Toward Human-Centric Transportation and Energy Metrics: Influence of Mode,Vehicle Occupancy,Trip Distance,and Fuel Economy

Abstract

Traditional metrics measuring transportation and energy outcomes can be augmented to better represent impacts on people’s lives and systems-level performance. This study introduces, analyzes, and tests two novel metrics: human-centered road capacity (road capacity for people) and energy intensity (energy use for people’s transportation) using empirical cumulative distribution functions of associated parameters for scenario development. Current national-level distributions of available data in the United States for factors contributing to the two new integrated metrics are used as context to evaluate potential outcomes. These factors include vehicle occupancy, mode share, fuel economy, and trip distance. Variations in input values provide insights on how these factors shape efficiencies in road capacity and energy intensity. Parametric sensitivity analysis indicates that the impact of each input depends on the metric being evaluated. For the human-centered road capacity mobility metric, increasing vehicle occupancy has the largest effect—twice that of increasing mode share for bike, walk, and transit. For the energy intensity mobility metric, the effect of improving fuel economy is the largest. Additionally, a novel interactive tool to visualize the results for various parameter combinations is designed to allow researchers and decision makers to test the metrics. The findings show deficiencies in continuing to use traditional vehicle-centric metrics and suggest that the diffusion of new human-centric metrics that benchmark outcomes associated with road capacity and energy may be significant in motivating new sustainable transportation investments and efficient utilization of infrastructure, mobility assets, and services.

Keywords

data and data science multimodal information systems multimodal analysis analytic data visualization pedestrians bicycles human factors accessibility planning and analysis access/accessibility sustainability and resilience transportation and sustainability accessibility and mobility

Throughout history, transportation systems have served humanity by providing mobility and access to services and opportunities at greater distances, thus advancing societies and connecting cultures worldwide. Several challenges still lie ahead for transportation, however, including energy consumption, pollution, emissions, congestion, safety, travel time reliability, equity, and affordability ( 1 – 3 ). In general, improved efficiency and productivity in the use of assets and resources can offer important pathways to deliver sustainable transportation ( 4 , 5 ). Additionally, transportation itself is not what we value; rather, we find utility in the places, things, and people we are going to or accessing. As Mumford ( 6 ) stated, “A good transportation system minimizes unnecessary transportation.”

This article contributes two new human-centric metrics: person per hour per lane (PPHPL) for the person capacity of a lane of road; and person energy intensity for transportation (PEIT). These are designed to fill a gap in the metrics by placing emphasis on transportation as the movement (and access) of people and goods, beyond vehicles. These metrics will be increasingly important as many new systems and services consider both user-relevant (e.g., affordability, safety, reliability) and system-level outcomes (e.g., efficient utilization of fuel, infrastructure, land, human and financial capital) for energy-efficient mobility and accessibility. An initial hypothesis is that understanding human-centric mobility parameters and models tied to measurable outcomes will become an increasingly critical component of decision processes that focus on investments in people, extending beyond the transportation realm of investing in vehicle assets and movement.

This paper first considers traditional vehicle-centric metrics and the parameters used to calculate them. Critical elements missing from traditional metrics are then added, while acknowledging that this is far from an all-encompassing suite of metrics. The study addresses a key gap by comparing differences between old metrics and the new proposed metrics. A unique contribution is a sensitivity analysis of individual and societal metrics to all input parameters (average vehicle occupancy, trip distances, mode share, energy efficiency and width multiplier).

As noted in the literature, emerging shared-use services (e.g., ride-hailing, micromobility) and automated vehicles will lead to shifts in mobility outcomes ( 7 , 8 ). These services also reveal that new methods are required to assess system performance such as the case with free-floating carsharing ( 9 , 10 ). A more integrated package of metrics is therefore necessary that allow for energy and space utilization measurements with relevant human factors of vehicle occupancy, mode choice, and distance traveled to inform new goals, systems management, or performance evaluation processes.

This paper is divided into four sections: literature review, methods and data, results, and discussion and conclusion. The analysis and scenarios in the methods and results focus on specific parameters, their distributions, and—with inputs from experts—the relationship and mapping of new metrics to emerging priorities for future mobility decision processes. The paper concludes with initial discussion of how both the terms and emerging metrics for mobility energy productivity and infrastructure productivity can aspire to consider new human-centric space, and how energy-effectiveness metrics help inform sustainable mobility futures.

Literature Review

Transportation agencies use modeling and analysis to evaluate plausible future outcomes, but often certain types of relevant outcomes are ignored from the beginning. As a motivation for this study, key gaps identified are the ineffective metrics and measurement methods for understanding the impacts of transportation and energy at the individual level to serve people in a more space- and energy-efficient manner. This includes a current weakness of travel demand models in having limited capability to account for different modes, variations in nonmotorized trips, average vehicle occupancy across different regions, or person miles traveled per vehicle miles traveled, which may be increasingly relevant to study the sustainability impacts of emerging services such as ride-hailing and automated vehicles. Even with today’s metrics and after project completion, outcomes are often far from actual performance, missing realized gains, real-world optimization opportunities, or goals of integrated solutions for communities. For example, Yang et al. ( 11 ) note how flawed design and performance evaluation approaches can create mismatches in supply and demand that do not consider users’ needs for mobility or efficient access, time, cost, or other components associated with designing for and with people. Merlin ( 12 ) addresses the importance of more people in fewer vehicles to achieve key sustainability goals, more so than automated vehicles alone. Jones and Leibowicz ( 13 ) explore the value of ensuring higher levels of shared or “pooled” uses of transportation assets. Brown and Dodder ( 14 ) address the uncertainty in travel demand and possible energy and emissions implications of fully connected and automated vehicles, similar to Bush et al. ( 7 ), who focus on key levers for more efficient mobility systems, including emphasis on a range of occupancy levels that may be achieved in future automated vehicle trips. Boisjoly and El-Geneidy ( 15 ) and the Committee of the Transport Access Manual ( 16 ) point to accessibility—measured as the ease of reaching valued destinations—as a key land use and transport performance metric. In addition to key gaps and weaknesses in travel demand models, effective metrics are essential to design transportation demand management (TDM) measures to account for the co-existence between conventional and more advanced transport systems, the need for a regulatory framework able to manage transport demand by ensuring goals beyond economic benefits such as environmental sustainability and social equity, and the need to account for new technologies and services. The promises of electric, shared, automated, and connected vehicles solving transport problems need to be coupled with TDM strategies to become a reality ( 17 , 18 ). This approach complements the need for new metrics in the state of the art.

This literature review will next focus on established vehicle-centric transportation and related energy performance metrics. This includes emphasis on literature related to metrics and transitions in metrics, initially focused on: (i) the movement of vehicles, road capacity, and past engineering standards and (ii) energy use in transportation systems.

Vehicle-Centric Transportation Metrics

Two metrics that have been used in analyzing infrastructure performance at section and network levels for decades ( 19 ) are annual average daily traffic and vehicle miles traveled (VMT). They are limited, however, as they do not include elements that are more critical with evolving services (e.g., ride-hailing, automated, and shared-use vehicles) such as deadheading, reallocation, and redistribution—which involve additional VMT when vehicles are not in service to users ( 20 – 22 ).

These evolutions are occurring in parallel with traffic engineers still using fundamental traffic flow diagrams including density (vehicles per unit distance) and velocity (how fast the vehicles are traveling) to measure traffic flow as “vehicles per hour per lane” (VPHPL). This metric does not include additional elements such as road width, vehicle size, nonmotorized modes, vehicle occupancy, or trip distance ( 23 ), and only recently has research informed practices of performance measurement focused on understanding road usage patterns in urban areas ( 24 ).

Similarly, traffic engineers have historically used level of service (LOS) standards from the Highway Capacity Manual to design roads, with the idea of meeting certain criteria based on road capacity and density of the road ( 25 ). Although the LOS standard is useful, it can create bias, inefficiencies, and unintended consequences to the system. For example, widening a road to improve LOS (for vehicles) will also worsen conditions for pedestrians and other modes. Widening a road will also attract more traffic, a concept known as induced demand ( 26 ). Some states, such as California, are moving away from LOS standards and looking at measuring VMT and energy or emissions per passenger mile traveled ( 27 , 28 ). Additional studies have also considered the importance of linking metrics for accessibility and equity with mode share ( 29 ), person miles per vehicle miles traveled with energy and emissions intensity ( 30 ), and vehicle occupancy for decarbonization of U.S. passenger transport with electrification and automation uncertainty ( 31 ). Accessibility analysis recognizes the full range of factors that affect the ability to access services and activities, including multimodal mobility, network connectivity, proximity, travel information, and affordability ( 16 ). Attention to land use and accessibility metrics, as distinct from mobility metrics alone, has been noted in relation to implications and benefits ( 32 ).

Energy-Centric Transportation Metrics

Historically, and based on a review of the official U.S. government fuel economy website (fueleconomy.gov), currently the emphasis in energy-related metrics is placed on vehicle ownership and the miles-per-gallon efficiency of different vehicles ( 33 ). Another example at the state level is legislation SB375 in California, focusing on how best to reduce energy use and emissions in the transportation sector through a reduction in VMT, requiring travel models from the metropolitan planning organizations to demonstrate how certain levels of reductions may be achieved through various management strategies ( 34 ).

A useful summary of new mobility-related energy metrics shows that recent studies have focused on the energy productivity of mobility (human-centered focus on accessibility weighted by time, costs, and energy) and consider other elements such as the quality of mobility ( 35 ). This study builds on these efforts, with a focus on parameter variation, and complements interests in linking mobility, land use, and accessibility. This effort emphasizes a multimodal approach on access to opportunities and includes cost, time, and energy inputs (including on a passenger-mile basis). This study also focuses on a fine grained spatial resolution, tailored to individuals, to fill gaps of historic metrics.

There are additional considerations on the use of renewable or clean energy for mobility choices, with increased interest in this topic specifically for the clean electrification of mobility systems ( 36 – 38 ). As an example analysis of future transportation scenarios, the U.S. Department of Energy’s SMART Mobility Laboratory Consortium developed scenario results on changes in energy impacts based on vehicle miles traveled, person miles traveled, and speeds with lower or higher sharing (or occupancy) and comparing mobility-as-a-service deployments of automated vehicles to personally owned connected and automated vehicles ( 39 ).

Methods and Data

As described in the literature review, multiple metrics can be developed to analyze transportation projects. To better understand the importance of different metrics and prioritize future scenarios in the U.S.A., two metrics for further analysis and scenario development are developed: (i) human-centered road capacity enabled by transportation infrastructure and (ii) person energy intensity for transportation (Table 1). These integrated metrics are developed with different individual data inputs. First the metrics are described and then the parameters and specific data sources (and assumptions) for each parameter are listed.

Table 1.

Traditional and Proposed Metrics for Review of Integrated Mobility, Energy, and Costs

Area	Traditional vehicle-centric metric	New human-centric metric
Road capacity	Vehicles per hour per lane	Persons per hour per lane
Energy use	Miles per gallon	Person energy intensity for transportation

Road Capacity

Traffic engineers use traffic flow diagrams to measure VPHPL and design using criteria such as road width, speeds, and braking distance. Moving from a vehicle-centric metric, this paper proposes a new human-centric metric, “persons per hour per lane” (PPHPL), to be more inclusive by acknowledging several modes of transportation using the system and include additional parameters such as vehicle occupancy, mode width space requirement, and trip distance. PPHPL is defined by Equation 1:

PPHPL = \sum_{j = 1}^{n} VPHP L_{j} \times M S_{j} \times AV O_{j} \times W M_{j}

(1)

where n is the number of modes using the transportation infrastructure, VPHPL is the traditional vehicles per hour per lane of mode j, MS is the percentage of mode share for mode j, AVO is the average vehicle occupancy for mode j, and WM is the width multiplier representing the number of users per mode that can pass through the width of the lane. To expand beyond mobility and acknowledging the need to measure accessibility as the ease of reaching valued destinations at shorter distances, an additional variation of PPHPL is included. This is trip distance (TD), an additional parameter to represent the number of people that can travel through a road stretch (20 km as the chosen example) per hour per lane (multiplying the right side of Equation 1 by 20 km/TD), named PPHPL-20 km. Because—to the authors’ knowledge—there is no national data set on VPHPL, this parameter was derived using a range of values for vehicle speeds, length of vehicle, and safe following distance, leading to the number of vehicles that can typically go through a lane in an hour. Figure 1 compares the traditional VPHPL metric and the proposed PPHPL metric.

Figure 1.

Comparison of road capacity metrics (traditional versus human-centered).

Person Energy Intensity for Transportation

In this section, a person-centric metric is created that includes energy use (and accessibility) for moving people in the transportation system. Note that for each of these approaches, it is possible to create a more detailed estimate by modeling individual agents. However, for simplicity, the authors have chosen to use population-level metrics because large-scale impacts are being estimated. This simplicity allows planners to quickly iterate through various scenarios without the computational complexity of a full simulation. First, energy consumption (EC) is determined by Equation 2 depending on quantity of travel (QT), mode share (MS) of travel, energy efficiency (EE) of the vehicle used for travel, and average vehicle occupancy (AVO) for the vehicle used for travel.

EC = \sum_{j = 1}^{n} QT \times M S_{j} \times (\frac{1}{E E_{j}}) \times (\frac{1}{AV O_{j}})

(2)

where n is the number of modes, QT is the total distance traveled by a person in a day (or year) using all modes (e.g., kilometers or miles), MS is the percentage of mode share (based on distance) for mode j, EE is the energy efficiency for vehicle of mode j (e.g., kilometers per fuel-liter or mpg), and AVO is the average vehicle occupancy (i.e., number of people) for mode j. The units of EC are in volume of fuel (liters or gallons) per person per day (or year).

The person energy intensity for transportation (PEIT) metric is then calculated using Equation 3 (PEIT₁) by dividing the energy by person distance traveled (as a mobility energy use measure) and using Equation 4 (PEIT₂) to calculate more of an accessibility energy use measure depending on the number of trips by a person in a day (or year).

PEI T_{1} = \frac{EC}{QT} = \sum_{j = 1}^{n} M S_{j} \times (\frac{1}{E E_{j}}) \times (\frac{1}{AV O_{j}})

(3)

PEI T_{2} = \frac{EC}{Trips} = \sum_{j = 1}^{n} TD \times M S_{j} \times (\frac{1}{E E_{j}}) \times (\frac{1}{AV O_{j}})

(4)

where n is the number of modes, TD is trip distance (e.g., kilometers or miles), MS is the percentage of mode share for mode j, EE is the energy efficiency for vehicle of mode j (e.g., kilometers per fuel-liter or mpg), and AVO is the average vehicle occupancy (i.e., number of people) for mode j. The units for PEIT₁ are in volume of fuel per distance (liters/km or gallons/mile) per person, and the units for PEIT₂ are in volume of fuel (liters or gallons) per trip per person.

Results

For each of the individual parameters used to explore influences on and variations in “road capacity” and PEIT metrics, statistical empirical distributions are built based on available data, including the 2017 National Household Travel Survey (NHTS), regional bus transit agencies, and U.S. fuel economy data sets. The NHTS data were filtered to include trips up to 100 mi for the transportation modes of walking, biking, private vehicle, and transit bus. For each parameter, 20,000 random samples of 200 trips each from this filtered data set were selected and the sample means for each parameter were used to construct corresponding cumulative distribution functions (CDFs) in Python. More details about the sampling process for each parameter are given below:

AVO: The AVO distribution for private vehicles was constructed using samples from the NHTS data set, further filtered to include only private vehicle trips.

MS: The distributions for walking/biking and transit bus were obtained by using sample means for the number of trips in each mode divided by the total number of trips of all modes, analyzed to construct empirical distributions using the NHTS database. For private vehicles, the MS is obtained by subtracting from 1.0 the mode shares for walking, biking, and buses.

WM: The width multiplier is one for private vehicles and bus transit and 2.5 for biking and walking because two or three users in these modes can easily pass through the width of a typical road lane (the actual number could be higher; this is used as a conservative assumption).

EE: The U.S. fuel economy website (fueleconomy.gov) provides data for energy efficiency of vehicles in miles per gallon (mpg) ( 33 ).

TD: Travel distances for walking and biking, private vehicles, and buses are obtained by using sample means for the trip distances based on the NHTS data set to construct empirical distributions.

The following figures present the empirically generated CDF for each of the following parameters at a national level for the U.S.A.: mode share (Figure 2), average vehicle occupancy (Figure 3), trip distance (Figure 4), and vehicle energy efficiency (Figure 5).

Figure 2.

Cumulative distribution function (CDF) of mode share for transit buses, walk/bike, and private vehicles.

Figure 3.

Cumulative distribution function (CDF) of average vehicle occupancy for private vehicles and transit buses.

Figure 4.

Cumulative distribution function (CDF) of travel distance for walk/bike, private vehicles, and transit buses.

Figure 5.

Cumulative distribution function (CDF) of vehicle energy efficiency in kilometers per liter of gasoline (km/L) for transit buses and private, light-duty vehicles (cars and light trucks).

After creating the vector for each parameter, the parameters based on Equations 1, 3, and 4 were combined to calculate a distribution of metric valuations for road capacity and PEIT. Finally, sensitivity analysis was performed by varying one vector parameter at a time while keeping all other vector parameters constant, which makes it possible to model scenarios using statistical variation of individual vector parameters and determine probabilistic ranges to better understand the influence of each parameter for metric variation. The following subsections present the distribution of results per metric, based on the available data sources noted.

Road Capacity

Figure 6 presents results for the road capacity metric, showing the traditional metric (VPHPL) and the proposed metric with (PPHPL2) and without (PPHPL1) trip distance in the y-axis for private vehicles, transit buses, and walk/bike (x-axis).

Figure 6.

Violin plots of road capacity metrics (VPHPL and PPHPL) for private vehicles, transit buses, and walk/bike.

As seen in Figure 6, the traditional VPHPL metric shows that private vehicle modes have higher road capacity when the focus is to measure vehicles, whereas the PPHPL metric, focusing on measuring movement of people, provides a more nuanced understanding that private vehicles can be moving fewer people than a transit bus or walk/bike mode. When trip distance is considered (PPHPL2 or PPHPL-20 km), walk and bike modes become more efficient, as they can move more people per hour per lane on a 20 km road stretch. At the same time, walk and bike modes tend to be used for shorter trips, so trip distance is an important caveat.

Figures 7 and 8 present sensitivity analyses of how much the new road capacity metric PPHPL could increase by varying the individual parameter one standard deviation in the direction that the metric could increase. The traditional VPHPL metric stays constant, as it does not consider either of the new parameters being evaluated.

Figure 7.

Cumulative distribution function (CDF) of PPHPL1 comparing a baseline versus two scenarios by adjusting each parameter (average vehicle occupancy and mode share for transit bus and walk and bike) by one standard deviation.

Figure 8.

Cumulative distribution function (CDF) of PPHPL2 (PPHPL-20 km) comparing a baseline versus three scenarios by adjusting each parameter (average vehicle occupancy, mode share for transit bus and walk and bike, and trip distance) by one standard deviation.

In Figure 7, increasing average vehicle occupancy by one standard deviation results in a greater effect on the PPHPL1 metric than increasing mode share of walk and bike, and public transit by one standard deviation. Compared with the baseline, increasing vehicle occupancy by one standard deviation improves the median of the PPHPL1 metric by approximately 6%. Increasing the mode share of bus, walk, and bike by one standard deviation improves the median of PPHPL1 metric by approximately 3%. In other words, the potential gain of the PPHPL1 metric by increasing vehicle occupancy is twice the potential gain of increasing mode share for transit bus, walk and bike.

When considering travel distance for the PPHPL2 metric (Figure 8), the effect of vehicle occupancy and mode share is flipped. Compared with the baseline, increasing vehicle occupancy by one standard deviation improves the median of the PPHPL2 metric by approximately 25%. Increasing the mode share for bus, walk, and bike by one standard deviation improves the median of the PPHPL2 metric by approximately 36%. Not surprisingly, the greatest effect for the PPHPL2 metric comes from lowering trip distance by one standard deviation, with an effect of approximately 53% more capacity in PPHPL on a 20 km road stretch. This means that providing valued destinations at shorter distances (e.g., increasing density) would significantly improve the efficiency of a human-centered road (or integrated mobility) system.

Person Energy Intensity for Transportation

The results for PEIT are presented in Figure 9, showing the distribution of the energy consumed for person travel in fuel liters per kilometer (PEIT1) and fuel liters per trip (PEIT2).

Figure 9.

Cumulative distribution function (CDF) of person travel in fuel liters per kilometer (PEIT1) and person travel in fuel liters per trip (PEIT2).

Figures 10 and 11 present sensitivity analyses of how much the PEIT metric changes by varying individual parameters one standard deviation. Both the PEIT1 and PEIT2 metric figures show how a decrease in PEIT will mean less energy consumed for transportation, with units of PEIT1 in volume of fuel per distance (liters/km or gallons/mile) per person and the units for PEIT2 in volume of fuel (liters or gallons) per trip per person.

Figure 10.

Cumulative distribution function (CDF) of PEIT1 comparing baseline versus three scenarios by adjusting each parameter (average vehicle occupancy, mode share for transit bus and walk/bike, and vehicle fuel economy) by one standard deviation.

Figure 11.

Cumulative distribution function (CDF) of PEIT2 comparing baseline versus four scenarios by adjusting each parameter (average vehicle occupancy, mode share for transit bus and walk/bike, vehicle fuel economy, and trip distance) by one standard deviation.

Results from Figure 10 show how increasing vehicle fuel economy by one standard deviation has the greatest impact on the PEIT1 metric, as an approximately 14% reduction in fuel liters per kilometer compared with baseline. Increasing average vehicle occupancy by one standard deviation has the second-highest effect on the metric (≈4.5% reduction compared with baseline), and the lowest effect comes from increasing mode share of walk/bike and bus by one standard deviation (≈2.5% reduction compared with baseline).

When the focus is on energy consumed in fuel liters per trip (PEIT2) as presented in Figure 11, the effect of lower trip distance is similar to the vehicle fuel economy (reduction of ≈15% compared to baseline). Mode share and vehicle occupancy have similar effects to improve the PEIT2 metric (reduction of ≈5% compared to baseline), but only about one-third of the effect compared with improving vehicle fuel efficiency or reducing trip distances.

Interactive Exploration of Parameter Impact

The previous figures illustrate the impact of individual parameters on the metrics. However, achieving relevant goals will likely involve modifying multiple parameters at the same time, which may have complementary or even multiplier effects on relevant outcomes or targets set with these or other new metrics. Further, the extent to which parameters can be modified likely depends heavily on existing geographical, climate, and land use patterns. For example, a dense metropolitan urban area may choose to focus on mode shifts to transit and increase vehicle occupancy rates, whereas a Sunbelt suburb might instead focus on increasing density (lower trip distance) and nonmotorized infrastructure (shifts to walking and bicycling) and improving fuel efficiency.

Introducing an interactive component into the previous figures can empower engineers and policymakers to experiment with different, locale-specific options; implications for decarbonization goals; and new data collection methods around vehicle occupancy, mode share, and fuel economy (or future zero-emission vehicles) to help reach their goals. Such components and add-on possibilities can also enhance the impact of these metrics by providing software tools to make them practitioner-ready, useful, or useable for scenario planning. An initial version of an interactive open-source tool has been built as a companion to this paper. It implements the parameter distributions and metric calculations as Python modules. The modules are used in two Jupyter notebooks—a static version that reproduces the graphs from this article and an interactive version that uses ipywidgets to add interactive controls to the graphs. Practitioners can run copies of the notebooks in the cloud using Binder and experiment with various parameters. The intent of including this tool as a companion to this paper is to spark a more wide-ranging discussion on human-centered metrics and their customization based on regional or local data.

Discussion and Conclusion

Consistent with the U.S. Department of Energy’s Energy Efficient Mobility Systems program vision for “an affordable, efficient, safe, and accessible transportation future in which mobility is decoupled from energy consumption” ( 40 ), the objective of this study is to highlight key human-centric metrics and parameters shaping such outcomes. The study also aims to align with the need identified by the U.S. Department of Transportation for new integrated metrics that use human-centered approaches. The study may also be useful in the context of the work by both departments to assess potential evolutions to new and future mobility choices characterized by increasing integration of automated, connected, efficient, and shared transportation ( 41 , 42 ). These new options are documented as having the potential to improve energy efficiency, land use, affordability, accessibility, mobility, connectivity, and quality of life ( 43 ). However, these desired benefits may not be realized if the management of transportation systems continues to focus on the movement of vehicles rather than the movement of people.

This paper shows that setting up human-centered metrics and including key parameters such as vehicle occupancy, mode share, fuel economy, and trip distance matters to better understand the efficiency of transportation and energy (making a stronger quantitative case to go beyond vehicle-centered metrics such as the movement of vehicles). When focusing on vehicles for road capacity, a mode like transit seems to be less efficient than private vehicles. When the focus is on moving people, however, transit is shown to be more efficient. Based on the data sets for this study, for potential increases in the new road capacity metric for people focused on mobility (PPHPL1), the net effect of increasing vehicle occupancy is twice as much as the net effect of increasing the mode share of walk/bike and public transit. When the focus is on accessibility (PPHPL2), the net effect of reducing trip distance is more than twice the net effect of increasing vehicle occupancy, and about 1.5 times the net effect of increasing the mode share of walk/bike and public transit.

Not surprisingly, the fuel efficiency of the vehicle has the highest impact on the person energy intensity for the transportation metric focusing on mobility (PEIT1), with moderate effects from increasing average vehicle occupancy and lower effects from increasing mode share of walk/bike and bus. The net effect of improving fuel efficiency is more than three times the effect of improving vehicle occupancy and more than five times the effect of increasing mode share of walk/bike and bus. For energy intensity per trip focusing on accessibility (PEIT2), the net effect of fuel economy is very similar to the net effect of trip distance, and approximately three times the net effect of increasing vehicle occupancy or increasing mode share of walk/bike and public transit. Acknowledging the limitations of this study, it will be important to assess the relevance and ease of implementation of these new metrics versus other human-centered metrics such as equity, safety, and affordability.

More broadly, several questions still lie ahead. Will new understandings of outcomes related to people and asset utilization change investment strategies? How can we incorporate qualitative factors (e.g., risk perceptions and public trust as they relate to safety and quality of service) that come into play when we focus on human-centered transportation? This paper focuses on clearer paths first, but recognizes a need for qualitative directions, additional mode considerations (e.g., automated, electric, and shared mobility or even telework as a new choice), and specific regional considerations. In the long term, some variations of human-centered metrics and input parameters—over vehicle-centered metrics—may prove helpful and novel to inform future data collection, analysis, goal setting, investments, ongoing performance evaluation, and scenarios that help inform planning or decision-making. The authors also acknowledge that there are other outcomes and metrics that deserve attention—including travel time, utility, safety, comfort, environment, and other societal, economic, or health benefits and risks ( 44 – 46 ), and that there is an additional opportunity to analyze the variations in relative importance of metrics based on scale, context, subpopulation characteristics (e.g., income, age, gender), or different spatial-temporal elements (e.g., rural, suburban, or urban, and peak demand versus weekend travel hours). Although the focus of this paper was on demonstrating the importance, application, and variability of new human-centric metrics with integration of several data sets and selective explanatory factors at a national level, confounding factors could be considered in future analyses. In addition, the interactive tool is an example of the customized and flexible analyses that can be performed with more detailed and localized data at city, community, regional, or state levels.

Footnotes

Acknowledgements

This work was authored in part by the National Renewable Energy Laboratory (NREL), operated by Alliance for Sustainable Energy, LLC, for the U.S. Department of Energy (DOE) under Contract No. DE-AC36-08GO28308. A special thanks to Shawn Johnson at U.S. Department of Transportation (U.S. DOT); Erin Boyd, David Anderson, and Danielle Chou at DOE; and Alex Schroeder at NREL for enabling connections within the DOE and U.S. DOT ecosystems. The authors also acknowledge Scott Lian, Joseph Stanford, and Don Pickrell at Volpe National Transportations Systems Center, along with multiple NREL colleagues and participants in their project workshop hosted alongside TRB 2020 with the American Council for an Energy-Efficient Economy. Finally, thanks to the NREL team for software package development and submittal. The team filed SWR-21-13, launching a notebook in the cloud so others can experiment.

Author Contributions

The authors confirm contribution to the paper as follows: study conception and design: A. Henao, J. Sperling, S. Smith; data collection: D. Weigl, S. Atnoorkar, A. Wilson; analysis and interpretation of results: A. Henao, D. Weigl, K. Shankari; draft manuscript preparation: A. Henao, J. Sperling, D. Weigl, S. Atnoorkar, A. Wilson, E. Nobler. All authors reviewed the results and approved the final version of the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy Vehicle Technologies Office, and the U.S. Department of Transportation, Office of the Secretary for Research and Technology (OST-R).

ORCID iDs

Alejandro Henao

Joshua Sperling

Dustin Weigl

Swaroop Atnoorkar

Alana Wilson

Erin Nobler

K. Shankari

Scott Smith

The views expressed in the article do not necessarily represent the views of the U.S. Department of Energy, the U.S. Department of Transportation, or the U.S. Government. The U.S. Government retains and the publisher, by accepting the article for publication, acknowledges that the U.S. Government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this work, or allow others to do so, for U.S. Government purposes.

References

Center for Neighborhood Technology. Housing + Transportation Affordability Index. H+T Affordability Index. https://htaindex.cnt.org.

Litman

Transportation Affordability: Evaluation and Improvement Strategies. Victoria Transport Policy Institute, 2017.

Pew. Household Expenditures and Income. https://www.pewtrusts.org/en/research-and-analysis/issue-briefs/2016/03/household-expenditures-and-income. Accessed October 4, 2021.

Environmental Protection Agency. Guide to Sustainable Transportation Performance Measures. Publication EPA 231-K-10-004. United States Environmental Protection Agency, 2011.

Lah

Decarbonizing the Transportation Sector: Policy Options, Synergies, and Institutions to Deliver on a Low-Carbon Stabilization Pathway. WIREs Energy and Environment, Vol. 6, No. 6, 2017, p. e257. https://doi.org/10.1002/wene.257.

Mumford

The Highway and the City. New American Library New York, NY, 1964.

Bush

Vimmerstedt

Gonder

Potential Energy Implications of Connected and Automated Vehicles: Exploring Key Leverage Points Through Scenario Screening and Analysis. Transportation Research Record: Journal of the Transportation Research Board, 2019. 2673: 84–94.

Gruel

Stanford

J. M.

Assessing the Long-Term Effects of Autonomous Vehicles: A Speculative Approach. Transportation Research Procedia, Vol. 13, 2016, pp. 18–29. https://doi.org/10.1016/j.trpro.2016.05.003.

Ampudia-Renuncio

Guirao

Molina-Sánchez

Engel de Álvarez

Understanding the Spatial Distribution of Free-Floating Carsharing in Cities: Analysis of the New Madrid Experience through a Web-Based Platform. Cities, Vol. 98, 2020, p. 102593. https://doi.org/10.1016/j.cities.2019.102593.

10.

Firnkorn

Müller

What Will Be the Environmental Effects of New Free-Floating Car-Sharing Systems? The Case of Car2go in Ulm. Ecological Economics, Vol. 70, No. 8, 2011, pp. 1519–1528. https://doi.org/10.1016/j.ecolecon.2011.03.014.

11.

Yang

van Dam

K. H.

Majumdar

Anvari

Ochieng

W. Y.

Zhang

Integrated Design of Transport Infrastructure and Public Spaces Considering Human Behavior: A Review of State-of-the-Art Methods and Tools. Frontiers of Architectural Research, Vol. 8, No. 4, 2019, pp. 429–453. https://doi.org/10.1016/j.foar.2019.08.003.

12.

Merlin

L. A.

Transportation Sustainability Follows From More People in Fewer Vehicles, Not Necessarily Automation. Journal of the American Planning Association, Vol. 85, No. 4, 2019, pp. 501–510. https://doi.org/10.1080/01944363.2019.1637770.

13.

Jones

E. C.

Leibowicz

B. D.

Contributions of Shared Autonomous Vehicles to Climate Change Mitigation. Transportation Research Part D: Transport and Environment, Vol. 72, 2019, pp. 279–298. https://doi.org/10.1016/j.trd.2019.05.005.

14.

Brown

K. E.

Dodder

Energy and Emissions Implications of Automated Vehicles in the US Energy System. Transportation Research Part D: Transport and Environment, Vol. 77, 2019, pp. 132–147.

15.

Boisjoly

El-Geneidy

Measuring Performance: Accessibility Metrics in Metropolitan Regions around the World. Brookings, 2017.

16.

Committee of the Transport Access Manual. Transport Access Manual: A Guide for Measuring Connection between People and Places. University of Sydney, 2020.

17.

Guirao

Transport Demand Management. In International Encyclopedia of Transportation ( Vickerman

, ed.), Elsevier Ltd., UK, 2021, pp. 374–379.

18.

Currie

Lies, Damned Lies, AVs, Shared Mobility, and Urban Transit Futures. Journal of Public Transportation, Vol. 21, No. 1, 2018. https://doi.org/10.5038/2375-0901.21.1.3.

19.

Federal Highway Administration. Traffic Data Computation Method. FHWA, Washington, D.C., 2018.

20.

Erhardt

G. D.

Roy

Cooper

Sana

Chen

Castiglione

Do Transportation Network Companies Decrease or Increase Congestion?

Science Advances, Vol. 5, No. 5, 2019, p. eaau2670. https://doi.org/10.1126/sciadv.aau2670.

21.

Henao

Marshall

W. E.

The Impact of Ride-Hailing on Vehicle Miles Traveled. Transportation, Vol. 46, No. 6, 2019, pp. 2173–2194. https://doi.org/10.1007/s11116-018-9923-2.

22.

Lee

Kim

H. -J.

Park

G. -L.

Relocation Action Planning in Electric Vehicle Sharing Systems. Berlin, Heidelberg, 2012.

23.

Levinson

D. M.

Marshall

Axhausen

Elements of Access: Transport Planning for Engineers, Transport Engineering for Planners. Network Design Lab, 2017.

24.

Wang

Hunter

Bayen

A. M.

Schechtner

González

M. C.

Understanding Road Usage Patterns in Urban Areas. Scientific Reports, Vol. 2, No. 1, 2012, p. 1001. https://doi.org/10.1038/srep01001.

25.

Elefteriadou

(Ageliki). The Highway Capacity Manual 6th Edition: A Guide for Multimodal Mobility Analysis. ITE Journal, Vol. 86, No. 4, 2016.

26.

Mokhtarian

P. L.

Samaniego

F. J.

Shumway

R. H.

Willits

N. H.

Revisiting the Notion of Induced Traffic through a Matched-Pairs Study. Transportation, Vol. 29, No. 2, 2002, pp. 193–220. https://doi.org/10.1023/A:1014221024304.

27.

Cazzola

Crist

Good to Go? Assessing the Environmental Performance of New Mobility. International Transport Forum, Paris, 2020, p. 87.

28.

Ong

P. M.

Sciara

G. -C.

Pech

Cheng

Gonzalez

S. R.

Thomas

Strand

Schouten

Identifying, Evaluating, and Selecting Indicators and Data for Tracking Land Use and Transportation-Related Trends Related to SB 375 Goals. Publication CARB Agreement No. 15RD010. UCLA Center for Neighborhood Knowledge, 2018, p. 192.

29.

Cui

El-Geneidy

Accessibility, Equity, and Mode Share: A Comparative Analysis Across 11 Canadian Metropolitan Areas. Transport Findings, 2019. https://doi.org/10.32866/7400.

30.

Schäfer

A. W.

Yeh

A Holistic Analysis of Passenger Travel Energy and Greenhouse Gas Intensities. Nature Sustainability, Vol. 3, No. 6, 2020, pp. 459–462.

31.

Alarfaj

A. F.

Griffin

W. M.

Samaras

Decarbonizing US Passenger Vehicle Transport under Electrification and Automation Uncertainty has a Travel Budget. Environmental Research Letters, Vol. 15, No. 9, 2020, p. 0940c2.

32.

Handy

Planning for Accessibility: In Theory and in Practice. In Access to Destinations ( Levinson

D. M.

Krizek

K. J.

, eds.), Elsevier Publishers, pp. 131–147. http://nexus.umn.edu/Books/A2D.html.

33.

Fuel Economy Guide: Model Year 2020. U.S. Department of Energy, 2020.

34.

Mawhorter

Martin

Galante

Chapter 22 - California’s SB 375 and the Pursuit of Sustainable and Affordable Development. In Transportation, Land Use, and Environmental Planning ( Deakin

, ed.), Elsevier, pp. 497–521.

35.

Hou

Garikapati

Nag

Young

S. E.

Grushka

Novel and Practical Method to Quantify the Quality of Mobility: Mobility Energy Productivity Metric. Transportation Research Record: Journal of the Transportation Research Board, 2019. 2673: 141–152.

36.

Amin

Altinoz

Dogan

Analyzing the Determinants of Carbon Emissions from Transportation in European Countries: The Role of Renewable Energy and Urbanization. Clean Technologies and Environmental Policy, Vol. 22, No. 8, 2020, pp. 1725–1734.

37.

Couture

Busch

Hansen

Guerra

Murdock

H. E.

Ranalder

Adib

Andre

Corcoran

Corscadden

Renewables in Cities. 2019 Global Status Report. 2019.

38.

Romero-Lankao

Wilson

Sperling

Miller

Zimny-Schmitt

Sovacool

Gearhart

Muratori

Bazilian

Zünd

Young

Brown

Arent

Of Actors, Cities and Energy Systems: Advancing the Transformative Potential of Urban Electrification. Progress in Energy, Vol. 3, No. 3, 2021, p. 032002. https://doi.org/10.1088/2516-1083/abfa25.

39.

U.S. Department of Energy. SMART Mobility Capstone Reports. https://www.energy.gov/eere/vehicles/downloads/eems-smart-mobility-capstone-reports-and-webinar-series.

40.

U.S. Department of Energy. Energy Efficient Mobility Systems. Energy.gov. https://www.energy.gov/eere/vehicles/energy-efficient-mobility-systems. Accessed February 15, 2022.

41.

Adler

M. W.

Peer

Sinozic

Autonomous, Connected, Electric Shared Vehicles (ACES) and Public Finance: An Explorative Analysis. Transportation Research Interdisciplinary Perspectives, Vol. 2, 2019, p. 100038. https://doi.org/10.1016/j.trip.2019.100038.

42.

Axsen

Sovacool

B. K.

The Roles of Users in Electric, Shared and Automated Mobility Transitions. Transportation Research Part D: Transport and Environment, Vol. 71, 2019, pp. 1–21.

43.

Soteropoulos

Berger

Ciari

Impacts of Automated Vehicles on Travel Behaviour and Land Use: An International Review of Modelling Studies. Transport Reviews, Vol. 39, No. 1, 2019, pp. 29–49. https://doi.org/10.1080/01441647.2018.1523253.

44.

Boehmer

Broehm

Robb

Introducing the CDC/USDOT Transportation and Health Tool. Journal of Transport & Health, Vol. 3, No. 2, 2016, p. S50.

45.

How Do You Measure Mobility? U.S. Department of Energy, 2020.

46.

Federal Highway Administration. Health Performance Metrics. FHWA, 2020.