Literature Reviews

Autonomous vehicles have the potential to reduce fuel consumption through automated acceleration and braking, platooning to reduce air resistance, vehicle design, fuel switching, routing efficiency, and traffic congestion reduction [1]. However, there is also the possibility that automation of vehicles will lead to increase in vehicle usage, and subsequently fuel consumption and emissions [1].

The effect of automation of heavy-duty vehicles can reduce energy consumption and benefit the environment, depending on the fuel source [2]. One study estimated that an automated diesel heavy duty truck reduces greenhouse gas emissions by 10 percent compared to a conventional heavy-duty truck [2]. Meanwhile, an automated battery electric heavy duty truck would reduce life cycle greenhouse gas emissions by 60 percent compared to the conventional heavy-duty truck [2]. However, there are trade-offs between fuel sources for automated heavy-duty trucks, including the mineral resource losses [2]; the battery manufacturing required for automated electric heavy-duty trucks increase mineral intensity significantly compared to automated diesel heavy-duty trucks [2]. Additionally, automation decreases energy intensity of heavy-duty trucks, which decreases through automation, the increase in power generation required for electrified heavy duty trucks may outweigh the benefits from automation [2].

Further research is needed related to the effect of different electricity generation methods on automated heavy-duty truck emissions, as well as with different vehicle design and weight assumptions. Research is also needed related to environmental effects of other heavy-duty vehicles and equipment, such as automated buses and specialized equipment.

Transportation Network Companies (TNCs), or ride-hail companies, have the potential to reduce emissions by reducing single-occupancy trip distances through pooled rides and reducing the need for private vehicle ownership. Ride-hail services can also support transit use by providing riders with an option to connect to transit stations, and by complementing transit in times and places it does not operate.

Ride-hail services can, in theory, reduce emissions by linking passengers traveling in similar directions. In practice, however, those benefits are limited. Most trips are not pooled; one study found just ten percent of trips were pooled, and 27 percent involved multiple passengers [1]. Deadheading, or trips made with no passenger in the vehicle (often between where one passenger is dropped off and the next is picked up), contributes to additional emissions. A significant portion of ride-hail trip miles (40 percent, from a study of TNCs in Canada) are deadheading trips. Both pooled and unpooled ride-hail trips emit more pollutants relative to trips taken by single-occupancy vehicles [1].

Ride-hail might also reduce emissions by offering an alternative to private vehicle ownership, or by connecting riders to transit stations. The evidence is mixed regarding the extent to which riders substitute ride-hail for public transit, with studies finding that ride-hail reduces net transit ridership between 14 and 58 percent, depending on the city studied and the type of transit [2], [3]. The more abundant and reliable ride-hail becomes, particularly in urban areas with a rich array of alternative travel modes, the more likely people are to willingly shed their private vehicles [4]. Moreover, electrifying ride-hail can go a long way toward reducing greenhouse gas emissions, particularly electrifying vehicles for full-time ride-hail drivers [1], [5].

A shift from dining in to at-home consumption can produce additional food packaging waste [1]. On-demand meal delivery may also affect travel activity, potentially increasing emissions. A study of delivery data in London, United Kingdom found that meal delivery by vehicle is “highly energy inefficient, producing 11 times more GHG [greenhouse gas emissions] per meal delivered by vehicle than by bicycle” [2]. However, this study did not identify if any travel activity was displaced by the substitution of meal delivery services; future research could explore if customers order from locations further away or substitute meal delivery for home cooking, activities that would increase energy consumption and resultant emissions. Policies to support bicycle use for delivery services can mitigate these increases [3], [4].

For robotic delivery services, the literature shows that the energy consumption and emissions of robotic delivery services do not necessarily outperform traditional ones, and are related to delivery distance, electrification, and operation [1], [5], [6].

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].

There is little research available on the environmental impacts of Universal Basic Mobility (UBM) programs. In a qualitative evaluation of eight UBM programs and pilots, UC Davis researchers concluded that UBM pilot program participants increased transit use more than shared mobility relative to shared mobility services, and decreased overall personal vehicle travel [1]. These results suggest that UBM programs may reduce environmental harms of private vehicle use, but additional research is needed.

The environmental benefits of demand-responsive transit and microtransit depend on the types of trips and vehicles they are replacing and generating. In theory, microtransit programs could pool passengers and thereby reduce emissions relative to drive-alone private vehicle trips [1], particularly if they use zero-emission vehicle technology. On-demand microtransit services tend to use vans that seat between four and twelve passengers. But empty vehicles [2], combined with vehicle miles lost to deadheading (trips with no passengers), can in some cases generate more emissions than private driving tips. Microtransit programs function as paratransit in some regions, and are notoriously expensive to provide in large part because they are often underutilized.

Demand responsive transit/microtransit programs in areas with limited public transit may offer a first- last-mile connection to transit, and thus enable less intensive car use. One study of suburban microtransit programs found that the majority of microtransit trips could not have been made with fixed-route public transit, and so microtransit largely either replaced ride-hail and private driving trips or generated new trips [3]. In particular, the study identified that microtransit induced trips among people without access to their own cars, and thus generated new vehicle miles traveled. More research is needed on the emissions and energy impacts of demand responsive transit and microtransit programs.

The environmental impact of Mobility-as-a-Service (MaaS) and related business models depends on how the services are offered, and the incentives of the operator [1]. For example, if ride hailing is incentivized over public transit and bike-shares, there would be fewer environmental benefits [2]. Additionally, private operated mobility services are generally focused on maximizing revenue, while public transport operators may focus more on public benefits including reduced environmental impact [3]. A study assessing welfare impacts of MaaS found that MaaS schemes with shared mobility have the potential to substantially reduce energy consumption, and even greater reductions were possible with improved cost transparency for use of cars and inclusion of externalities such as greenhouse gas emissions in the generalized cost [4].

Carshare’s emissions and fuel consumption depend on several factors, including the types of trips they are substituting for, how many new trips they generate, and how the vehicles are fueled. User demographics play a key role in determining these factors; for example, if people are primarily joining a carshare to shed their private car or delay purchasing one and thus reduce their overall car trips, then the program may reduce emissions [1]. If, however, users generally come from car-less or car-lite households, joining the carshare program may create more trips than subtract them, and thus may increase emissions. Carshare programs that use electric vehicles, such as BlueLA, reduce tailpipe emissions [2]. Electric carshare programs also expose users to cleaner vehicle types, and carshare users are more likely to later select electric cars when purchasing a vehicle than non-users [2]. However, even zero-emission vehicles generate fine air particulate emissions from tire friction, which worsens local air quality.

Carshares can offer emissions and energy savings by reducing net private automobile travel and by replacing more polluting private fleets with cleaner shared fleets. While some carshare trips replace public transit trips, in aggregate and when taking the life cycle energy and emissions impacts into account, carshares reduce net household greenhouse gas emissions [3]. One study found a majority of surveyed North American carshare users traveled more by car after joining the program, thus increasing emissions, but that those emissions increases were outweighed by emissions saved by users who gave up their personal cars [1]. Studies find that a carshare vehicle tends to replace approximately 15 private vehicles [4], [5]. Carshare vehicles may also be more fuel efficient than the overall private car fleet [6]. Another study, based on the same survey of North American carshare users, found a significant increase in walking, cycling and carpooling among people who joined a carshare [7].

While the market for zero emissions vehicles is growing, zero emission vehicles still exceed the budgets of some low-income potential drivers. Electric carshare programs offer a low-risk way to improve access to clean mobility among car-light or car-free households, without requiring users to pay the full costs of owning and maintaining their own electric vehicle [8].

In disadvantaged communities with especially high rates of air pollution, electric carshare programs may help reduce localized air pollution [9].

Connected autonomous vehicles (CAVs) are expected to optimize energy efficiency due to improved operational efficiencies and by moderating movements of automated vehicles (AVs) through Cooperative Adaptive Cruise Control (CACC), platooning, eco-driving strategies, Vehicle-to-Everything (V2X) communication and incorporation of various dynamic routing systems [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15]. For example, Djavadian et al., [16] proposed a dynamic multi-objective eco-routing strategy for connected & automated vehicles (CAVs) and implemented in a distributed traffic management system which shows the potential of reducing GHG and NOx emissions by 43 percent and 18.58 percent, respectively. Similarly, the eco-drive system for connected and automated vehicles proposed by Ma et al., [17] shows that more than 20 percent of fuel consumption can be saved. Mattas et al., [147] shows that while AVs lacking interconnectivity would likely increase emissions, a network of CAVs could lead to a decrease in carbon dioxide emissions of up to 5 percent.

V2X technology has potential to improve energy efficiency through applications such as traffic-light-to-vehicle communication, which can create energy savings and increased driving range [18]. However, vehicular communication systems also require infrastructure and energy to support [19]. Additional research is needed to understand potential environmental impacts of V2X technology, and whether there will be a net benefit when it comes to energy efficiency.

Micromobility has mixed implications for urban transportation sustainability. A comprehensive study of 500 travelers revealed that while personal e-scooters and e-bikes tend to reduce carbon dioxide emissions compared to replaced transport modes, their shared counterparts might increase emissions [1]. Another emphasized the potential of micro-mobility to reduce greenhouse gas emissions, but highlighted that the real impact depends heavily on what transport modes are substituted, the types of trips, and the specific urban contexts, and suggests that existing shared micromobility programs often substitute for active transportation.[2] Policies and infrastructure adapted to these realities can enhance the benefits of micro-mobility. Systematic reviews further underscored that the shift to e-mobility often replaces walking and public transport, which could lead to increased energy demands - this is, however, not an intrinsic property, but a product of the availability of the service, ease of docking, and perceived safety of the service [2].