APPENDIX A. REVIEW OF LITERATURE ON SPEED CONCEPTS AND THE DESIGN PROCESS

The concept of speed and the many ways it can be expressed and measured, as well as the variety of factors that can influence it, have been the subject of much research in recent decades. An emphasis in NCHRP 15-76 was to review the literature related to these topics and summarize the findings most relevant to the concept of target speed and its influences.

SPEED CONCEPTS

The transportation profession uses a metric of speed in a myriad of ways, and a wide variety of speed-related terms are employed within the design process. The variety of speed terms used by the profession leads to numerous discussions about how to describe a particular roadway segment in terms of speed. Among those terms, the concept of target speed is a relatively recent one that generally lacks a formal definition; as such, it is beneficial to explore the ways in which the term is used and, at least for the purposes of this research project, establish a definition that reflects the term’s expected use and places it in the appropriate context with other, more established, speed terms.

Target Speed

At its root, target speed relies on the word “target” to convey its message, distinctive from other types of speed. There are many definitions of the word “target”, but the most applicable definition provided by Merriam-Webster (Merriam-Webster Dictionary 2000) is “a goal to be achieved”. Similarly, the Cambridge Dictionary (Cambridge Dictionary 2000) states that a target is “a level or situation that you intend to achieve”. Perhaps because of the common occurrence and understanding of the word, the use of the word “target” to define speed has not been associated with a formal definition in the transportation profession, though there are many examples of its use, particularly within the last decade.

As early as 2000, the term appeared in a paper by Poe and Mason (Poe and Mason 2000), in which they mention that, for low-speed urban streets, “Too often, the speeds on these facilities exceed the intended target speed of the roadway.” No other discussion of the target speed term is offered in the paper, but the implied meaning of the term is consistent with the dictionary definitions provided above.

In 2005, Brindle (Brindle 2005) described target speed in the context of two other terms, saying that target speed is “the desired street speed. The new speed profile should be below the target speed for the full length of the street.” Street speed was defined as the “the highest mean, 85th or any other percentile speed actually observed along the street (or street section); i.e., the maximum value of the speed profile.” The speed profile was defined as “the variation of mean, 85th or any other percentile free speeds along a street or street section.” The inclusion of the word “desired” in this definition also agrees with the concept of a goal to be achieved.

More recently, the use of target speed within the profession has increased in frequency and importance. In 2006, the Massachusetts Highway Department included target speed in their

Project Development & Design Guide (PDDG) (Massachusetts Highway Department 2006). In the PDDG, target speed has not only a definition (“the desired operating speed along a roadway”) but also related guidance on how it should be used in the project development process, stating that it is determined early in that process and that an appropriate target speed should consider:

• The context of the roadway, including area type, roadway type, and access control.
• The volume, mix, and safety of facility users.
• The anticipated driver characteristics and familiarity with the route.

The 2010 ITE Recommended Practice (RP) titled Designing Walkable Urban Thoroughfares: A Context Sensitive Approach discusses the use of target speed as a design control related to the selection of thoroughfare type and context zone for a particular roadway (ITE 2010). The RP defines target speed as “the highest speed at which vehicles should operate on a thoroughfare in a specific context, consistent with the level of multimodal activity generated by adjacent land uses to provide both mobility for motor vehicles and a safe environment for pedestrians and bicyclists” (ITE 2010). The authors of the RP stated that target speed replaced design speed in the report, which meant that target speed was the primary control for determining minimum intersection sight distance as well as horizontal and vertical curvature and the appropriate minimum sight distance associated with that curvature. The RP also stated that the target speed is designed to become the posted speed limit, and where speed limit is established through measuring actual speeds on the roadway, the design of the facility should “encourage the desired operating speed to ensure actual speeds will match the target speed” (ITE 2010).

Bassan (Bassan 2016, Bassan 2017) described target speed as “the desirable travel speed in the defined highway category. The goal is for most vehicles in the traffic stream to be able to travel at such a speed during free-flow conditions in a specific highway category.” This definition was in the context of design policy for highways in Israel, with the addition that “most vehicles” referred to approximately 90 percent of passenger cars and did not include heavy vehicles.

City staff in Palo Alto, California, used target speed to describe a desired outcome on several streets that were part of a report in support of a proposed 2017 speed limit ordinance. That report (City of Palo Alto 2017) stated that target speed is “intended to be a design speed that anticipates the introduction of roadway design elements that may reduce the Operating Speed, potentially enabling radar enforcement in the future.” The report specified that target speed was not based on operating speed, but instead it considered “factors such as residential density, bicycle safety, roadside conditions, adjacent land use, and potential conflicts with pedestrians, bicyclists, and residential or business districts” (City of Palo Alto 2017). The target speed was not intended for application citywide, but rather it was identified for “street segments where the surveys show Posted Speed Limits to be outside of the acceptable deviation from the Operating Speed to allow for radar enforcement.” In a draft resolution included in an attachment to the report, the city staff who authored the report defined target speed similarly to the ITE RP, replacing the word “vehicles” with “motorists” and replacing the word “thoroughfare” with “street”.

The 2018 AASHTO Green Book (AASHTO 2018) contains a definition of target speed that is also nearly identical to that found in the ITE RP; the Green Book definition differs in replacing “safe environment” with “desirable environment”, and it expands the road user groups to include public transit users with pedestrians and bicyclists. The Green Book also says that the target speed is intended to be used as the posted speed limit (AASHTO 2018).

The Guide for Setting Posted Speed Limits on Manitoba Roadways (Manitoba Infrastructure 2019) describes target speed in the context of an engineering speed limit study, in which a target speed should be identified that is “based on the road function and physical characteristics of the roadway”. In that discussion, target speed is defined as “the speed at which vehicles should operate on a facility in order to promote the safety of all users, while providing the appropriate level of mobility for motor vehicles.” The guide (Manitoba Infrastructure 2019) says the following factors can influence target speed:

• Pedestrian exposure.
• Cyclist exposure.
• Stakeholder input.
• Roadside environment.
• Roadside hazards.
• Average lane width.
• Presence of a median.
• Horizontal and vertical alignment.
• Roadway surface type.
• Number of intersections with public roads.
• Number of private access driveways.
• On-street parking.

The Manitoba guide (Manitoba Infrastructure 2019) also states that large differences between the operating speed and the target speed are usually “a result of a road where the risks are not apparent to the driver. The engineering study should identify potential issues causing the divergence and propose potential solutions.” The authors of the guide (Manitoba Infrastructure 2019) state that typical causes of differences in operating speeds and target speeds include the following:

• The road is being used for a different function than its original intention (e.g., a local road used by through traffic, resulting in higher speeds).
• The road has been over-designed, compared to its function and surrounding land use.
• The function of the road and the surrounding land use are inconsistent.
• The risks that are present along the road have been overstated in the target speed evaluation.

The Manitoba guide says one method to determine the target speed is to use the Transportation Association of Canada (TAC) Canadian Guidelines for Establishing Posted Speed Limits (Law and Zein 2009), which is described as an easy, consistent, and repeatable reference source. The TAC guidelines provide an evaluation tool to assess appropriate posted speed limits (which could be used as target speeds), based primarily on the classification, function, and physical characteristics of a roadway. The methodology in the TAC guide

considers risks associated with engineering factors, and it recommends lower posted speed limits for roadways with higher risk. Supporting text in the guidelines says that they were “developed through a review of current domestic and international practices, technical documentation, and testing”. TAC provides a standalone spreadsheet tool to help the practitioner evaluate posted speed limits based on the guide’s methodology. Elements considered in the spreadsheet are geometry, roadside, classification, land use, access and intersection density, and vulnerable road users (Law and Zein 2009).

The Florida DOT Design Manual (Florida DOT 2020b) reproduces the ITE RP (ITE 2010) definition for target speed, with reference to FHWA’s website on Context Sensitive Solutions and Design (FHWA 2017). The Florida manual also states that setting a target speed “is more useful in context classifications with a wide range of acceptable design speeds. Ideally, the Target Speed, Speed Limit, and Design Speed should all be the same where speeds are 45 mph or less.” The discussion in the Florida manual contrasts target speed with design speed and speed limit in the amount of time it takes to implement each one, stating that target speed can be set immediately, while design speed and speed limit often take time to change and may even need to be changed over the course of several projects. As such, target speed serves as the “target or goal” for changes in design speed and speed limit over time. The Florida manual says that when target speed is used, it should be established by consensus of the appropriate engineering staff, and the district planning office should include the target speed along with other documentation of the project’s context classification.

The Oregon DOT Blueprint for Urban Design (BUD) (Oregon DOT 2020a, Oregon DOT 2020b) provides design guidelines for roadways in urban environments and contains a focus on target speed as part of those guidelines. It describes target speed as the desired operating speed and presents target speed in the context of a relationship with design speed and posted speed limit. The BUD states that ODOT has clear policy guidance related to posted speed selection; however, ODOT may consider changes to its guidance to more effectively achieve desired operating or target speed. In practice, the BUD advises, the target speed and design speed should be the same, and a roadway should encourage an actual operating speed at the target speed.

The authors of NCHRP 855 (Stamatiadis et al. 2018) discuss the concept of target operating speed as a metric used to define the interaction between context and roadway for drivers, as part of their recommendations for an updated and expanded functional classification system (FCS). Their discussion includes the statement that the goal of the FCS is to develop a facility where the operating speed is close to the design speed, resulting in an environment with smaller speed differences among drivers. They added that these smaller speed differences will eliminate discrepancies between design speed and operating speeds, which would create a more uniform speed profile among drivers and could improve safety.

Based on the common uses of the term and its application within existing literature and policy, as well as the need to place the term in the context of the design process, the definition for target speed developed by the research team for this project is as follows: a speed selected for a roadway during the design process that guides planning, design, and operational decisions and results in a roadway with a driving environment that encourages drivers to operate at or below that selected speed.

Other Speed Terms

As the discussion in the previous section illustrates, target speed is often defined in the context of, or otherwise related to, other speed terms commonly used in the design process. Table A-1 lists a selection of common terms and suggested definitions. This section provides additional details on key speed terms related to target speed.

Table A-1. Definitions of selected speed terms.

Design Speed

The 2018 AASHTO Green Book (AASHTO 2018) defines design speed as “a selected speed used to determine the various geometric features of the roadway. The assumed design speed should be a logical one with respect to the topography, anticipated operating speed, the adjacent land use, and the functional classification of the highway.” The supporting text for this definition in the Green Book has varied throughout the document’s history, but the general idea of the purpose of design speed is that it provides a basis for the selection of design criteria and features that promote an overall design that encourages an appropriate speed for drivers to travel and is consistent with driver expectation for a roadway facility with particular characteristics. The current description of design speed discusses “a desired combination of safety, mobility, and efficiency within the constraints of environmental quality, economics, aesthetics, and social or political impacts.”

That desired combination can differ between high-speed and low-speed facilities. The Green Book (AASHTO 2018) provides the fundamental guidance that the selected design speed “should be consistent with the speeds that drivers are likely to travel on a given roadway”, but additional detail is provided for low-speed environments. The Green Book says that lower speeds “are desirable for thoroughfares in walkable, mixed-use urban areas and this desire for lower speeds should influence the selection of the design speed”, adding that it is for these types of designs that a target speed should be selected, consistent with the definition provided in the previous section. The Green Book also advises that, for lower-speed facilities, the typical design elements that are directly related to design speed (e.g., curve radius, superelevation, sight distance) are not necessarily the largest influences on the speeds that drivers choose to travel; instead, those influences are likely to be posted speed limits, access (i.e., vehicles turning at driveways and intersections), physical constraints (e.g., lane widths; presence of curbs; setbacks to pedestrian facilities, trees and poles and adjacent buildings), spacing and timing of traffic signals, and other traffic control devices.

The topic of “design speed versus target speed” typically focuses on low-speed urban and suburban roadways, especially where the 85^th percentile speed is higher than the posted speed limit. The Green Book (AASHTO 2018) states that the design of arterial streets in urban areas, along with the regulation of traffic control devices, should be intended to permit running speeds of 20 to 45 mph; it says that speeds in the lower portion of that range apply to local and collector streets through residential areas and arterials through denser business areas, while speeds in the higher portion of the range are more suitable for arterials in outlying suburban areas.

An illustration of the use of multiple speed terms in this context is the variation in description of design speed and target speed in some guidelines. For example, NACTO’s Urban Street Design Guide (NACTO 2020) advises a design process that uses target speed instead of operating speed to determine the design speed of roadway and roadside elements; however, the Global Designing Cities Initiative (Global Designing Cities Initiative 2020) states that design speed “is the target speed at which drivers are intended to travel on a street and not, as often misused, the maximum operating speed.” The NACTO approach describes design speed, target speed, and operating speed as separate concepts to be used within the design process, which would be more consistent with the AASHTO approach. To some extent, the conflation of design speed and target speed described in the Global Designing Cities Initiative definition seems to

refer to the changes in the AASHTO definition of design speed over time, but it also illustrates the potential interchangeability, and subsequent confusion, of terms that can be used by a variety of sources for different purposes.

A potential source of this confusion is in the application of design principles for lower-speed roads. As mentioned previously, the Green Book describes differences for low-speed environments, but the level of guidance and supporting research is not necessarily the same. A variety of research efforts on factors that affect design speed and its interaction with other speed terms have been conducted in recent decades, such as the study on horizontal alignment design by Krammes et al. (Krammes et al. 1994) and subsequent studies on similar topics, but many of them are focused on the rural or high-speed environment. As the Green Book acknowledges, drivers’ chosen speeds on urban roads are not necessarily sensitive to geometric design characteristics like curve radius as much as other elements such as lane and shoulder width or driveway density; thus, the application of the design speed concept to urban roads is not the same as on rural roads (Bassani and Sacchi 2012, Garrick and Wang 2005). Harwood et al. (Harwood et al. 2000) stated that, where a reduced design speed is necessary, practical steps must be taken in the design process to ensure that most drivers will be likely to travel at or below that design speed. They stated that, on facilities where lower design speeds are used, everything about that facility, including geometrics, traffic control, roadside design, and character of adjacent development, must suggest the appropriateness of a lower speed. Similarly, Tarris et al. (Tarris et al. 2000) emphasized the concept that, if a lower design speed is used, all design features of the street must be consistent with that lower speed; that is, low speeds can be achieved not with speed limit signing but only by creating streets whose geometrics, operational character, and development environment encourage motorists to voluntarily or necessarily use lower speeds. Bassan, in a discussion of speed-related terms and their use in Israel (Bassan 2017), stated that Israeli highway design policy considers the design speed and target speed to have identical design values on the urban street network where the design speed is no more than 60 km/h (37.3 mph) because the difference between operating speed and posted speed limit was typically less than 5 km/h (3.1 mph).

Operating Speed

Many references provide a definition of operating speed, but those definitions typically share common themes, if not identical language. Fitzpatrick et al. (Fitzpatrick et al. 1995) provided the following definition: the speed at which drivers are observed operating their vehicles during free-flow conditions. Many definitions also contain a supplemental statement similar to the one provided by Fitzpatrick et al. (Fitzpatrick et al. 1995): the 85th percentile of the distribution of observed speeds is the most frequently used descriptive statistic for the operating speed associated with a particular location or geometric feature. The AASHTO Green Book (AASHTO 2018), the MUTCD (FHWA 2023), NCHRP Report 966 (Fitzpatrick et al. 2021c), and FHWA’s Speed Concepts: Informational Guide (Donnell 2009), among others, have definitions that are similar to this.

Boodlal et al. (Boodlal et al. 2015) incorporated a similar term, running speed, in their definition, stating that running speed is “the speed at which an individual vehicle travels over a highway section.” They added that, during off-peak, low-volume conditions, operating speeds

and running speeds provide similar insights on how drivers select speeds based on the road geometrics, particularly on uninterrupted flow facilities.

In summarizing a history of the definition of operating speed, NCHRP Report 504 (Fitzpatrick et al. 2003) discusses phrasing used in previous definitions such as “highest overall speed” and “safe speed as determined by the design speed”; the report authors note that though the assumptions associated with those phrases may be valid for facilities designed at very high speeds, such as freeways, the appropriateness of those assumptions declines for roadways on the other end of the functional classification spectrum, such as local streets.

Posted Speed Limit

The NCHRP Report 966 (Fitzpatrick et al. 2021c) defines posted speed limit as the numeric speed limit value displayed on regulatory speed limit signs. FHWA’s Speed Concepts: Informational Guide (Donnell 2009) defines speed limit as the maximum lawful vehicle speed for a specific location. The FHWA Guide adds that there are two types of speed limits:

• Posted speed – The maximum lawful vehicle speed for a particular location as displayed on a regulatory sign. Posted speeds are displayed on regulatory signs in speed values that are multiples of 5 mph.
• Statutory speed – Numerical speed limits (e.g., 25 mph, 55 mph), established by state law that apply to various classes or categories of roads (e.g., rural expressways, residential streets, primary arterials, etc.) in the absence of posted speed limits.

The FHWA Guide (Donnell 2009) further states that posted speeds must be reasonable, on the basis that speed limits are only meaningful if most motorists comply voluntarily, which occurs only if a speed limit is considered to be reasonable. The MUTCD (FHWA 2023) mandates that posted speed limits shall only be based an engineering study made in accordance with established traffic engineering practices. The MUTCD adds that the engineering study “shall include an analysis of the current speed distribution of free-flowing vehicles”; beyond that, however, a universal process for conducting these studies does not exist, and practitioners have multiple sources available to aid in evaluating and identifying the appropriate posted speed limits. Previous research (Fitzpatrick et al. 2021b, Fitzpatrick et al. 2021c, Fitzpatrick et al. 1995, Fitzpatrick et al. 2003, Parker 1985, Parker 1992, ITE 2001) has shown that different agencies often consider different factors in their decisions to establish posted speed limits. The predominant factor typically reported in such research in the United States is the 85th percentile speed, consistent with the guidance in the MUTCD that posted speeds “should be within 5 mph of the 85th-percentile speed of free-flowing traffic.” Using the 85th percentile speed means that the driver often plays a key role in the speed limit-setting process, because the operating speeds used in establishing speed limits are typically measured when traffic is flowing freely. During free-flow conditions, research says, drivers select speeds that they believe optimize the tradeoffs between travel time and risk. Basing the speed limit on the 85th percentile indicates a belief that drivers are reasonable at assessing these tradeoffs, and their judgment is trustworthy in establishing a level where drivers who exceed that speed may be cited by law enforcement. While that may be so, additional conditions could exist that do not influence the 85th percentile speed but contribute to crashes (Donnell 2009).

Other reported factors that are considered in setting speed limits include roadway geometry, roadside development, crash history, pedestrian activity, and political pressure. In addition to the MUTCD, many states and cities have their own laws and criteria for setting speed limits, with some being more detailed than others. The MUTCD broadly serves as a reference regarding the setting of speed limits, while leaving other references and guidelines to provide more detailed criteria for selecting the posted speed limit.

NCHRP Report 966 (Fitzpatrick et al. 2021c) provides a procedure to calculate and identify a suggested posted speed limit for a street or highway segment. The procedure is based on the speed distribution for a segment of current drivers with adjustments for the consideration of safety. The procedure starts with identifying the roadway segment context and type. Next, the speed distribution of drivers on that segment is used to identify a potential suggested speed limit that is adjusted with consideration of the crash potential for the segment. Full details of the procedure are provided in the User Guide and the associated Speed Limit Setting (SLS) tool that were produced in NCHRP Project 17-76.

USE OF SPEED IN THE DESIGN PROCESS

Just as there are a variety of speed-related terms in common use, there are also multiple ways in which the concept of speed is used in the design process. In addition to the actual geometric design of the roadway, a comprehensive design process includes planning in advance of (and in conjunction with) the geometric design as well as the traffic engineering aspects related to the operation of the roadway after it is opened to traffic.

Planning

The concept of planning involves both the anticipated use of the roadway itself as well as land-use planning for the area surrounding that roadway. Both aspects of planning include a consideration of speed, though they are not necessarily applied in the same way.

Within land-use planning, the consideration of how land and traffic interact is often quantified or described by measures such as travel purpose, trip generation, and mode choice. In a discussion of the effects of urban form on travel, Crane (Crane 2000) provided a suggested list of travel outcome measures with urban form and land-use measures, summarized as follows:

• Travel outcome measures.
• Total miles traveled.
• Number of trips.
• Car ownership.
• Travel mode.
• Congestion.
• Commute length.
• Other commute measures (e.g., speed, time).
• Differences by purpose (e.g., business vs. personal travel, local vs. regional travel).
• Urban form and land use measures.
• Density.
• Extent of land use mixing.

• Traffic calming.
• Street and circulation pattern.
• Jobs/housing and/or land use balance.
• Pedestrian features.

In Crane’s list, speed has a direct application as a potential travel outcome measure and an indirect application within the consideration of traffic calming as a measure of urban form. The discussion associated with those measures describes a variety of ways in which those measures can be evaluated (e.g., simulation, observation, statistical analysis) and their interactions and relationships quantified. Crane posited that, although some relationships between land use and travel appear straightforward, such as that between density and trip length, closer examination suggests that these simple observed correlations are not so simple. Rather, Crane said, they represent the complex interactions of many factors. Land/travel linkages are both multidimensional and difficult to deconstruct, and little if any hard evidence indicates how the built environment can reliably manipulate travel behavior. That does not suggest that the land use has no effect on speed, but rather that the effect is difficult to quantify.

Dumbaugh and Rae (Dumbaugh and Rae 2009), in a continuation of that premise, explored relationships between traffic safety and community design in urban settings. While their focus was on factors that reduce crashes and crash severity, they did provide some insights that related speed to planning practice. Three specific implications for planning practice were described as follows:

• Manage the Mobility and Access Functions of Urban Arterials – Urban arterials are often designed and intended for vehicles operating at higher speeds; however, the presence of commercial and retail uses also requires those streets to accommodate lower-speed, access-related functions, resulting in the speed differentials and traffic conflicts that are associated with crashes. Dumbaugh and Rae stated that two strategies are available for addressing the problem of mixed traffic on urban arterials: reducing or eliminating the provision of access, and incorporating design features (on the roadway and/or the built environment) that encourage lower operating speeds. The former typically adds emphasis on mobility over access, while the latter emphasizes access over mobility. Both can have positive effects on crash reduction, but the appropriate approach depends on the type of environment that is desired on that roadway corridor.
• Orient Retail and Commercial Uses Toward Lower-Speed Thoroughfares – Safety may be further enhanced by positioning retail and commercial uses toward lower-speed thoroughfares. This is primarily accomplished, they concluded, through either designing roadways for lower operating speeds (e.g., 35 mph or less) or by providing low-speed internal access roads within developments adjacent to arterials with higher speeds.
• Plan Land Use, Speed Management, and Access Control at the Scale of the Network – While general strategies exist for managing speed on urban arterials, there is not a single solution that fits all situations. Rather, their recommended approach to improve safety is to implement community-level design solutions that account for effects of different land uses and street network configurations on vehicle speed. Addressing safety in low-speed environments requires streets and street networks that are

• designed to both encourage lower operating speeds and meaningfully accommodate local access and circulation needs for both vehicles and non-motorized users. In contrast, they said, safety on higher-speed, mobility-oriented thoroughfares is enhanced when access to these facilities is strictly managed to reduce the hazards posed by traffic conflicts and speed differentials.

The anticipated use of the roadway is related to that roadway’s functional classification. According to the AASHTO Green Book (AASHTO 2018), functional classification defines the role of each roadway and suggests its position within the transportation network, though it does not by itself explicitly address the fit of a particular roadway within a community or the needs of non-motorized transportation modes. The Green Book states that, for design purposes, the roadway type should be based on the actual role that the road plays in the transportation system, as defined through the project development process.

The functional classifications presented in the 2018 Green Book (AASHTO 2018) include freeways, arterials, collectors, and local roads and streets. These classifications are listed in approximate order corresponding to the Green Book’s six recognizable stages of travel: main movement, transition, distribution, collection, access, and termination. Each of those six stages is accommodated by a separate facility with design features corresponding to that function. The function also generally corresponds to traffic volume, with freeways at the high end of the volume range and hierarchy and local roads and streets at the low end. As previously discussed in the context of land-use planning, the two major considerations in classifying the function of street networks are access and mobility, specifically providing the appropriate balance between mobility for through traffic and access for the origins and destinations of trip ends. As illustrated in Green Book Figure 1-3, that balance favors mobility for freeways and high-volume arterials, and it emphasizes access for local roads and streets and low-volume collectors. Between the two extremes of freeways and local streets reside arterials and collectors, which have appropriate access-mobility balances that are more difficult to define. Arterials connect to freeways (via ramps) as well as collectors and even local streets, while collectors connect to both arterials and local streets. While access management techniques can be used on arterials to reduce the potential for conflicts between mobility for through traffic and access for adjacent land use, those conflicts cannot be avoided completely. Access may be more generally balanced with mobility on a collector, but the context of the surrounding development still helps to identify the likely demand and determine the appropriate treatments, both for motorized and nonmotorized travel. Those roadway and roadside characteristics provide the setting for determining the appropriate details for the actual design of the roadway.

Recent research by Stamatiadis et al. (Stamatiadis et al. 2018) has produced an expanded functional classification system compared to those found in previous editions of the Green Book and other guidelines. The major objective of the expanded FCS, according to its authors, is to provide enhanced information to designers to better inform the design decision process. This is accomplished by considering more detail of a roadway’s design context to enable understanding of the role the roadway plays within the community; identifying the role of the roadway within the local, city, and regional transportation network; and identifying the multiple roadway user groups and their priority within the design corridor. The expanded FCS contains five context categories and five roadway types, which are defined as follows (Stamatiadis et al. 2018):

• Context Categories
  • Rural: Areas with lowest density, few houses or structures (widely dispersed or no residential, commercial, and industrial uses), and usually large setbacks.
  • Rural Town: Areas with low density but diverse land uses with commercial main street character, potential for on-street parking and sidewalks, and small setbacks.
  • Suburban: Areas with medium density, mixed land uses within and among structures (including mixed-use town centers, commercial corridors, and residential areas), and varied setbacks.
  • Urban: Areas with high density, mixed land uses and prominent destinations, potential for some on-street parking and sidewalks, and mixed setbacks.
  • Urban Core: Areas with highest density, mixed land uses within and among predominantly high-rise structures, and small setbacks.
• Roadway Types
  • Interstates/Freeways/Expressways: Corridors of national importance connecting large centers of activity over long distances.
  • Principal Arterials: Corridors of regional importance connecting large centers of activity.
  • Minor Arterials: Corridors of regional or local importance connecting centers of activity.
  • Collectors: Roadways of lower local importance providing connections between arterials and local roads.
  • Locals: Roads with no regional or local importance for local circulation and access only.

The five context categories and the basis of the roadway classifications described by Stamatiadis et al. are used in the 2018 Green Book as the framework for roadway design. The Green Book framework includes freeways and it, in its basic form, shows a single arterial category, while later detailed guidance separates arterials into principal and minor subcategories. The guidance associated with the Green Book framework adds that the framework together with an assessment of multimodal needs and performance measures should guide the flexible approach to the design of projects.

Roadway Design

The 2018 Green Book (AASHTO 2018) states that formal roadway classification is merely a useful starting point for the design process, rather than a design control; functional classification, it says, carries with it expectations about a roadway design, including speed and capacity. Thus, for design purposes, Green Book guidance says the roadway type should be based on the actual role that the roadway plays in the transportation system. In describing the five contexts, the Green Book says that they are defined based on development density, land use, and building setbacks, and it refers to the Stamatiadis work in NCHRP Report 855 (Stamatiadis et al. 2018) for more guidance on applying those contexts.

The Green Book continues to say that the context classes work together with the functional classification system to help designers identify the appropriate balance among the transportation modes that use a specific facility. The Green Book authors state that it is intended

for a road’s context to be identified through a simple review of the type of development in a field visit or using an aerial photograph, rather than a quantitative analysis, suggesting that the contextual expectations of the road provide the largest cues to context. Quantitative data such as topography, population density, and building square footage can be used in determining context, but the designer does not need to rely on them. Guidance indicates that both the current context and possible future context of a roadway should be considered in design.

Most relevant to Project 15-76, the description of the Urban Context in the Green Book lists high-density development, mixed land uses, and prominent destinations, with sidewalks and on-street parking as common features along with single-story and low- to medium rise buildings for residential, commercial, and educational uses. The application of this guidance in conjunction with functional classification leads to the selection of design speed and consideration of other speed measures, as discussed earlier in this chapter. In the Urban Context, the Green Book says that driver speed expectations are generally lower than in suburban areas, while pedestrian and bicycle activity are generally higher.

For arterials in urban areas, the Green Book says the principal objective is mobility of all users for the context and the appropriate degree of service to local development; design speeds are broadly ranged from 25 to 45 mph (40 to 70 km/h), and other design considerations include volume, level of service, sight distance, horizontal and vertical alignment (with superelevation), and cross-slope. In contrast, while collectors in urban areas are described as having a design speed range of 30 to 40 mph (50 to 60 km/h), the purpose of the collector is shifted more toward access (equally divided between access and mobility), compared to arterials; while the other design considerations for a collector are similar to those of an arterial, the details of those design considerations are different (e.g., maximum grades on collectors are less than on arterials).

In its discussion on the use of the expanded FCS and the appropriate categories of target operating speed, NCHRP Report 855 (Stamatiadis et al. 2018) states that those low, medium, and high categories are intended to coincide with high and low design speed ranges in the AASHTO Green Book. The upper limit of the low category was chosen in consideration of the 20-mph survivability speed for non-motorized users in a crash with motor vehicles, along with the recent trend of some metropolitan areas adopting a maximum 25 mph speed limit within their jurisdictions. The lower limit for high speeds was based on the Green Book definition of high-speed roads, which are those with speeds of 50 mph and above. The report adds that the designer should examine the available speed range to select the operating speed most appropriate for all users given the facilities and context.

Related to that discussion is the recent consideration of the National Roadway Safety Strategy (US DOT 2022), which includes recommendations to revise relevant guidelines and regulations on setting context-appropriate speed limits and designing roadways that are inherently made to encourage drivers to follow those speed limits, choosing to slow down rather than relying on enforcement to manage speeding.

Traffic Engineering

Once the design is largely finalized, and even after the road has been constructed, traffic control devices facilitate efficient movement of vehicles and nonmotorized users and serve to

minimize conflicts. The Green Book (AASHTO 2018) states that traffic control should be used in conjunction with the context to improve motor-vehicle capacity and multimodal level of service, adding that efficient traffic control on the arterial system can help to relieve congestion on collectors and local streets. Two broad categories of traffic control are discussed here: intersection traffic control and corridor traffic control.

Corridor traffic control is concerned with facilitating the safe and efficient movement of traffic along a given roadway segment, such as an arterial. Traffic control devices such as signs, markings, and signals are placed on or adjacent to an arterial to regulate, warn, or guide traffic. Appropriate devices are designed to fulfill a specific traffic control need. The Green Book (AASHTO 2018) advises that the need for traffic control devices should be determined by an engineering study conducted in conjunction with the geometric design of the street or highway.

The Green Book (AASHTO 2018) also says that successful operation of an arterial in an urban area depends largely on proper pavement marking, especially on arterials having multiple lanes and particularly when special provision is made for left turns. Overhead lane signing can be very helpful. Signs enable drivers to plan their maneuvers, and to change lanes where needed, well in advance of an intersection or decision point. Advance signs are especially helpful under adverse weather conditions, such as rain or snow. Adequate pedestrian crossing treatments and effective speed management enhance pedestrian movements on arterials in urban areas (AASHTO 2018).

Intersection traffic control includes devices such as traffic signals and other treatments specific to facilitating safe and efficient traffic flow through the conflicting movements of an intersection. The Green Book guidance (AASHTO 2018) states that the ultimate goal of any intersection design should be to serve the traffic demands of all users at a level and quality of service that is consistent with the design of the roadway facility (e.g., arterial) and with as few crashes as practical; therefore, all intersection elements, including traffic signals, should be integrated into all aspects of the design process to accomplish this goal.

Uniform design and installation application of traffic control devices in the United States is provided in the MUTCD (FHWA 2023). The introduction to the MUTCD states that the purpose of traffic control devices is to “promote highway safety and efficiency by providing for the orderly movement of all road users on streets, highways, bikeways, and private roads open to public travel throughout the Nation.” That purpose is also stated to be the underlying principle for the use of such devices. In expanding on this basic premise, the MUTCD states that vehicle speed should be carefully considered as an element that governs the design, operation, placement, and location of various traffic control devices. Details such as sign size and marking width are provided in context of the speed of approaching drivers, with larger sizes associated with higher speeds so that drivers have sufficient time to see those devices at prevailing speeds. Recommendations for signing and marking in advance of intersections and crosswalks are also tied to speed, so that drivers can prepare to respond to those conditions as they approach. Thus, the speeds on a given roadway affect the traffic engineer’s decision on what treatments and strategies are appropriate for managing the traffic on that roadway.

FACTORS THAT AFFECT OPERATING SPEED

Overview

Multiple factors, including the posted speed limit, can influence a driver’s speed choice; however, the exact relationships may not be clear or may not been conclusively proven. Adding to the challenge of quantifying the relationship is the interaction between these factors and the overall visual scene for the driver. The contextual expectations of the road can communicate an appropriate speed to a driver, though the driver must be willing to accept that message in a contextual response.

A limited-access road (e.g., a freeway) with multiple lanes and wide roadside clear zone communicates the appropriateness of high operating speeds, while a residential street with on-street parking, multiple driveways, and the likelihood of pedestrian activity communicates the need for low speeds. The posted speed limit needs to be in agreement with the design of the road if desired operating speeds are to be achieved. When the design of the road—in terms of how it visually looks to a driver—results in implying that a higher operating speed is reasonable, engineering treatments may be needed to adjust the message being communicated to the driver.

Table A-2 provides a list of factors found in the research literature and in current guidance documents, along with their sources. The following sections describe those research findings in more detail. Appendix B provides additional information on current practice and guidance from city and state road agencies.

Table A-2. Factors that could influence setting or achieving target speed.

Multiple Roadway Environments

NCHRP Report 504 (Fitzpatrick et al. 2003) provides a detailed review of factors that influence speed on a variety of roadway environments.

• For urban streets in low-speed (less than 40 mph) environments, Poe and Mason (Poe and Mason 2000) found that three geometric variables (degree of curvature, lane width, and hazard rating) helped explain the variability in speed on horizontal curves.
• Rowan and Keese (Rowan and Keese 1962) concluded that substantial speed reductions occurred in urban environments when sight distance was less than 1,000 to 1,200 ft, and that the introduction of a curbed urban cross section and the adjacent land use (residential or commercial development) had an influence on speed reduction. Lateral restrictions (trees and shrubbery) were found to be a greater influence on speed reduction than development density.
• Oppenlander reviewed the literature to identify variables influencing spot speed on multiple roadway types (Oppenlander 1966). The roadway characteristics determined to be most significant included functional classification, curvature, gradient, length of grade, number of lanes, and surface type. Sight distance, lateral clearance, and frequency of intersections were also determined to be influential.
• Multiple other studies summarized in NCHRP Report 504 noted relationships between operating speed and the following elements for a variety of roadway types: degree of curve, curve radius, curve length, deflection angle, inferred speed, lane width, hazard rating, access density, posted speed limit, roadside, median presence, grade, sight distance, curbs, road surface, superelevation, land use, center line markings, and warning signs.

Within the research conducted for NCHRP Report 504 (Fitzpatrick et al. 2003), the authors determined that the strongest relationship was between operating speed and posted speed limit; the 85th percentile speed increases as posted speed increases. Posted speed limit was the only statistically significant variable at a 5 percent alpha level; however, a number of other variables were analyzed and found to have some influence that was not statistically significant, including median type, on-street parking, and pedestrian activity level. Initial (i.e., non-statistical) evaluations showed potential relationships between operating speed and the following variables:

• Functional class was influential, where local roads had the lowest speeds collected, while arterials had the highest.
• Access density showed a strong relationship with 85th percentile speed, with higher speeds being associated with lower access densities.
• The absence of either center line or edge line markings was associated with lower speeds.
• As distance between features that have influence on a driver’s speed, such as a signal or sharp horizontal curve, increased, speeds increased to a point and then plateaued.
• Roadways with shoulders that had widths of 6 ft or more had speeds above 50 mph, with one exception. Roadways with shoulders between 0 and 4 ft also had speeds up to 50 mph, but most speeds observed were lower than on the roadways with wider shoulders. Roadways with curbs had speeds across the entire range seen on roadways with shoulders (25 to almost 60 mph). There was no evidence that the presence of curbs resulted in lower speeds for a facility.
• Higher signal density was associated with lower speed.
• Wider pavement was associated with fewer lower speeds.
• Fewer lower speeds were associated with wider medians.

• No relationship was apparent between lane width and speed.
• Speeds on streets with single-family residential development tended to have lower speeds; however, some sites with residential development had speeds near 50 mph. A sizeable range of speeds occurred within each development type.

A Texas study (Fitzpatrick et al. 2001) studied influences on operating speed on suburban arterials with posted speed limits between 30 and 55 mph. Regression analysis examined the effects of selected variables on horizontal curves and tangent sections. When all variables were considered, posted speed limit was the most significant variable for both curves and straight sections, accounting for roughly half of the variability in the data. Other significant variables for curve sections were deflection angle and access density. In analyses that excluded posted speed limit, only lane width was a significant variable for tangent sections, while median presence and roadside development were significant for curve sections. The analysis that included posted speed limit, however, produced stronger relationships between speed and significant variables than did the analysis that excluded posted speed limit.

Additional studies have found posted speed limit to have a significant effect on free-flow speed on urban streets. A study (Ali et al. 2007) examined 35 four-lane urban street segments with posted speed limits between 35 and 45 mph. Along with posted speed limit, median width and segment length ratio (defined as the ratio of study segment length to the maximum signal spacing) were also significant. Figueroa and Tarko (Figueroa and Tarko 2004) included both rural and suburban roadways in their study. In addition to the posted speed limit, significant variables most commonly associated with suburban roads included intersection and driveway density, as well as presence of a two-way left-turn lane.

NCHRP Synthesis 535 (Sanders et al. 2019) reviewed a variety of speed-reduction treatments for potential effectiveness in improving pedestrian safety. Their primary categories of engineering treatments related to roadway design and traffic engineering were treatments with vertical deflection (e.g., speed humps and speed tables), treatments with horizontal deflection (e.g., modern roundabouts, mini traffic circles, and chicanes), lane and road reconfiguration (i.e., cross-section changes such as road diets, bulb-outs and curb extensions, median refuge islands, and lane narrowing), and signs and signals (e.g., speed-activated feedback signs, in-street pedestrian crossing signs, and signal timing). They found varying effects on speed reduction for each of these treatments, or combinations thereof. They concluded that the bulk of the countermeasure literature they reviewed covered treatments more appropriate for local roads or lower-speed collectors, but they were often not recommended or allowed (based on local policy) on higher-speed streets typically associated with the highest injury severity for pedestrians. For that context, they said, redesigning the street to communicate a lower speed, such as through cross-section changes, can effectively accomplish the goal of lowering speed.

The Canadian Guidelines for Establishing Posted Speed Limits (Law and Zein 2009) considers the following components to be influential in their speed limit evaluation process for all road classifications:

• Land use (urban or rural).
• Median (divided or undivided).
• Number of through lanes per travel direction (1 or 2+).

• Length of corridor.
• Horizontal geometry.
• Vertical geometry.
• Average lane width.
• Roadside hazards.
• Pedestrian exposure.
• Cyclist exposure.
• Pavement surface.
• Intersection density.
• Access density.
• Interchange density.
• On-street parking.

The TAC policy requires the consideration of each of the above items in determining the appropriate speed limit for a given roadway section.

Lower-Speed or Urban Environments

Aronsson (Aronsson 2006) reviewed speed, driver behavior, and site data to identify speed relationships for urban streets in Sweden. Both macroscopic (aggregated) and microscopic (simulation) speed data studies were part of the research. Results from the microscopic study showed that average number of crossing pedestrians and traffic flow had significant (R² = 0.91) impacts on average travel speed. Results from the macroscopic study showed that street function (i.e., functional classification) and number of lanes also had a high degree of explanation (R² = 0.72 for all street types). Separated bicycle lanes, presence of roadside parking, and number of minor intersections per kilometer were also significant for some of the street types in the macro study.

Bassani et al. (Bassani and Sacchi 2012) developed operating speed models for urban arterials and collectors in Italy. In their analysis of the available data, they found that posted speed limit, choice of travel lane (i.e., right, center, or left lane), lane width, median width, and carriageway width were all significant variables, with positive coefficients, in one or more models.

Dinh and Kubota (Dinh and Kubota 2012) analyzed speed and roadway data on urban residential streets in Japan with a posted speed limit of 30 km/h (18 mph). Their models contained the following significant variables:

• The length of a street section was positively associated with both 85th percentile speed and mean speed on tangent sections. The expected 85th percentile speed and the expected mean speed increased by 0.022 km/h (0.014 mph) and 0.015 km/h (0.009 mph), respectively, for every meter (3.28 ft) of section length.
• Drivers tended to select higher speeds with increased carriageway width (1.27 km/h [0.79 mph] for 85th percentile speed and 1.24 km/h [0.77 mph] for mean speed per 1-m [3.28-ft] increase in width).
• Compared to one-lane streets, the expected 85th percentile speed on two-lane streets was 1.44 km/h (0.89 mph) higher.

• The expected mean speed increased by 0.49 km/h (0.30 mph) for every 1-m (3.28-ft) increase in right safety strip width.
• The presence of sidewalks on both sides resulted in a higher 85th percentile tangent speed, by 1.20 km/h (0.75 mph), compared to no sidewalks or a sidewalk on only one side.
• A higher density of roadside objects was associated with a lower driving speed. The expected 85th percentile speed was approximately 0.37 km/h (0.23 mph) slower for every increase of one roadside object per 100 meters (328 ft), while the similar expected mean speed reduction was 0.35 km/h (0.22 mph) for every roadside object per 100 meters (328 ft).

An Atlanta, Georgia study (Dixon et al. 2008) developed operating speed models for low-speed urban streets, based on speed data collected from vehicles driven throughout the metropolitan area. Ultimately, the research team developed four road type models. For a tangent location, they developed speed models for two-way, two-lane roads and models for two-way, four-lane roads. Similar models were also developed for horizontal curve locations where the radius was 1700 ft or less. Key roadway characteristics in the models include the following:

• Roadside features (i.e., roadside rating and adjacent land use) more directly influenced the operating speed on two-lane roads than on four-lane roads.
• In general, raised curb, denser roadside and land use conditions, and higher driveway and intersection density contributed to reduced operating speeds.
• Larger curve radii and improved available sight distance contributed to higher operating speeds.

Thiessen et al. (Thiessen et al. 2017) developed three operating speed models with data from 249 tangent road segments in Edmonton, Alberta, Canada. The three models, reflecting arterials (A), collectors (C), and both arterials and collectors (A&C), used data from roadways with posted speed limits between 40 and 100 km/h (25 and 62 mph), though most collectors were posted at 50 km/h (31 mph). Key findings from the study included the following:

• All models indicated that operating speeds increased as road length increased. Specifically, the C model showed a 1-km/h increase in speed with every 100 to 110 m (328 to 361 ft) of additional length. the A model showed a 1-km/h (0.6 mph) increase for each additional 330 m (1083 ft).
• Higher operating speeds were expected in areas with higher speed limits. The A&C and A models also showed a positive association between speed limit and speed variability.
• Sidewalks were found to have statistically significant effects on speeds in the A&C model. In general, as sidewalks were moved farther away from the roadway, those roads showed increases in speed.
• All three models showed a negative correlation between access density and operating speeds.
• The A&C Model and the A Model showed reduced operating speeds as the number of objects and their nearness to the road increased. Both models had decreased speed as tree density increased and as distance from the road decreased.
• Increases in median width were associated with increases in operating speeds in the A&C and A models.

• Locations where a pedestrian crossing was present on collectors had lower operating speeds.

Multiple studies (Fitzpatrick et al. 2009, Trinkaus 1998, Ash 2006, Young and Dixon 2003, Lazic 2003) have investigated the relationship between operating speed and the presence of a school or a school zone speed limit. While drivers are commonly not in full compliance with the speed limit, speeds are typically lower during an active school zone.

Park et al. (Park et al. 2020) investigated relationships among roadway characteristics, speed, and crashes. In a path analysis of 649 sites with posted speed limits of 25 to 45 mph, they found that the absolute difference between posted speed limit (PSL) and average speed was positively related to KABC crashes (i.e., crashes with a severity level, no property-damage only crashes) at a statistically significant (α = 0.1) level. They also found that PSL, presence of bicycle lane, presence of school zone, and functional classification were significantly (α = 0.05) and positively associated with the absolute difference between PSL and average speed. The roadway surface width, natural log of daily volume, and natural log of segment length were also significantly (α = 0.05) but negatively associated with the absolute difference between PSL and average speed. Thus, each of these roadway variables is indirectly associated with KABC crash frequency. The authors stated that these effects were consistent with past research and intuition: larger spread/variability in operational speed (i.e., greater difference between PSL and average speed) is indicative of reduced smoothness in operations, with higher potential for speed differentials, which may, in turn, result in increased risk of crash occurrence. They also said that another possible explanation is that these associations could indicate that sites with more speed variability tend to be those with mixed visual cues or prone to ambiguous contextual situations (e.g., wide streets in a residential setting) that may result in different drivers choosing different speeds, and perhaps by doing so a larger proportion of the driving population could be more likely to exceed roadway conditions and thus increase their risk of crashing. While their work did not directly identify a causal relationship, it did provide evidence of association between these roadway variables and speed (and ultimately crashes).

The second explanation from Park et al. (Park et al. 2020) points to an entirely different set of potential influencing factors. Much of the previous discussion on influences on target speed focus on items within the designer’s control or decision-making ability; however, other characteristics of the roadway and roadside environment can play a role that may be largely undefined at present. Particularly for rural roadways, the horizontal and vertical alignment, paved surface width, and other design elements do play important roles in influencing operating speeds. In contrast, those elements may have a less pronounced effect and/or be present in conjunction with other influences on urban and suburban roadways.

NCHRP Report 966 (Fitzpatrick et al. 2021c) discusses this in its summary on target speed by saying that the central issue to achieving target speeds involves the configuration and operation of roadways so that target speeds, compatible with context and all roadway users, are chosen by—and not forced upon—vehicle operators. However, since much of the roadway context, particularly the urban context, has already been established, a large part of the effort of achieving target speeds involves retrofitting the existing environment. Since only lane width, cross-sections, roadside elements such as street furniture or trees, and vertical and horizontal

deflections are available to alter the roadway from a physical standpoint, the User Guide says, a clear understanding of what combination of those, and in what configurations, achieves target speeds is needed (or at least greatly influences operating speeds). The User Guide (Fitzpatrick et al. 2022) goes on to say that the simplest and most straightforward, proven method to achieve target speeds on major streets in urban areas is traffic signal progression, on the premise that if drivers realize they will have a stop-free, steady, but appropriate, speed to travel, then they may be more likely to actually drive the PSL. For low-speed urban roads and streets that are unsignalized, it says that transportation professionals will have to achieve target speeds through appropriate combinations of physical design features, many of which are now being included under context-sensitive or complete streets principles.

With that in mind, it is important to consider variables in addition to design elements that have potential to provide visual cues to drivers on what speeds are appropriate or comfortable. Those variables are often found in the roadside or in the surrounding environment. Poe and Mason (Poe and Mason 2000) stated that the low-speed environment has different objectives than high-speed environments in trying to provide access and accommodate multiple roadway users, such as bicyclists and pedestrians. The goal, they said, is to maintain lower speeds to achieve the functionality of the roadway and improve overall safety; however, they added, the speeds on these facilities too often exceed the intended target speed of the roadway.

For example, the study by Poe and Mason included five categories of variables (roadway, cross-section, roadside, land use, and traffic control devices) to investigate their relationships to speed on urban collector streets in central Pennsylvania. Roadside variables included hazard rating (a categorical variable on a scale of zero to four that measured the frequency and severity of roadside objects within 5 ft of the travel way), access, and sidewalk presence. In addition to roadway and cross-section variables (i.e., degree of curve, grade, and lane width), they found roadside hazard rating to be a significant variable for speeds on tangents and within horizontal curves; as hazard rating increased (i.e., as the frequency and severity of obstacles within 5 ft of the travel lane increased), speeds decreased.

Eluru et al. (Eluru et al. 2013) explored relationships between speed and street characteristics for local roads and arterials with posted speed limits of 40 km/h (25 mph) and 50 km/h (30 mph) in Montreal. They found that the number of lanes, number of sidewalks, presence of bicycle route, and good pavement condition were all positively associated with speed. The presence of a one-way street, a high vertical grade, and the number of sides of the street for which on-street parking was present were all negatively associated with speed.

A study by Naderi et al. (Naderi et al. 2008) found the presence of trees in a suburban landscape significantly dropped the cruising speed of drivers in a simulator study by an average of 3 mph; this finding was described in the context that trees are connected to a sense of safety and can provide a sense of spatial edge. Similarly, Calvi (Calvi 2015) described effects of tree spacing and offset on speed and lateral placement in a simulation of driving on a rural two-lane highway. Calvi stated that when trees were close to the road edge, drivers were found to decrease their speed significantly and move toward the center line of the road; increasing the offset of the trees resulted in higher speeds and a lower left lateral displacement. Tree spacing in Calvi’s

study did not affect speed, but when tree spacing was decreased, the distance from the road edge increased.

While not a direct analysis of speed, Tice (Tice 2021) studied factors that affect driver attention in complete streets environments, using eye-tracking data from the Strategic Highway Research Program 2 (SHRP2) Naturalistic Driving Study (NDS). One premise of the study was that driver attention affects the driver’s choice of speed, and a better understanding of why drivers pay attention could lead to design principles that account for these factors and a design process that incorporates those factors into the design. The SHRP2 data provided the basis of a measurement of midsegment percent of time on-task and multitasking behavior for 200 sites in Tampa, Florida, and Seattle, Washington. Tice (Tice 2021) conducted a statistical analysis on the attention data to determine effects of a wide range of context variables. Context features with a strong correlation to vulnerable user presence that supported driver’s visual recognition of that presence were also strongly correlated with driver attention. Features like pavement width, block length, doorway density, and sense of enclosure (e.g., tree canopy coverage and building-to-building width along the corridor) had the largest effects. Features that did not have a significant effect on the potential visual connection with street users (e.g., lane width, right of way width, on-street parking, functional classification) had no impact on driver attention or had weak effect sizes, despite strong correlations with vulnerable user presence. In terms of relation to speed, Tice concluded that the width of the corridor is an important design element, because narrower corridors coincide with both wider fields of view and lower speeds for the driver (Tice 2021).

A study by Wang et al. (Wang et al. 2006, Wang 2006) found that roadway variables such as number of lanes, density of intersections and driveways, on-street parking, and presence of curb affected speed. In addition to those roadway variables, the study also found that the presence of sidewalks, roadside density (i.e., the density of trees and utility poles [number/mile] divided by their average offset from the roadway), and adjacent land use were significantly related to speed. Sidewalks and roadside density had a negative association with speed, while non-commercial land use had a positive association.

Sidewalks have been examined in a number of studies such as the one by Wang to investigate effects of pedestrian facilities and activity, but the presence of sidewalks, sometimes defined by offset distance from curb, does not have a consistent trend in its effects on speed among those various studies. The Tice study indicated that sidewalk width could be a good predictor of the presence of vulnerable users but was not a significant predictor of driver attention on-task; the presence of a sidewalk on a wider corridor affected the driver’s attention less than on a narrow corridor. Similarly, some studies show sidewalks have a positive effect on speed while others show a negative effect. This raises the question of what extent sidewalks have a non-intuitive effect that is difficult to quantify when studying just the sidewalk itself. For example, sidewalks in some locales may be more commonly found on arterials, which have higher speeds, compared to collectors and local streets, which have lower speeds. In that case, the sidewalk may not be a cause of the increased speed, but its presence could be correlated to that speed. For that reason, it is beneficial to consider additional factors related to pedestrian features besides the mere presence of a sidewalk.

A variety of elements contribute to the provision of pedestrian features and the presence of pedestrians, which can be considered under the concept of walkability. Walkability can be represented in a number of ways, but one definition suggested by Dovey and Pafka (Dovey and Pafka 2020) describes a set of capacities in a given neighborhood that are in three primary categories: the densities or concentrations of buildings and people, the mix of different functions and attractions, and the access networks used to navigate between them. Density shortens the distances between people and the places they desire to go, mix describes the variety of functions and purposes available in the area, and access is about the options for making those connections between two locations (and to what degree those options include or exclude walking).

The Designing Walkable Urban Thoroughfares Recommended Practice from ITE (ITE 2010) defines walkable as “Streets and places designed or reconstructed to provide safe and comfortable facilities for pedestrians, and are safe and easy to cross for people of all ages and abilities.” Similar to Dovey, it describes the following concepts in describing principles for walkable communities:

• Accommodating pedestrians, bicycles, transit, freight, and motor vehicles within a fine-grained urban circulation network where the allocation of right of way on individual thoroughfares is based on urban context.
• Providing a compact and mixed-use environment of urban buildings, public spaces and landscapes that support walking directly through the built environment and indirectly by supporting human and economic activities associated with adjacent and surrounding land uses.
• Achieving system-wide transportation capacity by using a high level of multimodal network connectivity, serving walkable communities with appropriately spaced and properly sized pedestrian, bicycle, transit, and vehicular components.
• Creating a supportive relationship between thoroughfare and context by designing thoroughfares that will change as the surroundings vary in urban character.

There are multiple ways to measure walkability, depending on the desired context. NHTSA provides a Walkability Checklist (NHTSA no date) that uses the participant’s subjective ratings of a neighborhood’s sidewalks, crossings, drivers, ease of following safety rules for adults and children, and pleasant surroundings. The checklist contains a list of potential action items a person can take to improve a neighborhood’s score based on those ratings.