practice and policy responses at federal, state, and local agencies, among others.

While AVs could induce demand due to easier travel and the empty travel generated from shared AV fleets [7], [8], [9], most studies show energy savings despite the Vehicle-Miles-Traveled (VMT) increase [10], [11], [12]. For example, Fagnant and Kockleman [11] estimated that shared autonomous vehicles (SAVs) may save 10 times the number of cars needed for personally owned vehicles travel but increase daily VMT by about 11 percent from empty vehicle travel. The energy use and GHG emissions could be reduced by 12 percent and 5.6 percent respectively, owing to changes in total number of vehicle starts, lower proportion of cold starts, and reduced parking needs. However, some studies also indicated an increase of emissions considering different AV penetration rates [13], [14], [15]. For example, Harper et al. [16] estimated that privately owned AVs searching for cheaper parking could increase light-duty energy use in Seattle by up to 2 percent.

In general, most studies conclude that AVs would reduce energy consumption and GHG emissions per mile driven due to improvements in operational efficiencies such as automated eco-driving, changes in vehicle size, and traffic smoothing, but there is not a clear consensus that these efficiency improvements will reduce total energy use and emissions. Current areas for future research include: 1) studying the full lifecycle environmental impacts of AVs, 2) investigating models that capture the full complexity of real-world scenarios such as dynamic traffic patterns, diverse weather conditions, varying road types, and unpredictable human behavior, 3) exploring how a fleet of electric AVs might interact with power grids, especially concerning charging demands and renewable energy integration, 4) exploring if the operational efficiencies gained from AVs, lower emissions and energy use remain as trip making and VMT increases due to empty, longer, and/or easier travel [17], [18].

Many studies show that Autonomous Vehicles (AVs) could change the layout of urban areas [1], [2], [3], potentially leading to dispersed development or densification of cities. By lowering travel expenses, AVs could influence residential and work locations, potentially leading to more pronounced urban sprawl. For example, Moore et al., [4] used a web-based survey of commuters in 2017 in the Dallas-Fort Worth Metropolitan Area (DFW) and predicted a substantial extent of urban sprawl up to a 68 percent increase in the horizontal spread of cities due to AVs. AVs could also increase urban density by decreasing the need for parking, leading to more dense and mixed use development.

AV could increase trip lengths and induce suburban and exurban development [5], [6], [7], [8]. Nadafianshahamabadi et al., [9] utilized an integrated model of land use, travel demand, and air quality. The modeling is designed for the Albuquerque, New Mexico metropolitan area to demonstrate that AVs encourage development at the urban fringe. While jobs and population typically migrate outward in tandem, trip lengths and overall travel demand continue to rise due to the relatively low density in these emerging areas compared to traditional urban employment centers. Similarly, Gelauff et al. [10] used equilibrium model to simulate spatial effects of AVs and found that population tends to increase in large metropolises and their suburbs, at the expense of smaller cities and non-urban regions given high automation with good public transport systems in Netherlands. Carrese et al. [11] used discrete choice modeling and traffic simulation to study the residential relocation due to different time perception. Results show that about 40 percent of respondents would move to the suburbs under the AV regime in Rome, Italy, and travel time would increase by 12 percent for suburban resident commuters.

Besides contributing to the development of new peripheral centers, AV has the potential to densify the existing urban landscape by reallocating space for residential, economic, and leisure activities [12]. Zakharenko [1] concluded that with the introduction of AVs, the need for daytime parking may shift to outlying areas, which would allow for denser economic activity and increased land rents in downtown areas. As AVs potentially reduce car ownership, it's anticipated that less space will be required for parking, which could give rise to more high-density and mixed-use developments [13], [14], [15]. Zhang and Guhathakurta [16] developed a discrete event simulation model to assess the impact of Shared AVs (SAVs) on urban parking land use in Atlanta, Georgia and concluded that SAV can reduce parking land by 4.5 percent at a 5 percent market penetration level and each SAV can emancipate more than 20 parking spaces. However, some research indicates that vehicles are traveling longer distances daily, and there could be an increase in parking space on the outskirts [17], [18].

In general, most studies found that private AVs can potentially lead to dispersed urban development, while SAVs are expected to contribute to densification of city centers. Current areas for future research include: 1) AV effects on people's residential and employment location decisions, recreation spaces and supply of infrastructure. 2) long-term effects of AVs on urban land use patterns to promote AV adoption with efficient use of land. 3) infrastructure adaptation to fully accommodate the new traffic dynamics and parking needs introduced by AVs [19].

Many researchers have used agent-based simulation to assess the effects of Automated Vehicles (AV)s on transportation system operations and efficiency (e.g., congestion and Vehicle Miles Traveled (VMT)) [1], [2], [3], [4], [5], [6], [7]. For example, Yan et al. (2020) simulated and then evaluated the performance of a shared autonomous vehicle fleet serving requests across the Minneapolis-Saint Paul region [1]. Yan et al. [1], [2], [3], [4], [5], [6], [7] estimated that the average shared AV could serve at most 30 person-trips per day with less than a 5 minute wait time but generates 13 percent more VMT. Yan et al. [1], [2], [3], [4], [5], [6], [7] also concluded that dynamic ridesharing could reduce shared AV VMT by 17 percent on average and restricting shared AV parking on the busiest streets could generate up to 8 percent more VMT.
Other methods such as static traffic assignment models and scenario analysis, have also been used to to understand the effect of AVs on congestion and VMT [8], [9], [10], [11], [12], [13]. For example, Harper et al. (2016) estimated the upper bound increase in travel with AVs for the non-driving, elderly, and people with travel-restrictive medical conditions by creating demand wedges and assuming that these traditionally underserved populations would travel as much as younger and/or healthier populations [9]. Harper et al. (2016) estimated that vehicle automation addressing latent demand for underserved population could increase VMT by as much as 14 percent, with females and non-drivers making up most of this increase [9].

Most studies are in agreement that AVs are likely to increase VMT and congestion, due to increased trip making, the ability for AVs to search for more distant and cheaper parking, and the additional VMT generated from people switching from personally owned vehicles to shared autonomous vehicles, generating empty travel [5], [9], [14]. Current opportunities for future research in this area include: 1) simulating AVs considering a heterogeneous population of travelers with different values of travel time (VOTT) and 2) incorporating parking to estimate the impact of AVs on transportation system operations [15].

Automated vehicle technologies hold significant promise for benefiting vulnerable populations and bridging urban-rural disparities. Demographically, numerous studies highlight the potential of automated vehicles to improve mobility for people with disabilities, elderly individuals, and low-income populations by offering accessible and affordable transportation options [1], [2], [3], [4], [5].
Automated vehicles offer a game-changing solution for individuals with disabilities, including those with vision impairments [6], [7], [8], cognitive impairments [9], [10], [11], or limited mobility [12], [13], [14]. Equipped with advanced sensors and navigation systems, these vehicles could provide safe and reliable transportation for people with disabilities. They could incorporate user-friendly interfaces and assistive technologies, such as wheelchair ramps and voice-activated controls, to ensure accessibility and ease of use [15], [16], [17]. By removing physical barriers and offering personalized assistance, automated vehicles empower individuals with disabilities to travel independently and participate more fully in their communities.
Geographically, the deployment of automated vehicles has the potential to address “transportation deserts” in small urban, rural, or remote areas, providing residents with access to essential services and opportunities that were previously out of reach [18], [19], [20]. For rural areas, where transportation infrastructure may be lacking and population densities are lower, automated vehicles, like other shared ride services, could provide on-demand mobility options and connect residents to employment opportunities, healthcare services, and education centers [21]. Similarly, in small urban areas, where public transportation may be less extensive compared to larger cities, automated vehicles could serve as a flexible and efficient transportation solution, improving mobility and access to resources for residents.
However, the literature also emphasizes the need for careful planning and implementation to ensure that these technologies do not exacerbate existing inequalities. Concerns such as the digital divide [22], [23], [24], affordability [1], [25], [26], [27], and infrastructure limitations [28], [29], [30], [31] in rural and small urban areas must be addressed to ensure that the benefits of automation are equitably distributed across demographic and geographic lines. In addition, the literature emphasizes the importance of community engagement and inclusive planning processes to ensure that the deployment of automated vehicle technologies is responsive to the needs and priorities of diverse communities [18], [32], [33], [34].

The automotive industry is undergoing a transformative shift driven by advancements in technology, changing consumer preferences, and global sustainability goals. As the industry evolves, the need for a skilled workforce equipped with the knowledge and expertise to navigate this changing landscape becomes increasingly critical. On one hand, automated vehicles (AVs) will likely displace some jobs such as taxi drivers, bus drivers, and truck drivers. On the other hand, widespread AV deployment will create new jobs and fundamentally change many others. For example, skills needed to manufacture and maintain these vehicles will be very different from those currently needed in these markets. Understanding anticipated shifts in job availability, roles and responsibilities, and required skill sets over time will serve as a crucial foundation for developing targeted training programs, implementing strategic workforce development initiatives, and ensuring that individuals possess the requisite skills and competencies to thrive in this dynamic and rapidly evolving sector [1].

The workforce shift and changes in labor demands are directly related to the acceptance of AVs. While previous study has found several elements that contribute to the shift in acceptance of AVs following education, as of 2019 there was a paucity of investigation into the specific components that influence this change at the individual level [2]. Another aspect influencing workforce development strategies and efforts in AVs is the accuracy of AV technology advancement timeframes. This is because the widespread deployment of AVs will have an influence on a variety of transportation-related jobs [3]. As a result, having an accurate AV deployment schedule will aid in the development of appropriate and suitable public policies, as well as the creation of well-planned budgets for workforce development [1].

Automated Vehicles (AVs) have the potential to prevent 95 percent of pedestrian injury crashes in the US, particularly when a driver violation or pedestrian visibility occurred more than one second before crossing [1]. Vehicles equipped with Advanced Driver Assistance Systems, such as front collision prevention/warning, lane departure prevention, emergency braking, and adaptive cruise control, are already accessible for purchase by consumers. These systems are believed to provide safety benefits due to their ability to reduce human errors in driving and minimize the likelihood of accidents. For example, Scanlon et al. [2] found that lane departure warning and lane departure prevention systems could prevent 28 to 32 percent of road departure crashes in the United States under current road infrastructure conditions. Cicchino [3] observed a 27 percent reduction in front-to-rear crash rates and a 20 percent decrease in front-to-rear injury crash rates with the use of forward collision warning systems. Furthermore, Cicchino [3] noted a 43 percent decrease in front-to-rear accident rates and a 45 percent reduction in front-to-rear injury crash rates with the implementation of low-speed autonomous emergency braking. It is estimated that over 400,000 injuries and nearly a million collisions could have been prevented in 2014 if forward collision warning with autonomous emergency braking had been installed in all vehicles nationwide [3].

Recent safety studies have focused on comparing Autonomous Driving Systems (ADS) safety with that of human drivers. For example, Kusano et al. [4] compared Waymo (an SAE Level 4 ADS) rider-only crash data to human drivers, and found a human crash rate 6.7 times higher compared to the ADS for crashes that caused injuries, and 2.2 percent higher compared to the ADS for policed-reported crashed vehicle rates. In 2024, a working group of industry, academic, and insurance experts developed the Retrospective Automated Vehicle Evaluation (RAVE) checklist, which sets out 15 recommendations to ensure the quality and validity, transparency, and accurate interpretation of retrospective ADS performance comparisons [5].