U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
| REPORT |
| This report is an archived publication and may contain dated technical, contact, and link information |
|
 |
| Publication Number: FHWA-HRT-13-047 Date: August 2013 |
Publication Number: FHWA-HRT-13-047 Date: August 2013 |
PDF files can be viewed with the Acrobat® Reader®
A complex interchange is defined as a facility that contains many lanes (i.e., four or more in each direction) and carries high traffic volumes through a maze of tightly spaced ramps and connectors.(10) Additionally, drivers often have to make multiple lane changes requiring intense attention and rapid decisionmaking. As transportation agencies struggle with adding freeway lane capacity in times of limited resources and shrinking right-of-way, new interchange designs are being built beyond the traditional diamond or cloverleaf configurations that are increasing the frequency of drivers having to navigate through a complex interchange. The MUTCD offers limited advice for signing at non-traditional interchanges; however, there are still improvements that could be made to current signing and marking practices that could further affect driver decisionmaking, comfort, and safety.(1)
This chapter identifies the current state-of-the-practice regarding signing standards and summarizes research that has been conducted regarding signing at complex interchanges.
To effectively design and place freeway guide signs at a complex interchange, a practitioner must be cognizant of the limitations of drivers while navigating through an interchange area. Information should be presented in a clear, concise, and consistent manner to help ensure that motorists unfamiliar with the route can easily interpret the information presented. Repetition of messages is also encouraged.
This section provides a brief review of issues related to driver information needs and processing as related to freeway guide signing. It also provides a brief description of the various components of the driving task. Research on legibility and information processing is also briefly discussed. Readers are advised to consult the relevant source material for a more detailed treatment of these topics. Finally, a methodology for evaluating the adequacy of guide signing is reviewed.
The concept of positive guidance is often used as a guiding principle for providing information to drivers. Positive guidance consists of creating and maintaining a driving environment that has the following characteristics:(11)
If the principles of positive guidance are applied consistently, drivers will subconsciously develop expectations about where to seek information. It is important to understand the demands that are placed on the driver during the driving task when the concepts of positive guidance are applied. The driving task is made up of a number of subtasks that require varying levels of time and cognitive activity. The three most basic subtasks are as follows:(12)
Performance of these subtasks allows drivers to maintain control of their vehicles, maintain their positions in the lane, and navigate to their final destinations. Drivers perform these subtasks continuously at various cognitive levels, although the amount of attention and cognitive resources allocated to each task may vary depending on the specific conditions that are present at a given point and time. A detailed description of each of these subtasks is given in the following subsections. The features of the positive guidance concepts were summarized into tables and are reproduced.(13)
The control subtask consists primarily of steering control and speed control.(12) Steering control involves maintaining the orientation of the vehicle with respect to the roadway, and it usually has the highest priority to the driver. Speed control involves using the brake and accelerator to select an appropriate speed for a given situation. Table 2 summarizes the basic characteristics of these two components of the control subtask.
| Characteristic | Steering Control | Speed Control |
|---|---|---|
| Priority | High | High |
| Driver level of effort | Varies depending on geometrics | Varies depending on geometrics and traffic |
| Information needs | Vehicle response characteristics relative position of vehicle | Vehicle braking and acceleration characteristics and road conditions ahead of driver |
| Demand on driver | Usually low because subtask is over learned | Greater than steering since driver must look farther down the road |
The guidance subtask involves maintaining a safe and efficient path relative to all factors in the roadway environment.(12) Some examples of actions included in the guidance subtask are car following, passing, and responding to traffic control devices. Table 3 summarizes the characteristics of the guidance subtask. The ability to perform the guidance subtask is a function of the driver's previous knowledge of similar conditions. Once a particular condition has been observed by drivers, they must process the information to determine an appropriate course of action. The level of cognitive demand that this places on drivers is dependent on their previous experiences in a given situation.
| Characteristic | Guidance |
|---|---|
| Priority | Varies depending on conditions but usually intermediate between control and navigation |
| Driver level of effort | Higher than control subtask with more conscious decisionmaking necessary |
| Information needs | Traffic conditions, road geometry, weather conditions, and other information that impacts the road environment |
| Demand on driver | Varies depending on the driver's previous experiences and prior knowledge |
The portion of the driving task that is most directly affected by freeway guide signing is the navigation subtask.(12) The navigation subtask consists of planning a trip from the beginning to the end and then executing the trip plan. The navigation subtask can be broken into two areas: (1) trip preparation and planning and (2) direction finding. Trip preparation and planning can consist of anything from drivers using their own mental map of an area to consulting maps or knowledgeable persons in order to plan a trip. If drivers are well prepared prior to beginning a trip, they will be more successful in the navigation subtask even if there is limited en route information. Direction finding occurs while drivers are en route and attempting to reach their destinations. This portion of the subtask involves interpreting direction guidance on signs to receive information about the appropriate path. The characteristics of the navigation subtask are summarized in table 4.
| Characteristic | Trip Preparation and Planning | Direction Finding |
|---|---|---|
| Priority | Performed pretrip, so no demands on driver while en route | Usually lowest of all subtasks, although demands may increase in complex or unfamiliar situations |
| Driver level of effort | Varies depending on driver familiarity with route | Usually low |
| Information needs | Location or origin and destination and physical or mental map of alternative routes | Guide signs, route markers, street name signs, landmarks, etc. |
| Demand on driver | Usually low | Usually low except in unusual circumstances |
Attention is an important component of the driving task.(12) When a subtask has a low demand, it can be performed with little conscious attention, allowing drivers to allocate attention to tasks that require more cognitive resources. When the demands of the driving task require that more attention be placed on a particular subtask, it comes at the expense of performing tasks requiring a higher level of attention. This process is known as load shedding. For example, a driver on an uncongested freeway can easily perform navigational subtasks. If traffic becomes extremely congested, the navigational subtasks become more difficult to perform because the driver must allocate more attention to the control and guidance subtasks.
Expectancy is also very important in the driving task.(12) Drivers need to have reasonable expectations about how their vehicles will perform, the geometry of the road downstream of their positions, and where to find navigational information. If the expectancy of the driver is violated, the performance of the driving task may suffer. This situation is particularly important in freeway guide signing where an unfamiliar driver relies on guide signs to provide information to perform the navigation subtask.
In a freeway environment, drivers must read, interpret, and react to freeway guide sign messages in a limited amount of time in order to obtain information for the navigation subtask. If a sign presents too much information, there is a possibility that a driver will not comprehend important navigational information. The following subsections present studies that have attempted to determine both the amount of time required to read a guide sign and the maximum amount of information that should be displayed on a guide sign.
Figure 1. Equation. Relationship between sign reading time and number of words.
Where:
N = Number of familiar words on the sign.
T = Reading time (s).
This relationship yields an average reading time of 333 ms per word. Researchers then modified this formula to incorporate a safety factor in case the driver was distracted while attempting to read the sign. The researchers arbitrarily determined that a safety factor of 2 should be provided. This safety factor was not the result of any research. The revised formula is shown in figure 2.(14) The modified equation yields a reading time of 667 ms per word.
Figure 2. Equation. Revised formula for relationship between sign reading time and number of words.
Issues related to sign placement, sign content, traffic conditions, and driver familiarity with the message can significantly alter the amount of time required to read a sign. One study examined drivers' eye fixations while driving on an interstate highway and found that drivers did not continually read signs.(15) Instead, drivers made a series of discrete fixations on signs that lasted between 100 and 600 ms. As drivers became more familiar with a sign, they spent less time reading the sign to obtain information.
A British study attempted to evaluate the impact of information overload on the time required for drivers to respond to guide signs.(16) The researchers evaluated guide signs on normal surface streets and found that the relationship between response time and the number of destinations was non-linear. Although the search times were greater when more destinations were present, there was no evidence that drivers' search abilities broke down when more destinations were present.
A 1989 study attempted to determine the time necessary to read signs while subjects were performing demanding driving tasks.(17) Non-freeway guide signs were used in this evaluation. The number of destinations on the signs ranged from four to nine, with a maximum of three destinations in each cardinal direction. They found that reading times varied from 0.88 s for signs with four names to 1.33 s for signs with nine names. When participants were asked to find destinations that were not on the sign, the reading times increased from 1.42 to 2.24 s. While reading times increased as the number of words increased, it was not always a substantial increase.
McNees and Messer conducted a study that examined the ability of drivers to successfully read and interpret freeway guide signs within limited time constraints.(18) Drivers were presented with two to five individual sign panels on a simulated overhead sign structure. Each sign panel contained 2 to 10 units of information per panel. Each place or street name, route number, cardinal direction, command, distance, or lane use arrow was counted as a separate unit of information. Subjects were asked to identify the proper travel lane that should be used to reach a predetermined destination. Time constraints were applied in order to simulate the impact of heavy driver task loads under freeway speeds. Signs were displayed for 2.5, 4.0, and 6.0 s in order to reflect unacceptable, acceptable, and desirable amounts of available reading time.
As expected, it took participants longer to read signs that had more information on them. When the exposure time was limited and a lot of information was presented, the participants had a lower accuracy for message interpretation.
Table 5 summarizes the accuracy results of the test subjects for different information loads and display times. The researchers recommended an optimum value of six units of information per sign. The average correct response rate varied as a function of the total number of units of information presented on each sign.
| Information per Panel (units) | Display Rate (s) | Percent of Drivers with Correct Response | |||
|---|---|---|---|---|---|
| 2 Panels | 3 Panels | 4 Panels | 5 Panels | ||
| 2 | 6 | 93 | 83 | 94 | 80 |
| 4 | 88 | 83 | 93 | 80 | |
| 2.5 | 83 | 82 | 80 | 84 | |
| 4 | 6 | 96 | 82 | 63 | 46 |
| 4 | 93 | 83 | 69 | 52 | |
| 2.5 | 91 | 94 | 92 | 76 | |
| 6 | 6 | 100 | 92 | 33 | 95 |
| 4 | 87 | 92 | 36 | 86 | |
| 2.5 | 99 | 92 | 52 | 92 | |
| 8 | 6 | 81 | 55 | 76 | 58 |
| 4 | 93 | 80 | 82 | 61 | |
| 2.5 | 75 | 45 | 91 | 36 | |
| 10 | 6 | 77 | 60 | 65 | 70 |
| 4 | 83 | 71 | 75 | 73 | |
| 2.5 | 57 | 80 | 78 | 79 | |
McNees and Messer used the reading time data collected in the study to generate a table of desirable and minimum reading times that should be provided to drivers for overhead guide signs.(18) The elapsed time between when a sign was initially displayed and when the subject made a correct lane choice was recorded. The researchers then developed a series of regression lines that they used to predict desirable reading times that should be provided for varying levels of information.
Table 6 summarizes these results. Desirable reading times represent a predicted reading time where at least 85 percent of the drivers would make the correct lane choice decision. The minimum reading time represents a time where 75 percent of the drivers would make the proper decision. Cells with a dash indicate situations that should not be used on the road. These situations represent cases where more than 20 units of information are presented on the sign structure.
Researchers also generated a table of desirable and maximum amounts of information that should be placed on overhead sign structures (see table 7). The table shows that placing five sign panels on a single structure is not a desirable design and should not be used if possible. The maximum amount of information on any sign structure should not exceed 20 units.
| Units of Information per Panel | Condition | Reading Time (seconds) | |||
|---|---|---|---|---|---|
| 2 Panels | 3 Panels | 4 Panels | 5 Panels | ||
| 2 | Desirable | 3.1 | 3.5 | 3.9 | 4.4 |
| Minimum | 2.7 | 2.7 | 3.0 | 3.3 | |
| 4 | Desirable | 3.6 | 4.2 | 5.0 | 5.7 |
| Minimum | 2.7 | 3.2 | 3.7 | 4.2 | |
| 6 | Desirable | 3.8 | 4.5 | — | — |
| Minimum | 2.8 | 3.4 | — | — | |
| 8 | Desirable | 3.9 | — | — | — |
| Minimum | 2.9 | — | — | — | |
| 10 | Desirable | 4.0 | — | — | — |
| Minimum | 3.0 | — | — | — | |
— Indicates situations that should not be used on the road.
| Number of Panels | Condition | Maximum Units of Information per Structure |
|---|---|---|
| 2 | Desirable | 12 |
| Maximum | 16 | |
| 3 | Desirable | 18 |
| Maximum | 20 | |
| 4 | Desirable | 16 |
| Maximum | 20 | |
| 5 | Desirable | — |
| Maximum | 20 |
— Indicates an undesirable design.
Before interpreting the message on a sign, drivers must be able to determine that a sign is present, and they must be able to read the message. The initial detection of a guide sign occurs when a driver can see the sign without being able to read it. The ability of a driver to detect a guide sign is usually a function of the size of the sign and the contrast between the sign and the surrounding background. At night, the luminance of the sign also impacts the ability of a driver to detect it. This section provides a brief overview of some of the factors that can influence the legibility of a sign.
Drivers must be able to read the sign before they can comprehend its message. The designer has the ability to alter a variety of factors that can influence sign legibility including font, letter height, and type of sign illumination or retroreflectivity.
Some of the earliest research on legibility was performed in the 1930s by Forbes and Holmes.(19) The researchers evaluated the day and night legibility of signs using highway series B and series D letters. Letter heights that were evaluated ranged between 6 and 24 inches. Approximately 400 young adult observers drawn from a college student population were used to view the signs. The observers approached the sign, and researchers made note of the distance at which the observers could read the message on the sign. All signs were ground mounted. Nighttime legibility studies were performed using reflectorized floodlighted letters for the series B signs and both floodlighted and reflectorized letters for the series D signs.
The researchers examined the 80th percentile legibility distances for the signs. The daytime legibility was consistently better than the nighttime legibility for both letter series; the nighttime legibility was between 8 and 20 percent lower than the daytime legibility. They found that a person with a visual acuity of 20/40 had a legibility of 33 ft per inch of letter height for B series letters and 50 ft per inch of letter height for D series letters. Since the subjects tended to be young, their eyesight was fairly acute, with a median value of 20/20.
In 1958, Allen evaluated the daytime and nighttime legibility of highway series E letters (typically used on freeway guide signs in the 1950s) using 48 participants.(20) The researchers tested four different levels of external illumination as well as button copy and retroreflective sheeting. Both ground-mounted and overhead guide signs were evaluated. Subjects approached the sign while traveling in a vehicle at 15 mi/h, and the distance at which the subjects could read the sign was noted. The average age of the subjects was 33. The average visual acuity of the subjects was 20/18, so they tended to have good vision.
Letter heights between 8 and 18 inches were evaluated. This study showed that the average daytime legibility of the message was about 88 ft per inch of letter height. When the sign was externally illuminated, the legibility declined by about 15 percent. There were several possible reasons for the differences between Allen's and Forbes and Holmes' studies. Allen's study used four-letter words that were familiar to drivers. Forbes and Holmes' study used six-letter words with deliberate misspellings in order to ensure that subjects read the entire word. By using familiar words, the legibility distances in Allen's study may have been increased. The visual acuity of the test subjects was also slightly better for Allen's study than for Forbes and Holmes' study.
There is some concern that the E(modified) font used on freeway guide signs is not suitable for use with prismatic materials due to its wider stroke width compared to other fonts. The E(modified) font may be susceptible to irradiation, where the letter stroke is so bright that it may bleed into open spaces in the letter. This blurring can reduce the legibility of the letters. The Clearview font was developed to mitigate some of these concerns by creating wider open spaces with the letters.
A study performed at Texas A&M University compared the daytime legibility of the E(modified), Clearview, and British Transport fonts and found no significant differences between the daytime legibility of these fonts.(21) The author recommended that the type of font used on a guide sign be determined based on nighttime legibility concerns. A recent study by TTI evaluated the difference in nighttime legibility between Clearview and E(modified) when a prismatic sheeting was used.(22) This study found that nighttime legibility distance increased by approximately 70 ft (10 percent) over E(modified) when the Clearview font was used with the prismatic sheeting.
Another study examined the relative effectiveness of the Clearview font across several material types.(23) The study found that the Clearview font did not create significantly better recognition distances than the E(modified) font, although it did perform better than a series D font during the day. At night, the Clearview font did appear to improve recognition and legibility distances over the E(modified) font. When the Clearview font was increased to 112 percent of its normal size, the legibility distances were approximately 50 ft greater than the E(modified) font.
Gordon examined the legibility of cardinal direction words using all capitals and mixed-case lettering.(24) The mixed-case font used initial capital letters that were of the same size as the lowercase letters. The researchers hypothesized that emphasizing the initial letter of the cardinal direction would improve legibility. The results indicated that the cardinal directions could be identified from 10 percent farther away when the mixed-case font was used.
In the early 1980s, Messer and McNees developed a level of service indicator for analyzing freeway guide signs.(25) The purpose of this indicator was to provide an objective means to determine if a guide sign was adequate for a driver with 20/40 vision. Their rating scheme was based on an assessment of three factors: navigation, workload, and response. The assessment of these factors was then used as input into an overall comprehensive level of service for the sign. This section briefly discusses this level of service concept.
The first component of the level of service concept was an assessment of the ability of a sign to guide motorists unfamiliar with an area to their destinations. Four factors were used to perform this assessment: sufficiency, consistency, expectancy, and relatability. Each one of these components was subjectively scored as good, fair, or poor, and the resulting rating was converted into a level of service. This portion of the assessment determined whether drivers received enough information to make an informed decision about their paths of travel.
Messer and McNees also assessed the workload that the sign placed on the driver.(25) They defined the workload of the sign as the ratio of the time required to process the information on a sign to the time available for this to occur. Workload ratings were then converted into a level of service for the sign. This level of service assessed whether a driver could read the information in the required amount of time.
Driver ability to react to the information on the sign within the space permitted was also examined as part of the level of service assessment. The total travel distance needed to respond to a sign was calculated and then divided by the physical distance available at the site to perform these actions. This measure indicated whether enough time was provided for a driver to react to the message.
Messer and McNees developed a tool to objectively assess the workload and response criteria, but the navigation criteria remained relatively subjective.(25) Their method permits comparisons between signing alternatives, but it may be too cumbersome to use in practice. The user must perform a variety of calculations to determine the amount of time or space available to see and react to a sign, and the conditions used to determine these values often represent idealized situations. This methodology is a step in the right direction but may have limited value to the practitioner.
Although this tool cannot illustrate what sign sequence should be used for every possible interchange geometry, its guidance and standards provide a starting point for the design and placement of the appropriate sequences for complex interchanges.
Lerner et al. developed a model and accompanying analysis software that predicted driver workload as a function of sign density, units of information per sign, and some limited roadway geometric data.(26) This work is promising, but only simple roadway geometries can be input into the modeling tool. The workload predictions for these simple geometries demonstrate the high impact of sign density on driver workload.
The 2009 MUTCD contains the basic principles that govern the design and use of traffic control devices for all roadways open to the public.(1) The fundamentals of the MUTCD state that signing on freeways and expressways should serve the following purposes:
However, complex interchanges occur where multiple roadways intersect and where there are an increased number of exits to surface streets, thereby creating an increased amount of information that must be provided to the driver. At this point, providing all of the necessary information to fulfill a driver's signing needs without overloading the driver's abilities becomes difficult.
Due to the variable nature of interchanges that can be encountered by a designer, the MUTCD provides an option to State and local agencies in developing word messages not provided in the manual in situations where roadway conditions make it necessary to provide drivers with additional guidance information. These new word messages may be used without experimentation and may be beneficial in the development of complex interchange signing.
The manual also provides basic guidance to the practitioner on where to locate signs. This is of particular importance at complex interchanges, as the distribution of information along the path can have a significant impact on driver understanding. Section 2A.17 of the MUTCD states, "Overhead signs should be used on freeways and expressways, at locations where some degree of lane use control is desirable, and at locations where space is not available at the roadside."(pg. 41)(1) The section then lists conditions where overhead signs may be beneficial, including complex interchange design, as well as other characteristics that this project has categorized to contribute to an interchange being complex. Later, the manual adds further direction, stating that if overhead signs are warranted, "The number of signs at these locations should be limited to only those essential in communicating pertinent destination information to the road user."(pg. 183)(1) However, these simple directions do not necessarily answer the question of how the information for a complex interchange should fit into this format.
Beyond this simple advice to use overhead signs, the manual further addresses sign locations in section 2A.16, stating that "Signs requiring separate decisions by the road user shall be spaced sufficiently far apart for the appropriate decisions to be made."(pg. 37)(1) Additionally, the manual suggests the concept of sign spreading when major overhead signs are spaced so that they are not all placed at a single location, possibly overloading the driver. Guidance states that "sign spreading should be used at all single exit interchanges and to the extent possible at multi-exit interchanges."(pg. 183)(1) Unfortunately, when addressing complex interchanges, this guidance can be hard to comply with, as the number of decisions points within a small travel space can be very high.
Other portions of the signing design that can be applied to complex interchanges include the following:
Additionally, section 2E.07 of the MUTCD gives the following list of special sign treatments that may be desirable due to specific operating conditions or road geometrics on urban freeways and expressways, and therefore apply to the complex interchanges this project is studying:(1)
For complex interchanges with multiple major and intermediate interchanges, it can become difficult to place two to three advanced signs without violating manual guidance previously mentioned about the amount of information on a sign as well as the number of signs at a single location.
There are two unique formats that can be applied to guide signs addressing complex interchanges that have optional lanes and multilane exits: arrow-per-lane signs or diagrammatic signs. Each of these signs is addressed within the MUTCD; however, there are no definitive correct applications as to what conditions would make one or the other of these designs appropriate. The following sections discuss the makeup of each of these types of sign formats.
An arrow-per-lane guide sign uses upward-pointing arrows above each lane to convey the direction of travel of that lane at the split or exit (see figure 3). Research has shown that these types of signs can be beneficial for splits and multilane exits with an option lane because the option lane can otherwise be difficult to interpret by the driver.
Figure 3. Illustration. Arrow-per-lane guide sign for a multilane exit.(1)
Section 2E.21 of the MUTCD states the following:
Where used, the Overhead Arrow-per-Lane guide sign at the exit or split shall be located at or in the immediate vicinity of the point where the existing lanes begin to diverge from the through lanes or, for a split, at the point where the approach lanes begin to diverge from one another, preserving the relation of the arrows displayed on the sign to their respective lanes. Overhead Arrow-per-Lane guide sign at the exit shall not be located at or near the theoretical gore.(pg. 193)(1)
The section goes on to provide the standard sign details, focusing on the look, orientation, and placement of the arrows.
The arrow for an option exit lane that also carries the through route shall have a single shaft that bifurcates into a vertically upward-pointing arrow and a curving arrow corresponding to the configuration of the through and exit lanes. For splits with an option lane, the arrow for the lane from which either direction of the split can be accessed shall have a single shaft that bifurcates into two upward-pointing curving arrows showing the approximate degrees of curvature of the two roadways beyond the theoretical gore.(pg. 194)(1)
Figure 4 shows an example sequence of arrow-per-lane signs for a split with an option lane. Important guidance for the arrow-per-lane signs states "No more than one destination should be displayed for each movement, and no more than two destinations should be displayed per sign."(pg. 198)(1)
Figure 4. Illustration. Arrow-per-lane sequence for a split.(1)
Diagrammatic guide signs show a graphic view of the exit/split geometry in relation to the main highway. As with arrow-per-lane signs, they are used when there is an option lane present at a freeway or expressway exit or split. The MUTCD provides the standards and guidance in section 2E.22 for this type of sign.(1) An example diagrammatic guide sign is shown in figure 5.
The standards prohibit the use of diagrammatic signs at cloverleaf interchanges except for the following:
Where the outer (non-loop) exit ramp of the cloverleaf is a multilane exit having an optional exit lane that also carries the through route; and at cloverleaf interchanges that include collector-distributor roadways, such as those that are accessed from the mainline by a multilane exit having an optional exit lane that also carries the through route. In this case, the Diagrammatic guide sign shall only show the configuration of the lanes at the exit point to the collector-distributor roadway and not the entire interchange configuration.(pg. 199)(1)
Figure 5. Illustration. Diagrammatic guide sign for a multilane exit.(1)
A C-D roadway is a one-way access road typically located next to freeway lanes that is used as the exit/entrance point for some or all of the ramps that would otherwise be merging with the freeway. Although the initial reaction to this type of geometry is to assume that it will simplify the interchange area by removing the exit/entrance points from the main lanes, this geometry leads to a new set of signing issues when advance and exit signing is needed to move traffic from the mainlines to the C-D road before the actual interchange or exit. Also, multiple exits may need to be signed as a single movement from the main lanes to the C-D. If this type of system is unfamiliar to a driver, it can violate driver expectations and create a need for greater signing information to ensure proper movements. The MUTCD contains limited guidance for signing for simple C-D roads, but solutions that extend these to modern C-D roads that may have access points for downstream exits miles before the actual intersection still need to be developed.
Preferential lanes are defined by the MUTCD as "Lanes designated for special traffic uses such as high-occupancy vehicles, light rail, buses, taxis, or bicycles."(pg. 253)(1) Managed lanes are a type of preferential lane that "Typically restricts access with the adjacent general-purpose lanes to designated locations only."(pg. 253)(1) Under varying operational strategies, the occupancy requirements of managed lanes can change, or it may even cost to use the lane based on time of day or congestion levels.
For a managed lane running adjacent to the mainlines of a freeway or expressway, the facility will require similar advanced guide and exit direction signs as the mainlines but will need sign sequences for both its entry and egress points to and from the lane. The MUTCD addresses these types of facilities and provides an example geometry and sign sequence.(1)
The combination of the signage for a managed lane facility and a complex interchange could quickly become overload for a road user, and special consideration would need to be taken to help the driver focus on the signs relevant to his or her travel.
In many ways, managed lanes are analogous to C-D roadways in that they provide limited access points that may serve multiple and distance downstream exit points. The principles developed for managed lane advance exit signing could be applied to C-D situations.
Concerning signage, the Institute of Transportation Engineers Freeway and Interchange Geometric Design Handbook stresses that signs should be placed where their message is integrated with other information that drivers use to make decisions, especially for departure from the freeway.(27) For example, the coordination of the sign, pavement markings, and the actual exit ramp itself, all in the same visual field, paint a picture for the driver for maximum clarity and comprehension.
Chapter 7 of the Traffic Signs Manual details the design of traffic signs in the United Kingdom (UK).(28) The design and layout of guide signs is based on x-height of the alphabet or font being used (the UK only uses two sign fonts: Transport Medium for positive contrast signs and Transport Heavy for negative contrast signs). The x-height is the height of lowercase letter "x" for that particular sign. All symbol and legend spacing, border widths, and radii are given in number of stroke widths (sw). The sw is one-quarter of the x-height.
The UK uses four types of directional signing or guide signing: stack type, map type, dedicated-lane type, and gantry-mounted type. Stack type signs are similar to the destination and distance signing used in the United States, while map type signs are diagrammatic signs.
Map type signs can be mounted on the roadway shoulder or on an overhead gantry. The design of the symbol on the map type sign is largely predicated on the junction. However, the lengths and widths of the route arms (i.e., directional arrows) are based on the route classification and the location of any legend or route shields. A width of 6 sw is used for primary routes, 4 sw for numbered non-primary routes, and 2.5 sw for non-numbered local routes. A width of 5 sw is reserved for routes indicated on a grade-separated junction and advance sign and for marking the approach arm of a roundabout at the end of an exit ramp on a grade-separated junction. The minimum length of a vertical route arm is 12.5 sw. Horizontal route arms are two-thirds the length of the destination legend associated with that arm. Inclined or angled route arms have a minimum length of 12 sw. When the advance sign is for a grade-separated junction, the width of all route arms is 5 sw, and the minimum length of the exit route arm is 24 sw. Map type signs can also show stubs. Stubs are shortened route arms that indicate a road but do not give a direction. The length of a stub is equal to its width. Warning and regulatory signs can also be placed within a map type sign. These signs are placed in line with the route arrow and can include distance plaques. Symbol signs (airport, parking, etc.) are also placed within the map type sign and are associated with a route arm. Route shields (e.g., U.S. highway shield) are not used on guide signs in the UK. Route numbers use a combination of a color legend, a color background panel, and a letter code to indicate the roadway classification (motorway, primary route, etc.). In addition, the background color of the map type sign indicates the classification of the traveled route.
The dedicated lane signs are used in advance of at-grade and grade-separated junctions. In the case of a grade-separated junction, the dedicated lane sign indicates the exit slip ramp. The directional arrows are 18 sw in length with an 8 sw head. If two or more lanes lead to the same direction, a horizontal bar is used. This is applicable for through lanes and exiting lanes. Lane widths on the sign should be equal for lanes with the same destination. The arrow indicating the widest lane should not be longer than two times the narrowest lane. Lane lines are always vertical, and the minimum length of a lane line is 3 sw. If this minimum length cannot be met, the lane line should be omitted. Destination distances are not to be shown on dedicated lane signs. The distance to the junction can be shown and is located in the lower corner of the sign. The dedicated lane arrow is vertical for the advance signing and is inclined only at the exit sign.
Gantry-mounted signs in the UK can include more than the destination name for a given direction. The destination names are separated using a comma. When using a non-lane drop gantry sign, two signs are created. The first is the through movement. This panel is the lower of the two and is centered over the main carriageway or main lanes. The second sign is positioned above the first and offset to the left (in the United States, this offset would most likely be to the right) such that the inclined directional arrow is not positioned over the lower sign. If the main lanes curve to the right (exit to the left), the lower sign arrows can also be inclined to the right. The length of the arrows is typically 16 sw. The downward arrows of the lane drop sign are to be centered over the traffic lanes. The legend is centered, and a horizontal bar is used. The sign should cover at least three-quarters of any lane to which it applies. In the case of a single lane, the sign panel may be wider than the lane but cannot cover more than one-quarter of a neighboring lane. The distance to the junction can be used and is added as a third sign or as a panel within the sign. Warning, regulatory, tourist, and destination distance information is not allowed on gantry-mounted signs.
The motorway in the UK is the equivalent of the U.S. interstate. Signing rules for the motorway follow the guidelines outlined in this section. A motorway sign is indicated by its blue background. The signs also include the junction number.
The Strassenverkehrs-Ordnung (i.e., Road Traffic Regulations), provides a summary of laws governing Germany's vehicles, traffic signs, and pedestrians published by the Federal Ministry of Transport, Building and Urban Affairs.(29) Part II, section 42, part 8 covers the use of guide signs and advance guide signs. The regulations give sign examples and descriptions for their use. Autobahn (equivalent to a U.S. interstate) signing uses a blue background. The signs can be shoulder-mounted or gantry-mounted.
Numerous studies have attempted to identify situations where drivers do not understand the lane assignment message being conveyed by a guide sign. This section summarizes some of the key points from these studies.
Two landmark laboratory studies conducted in 1970 served to develop guidance for the MUTCD.(1) NHTSA conducted the first of these laboratory studies.(30) The researchers showed subjects a series of signs with different guide signing concepts. The signs included conventional signs, diagrammatic signs that showed a plan view of the interchange, and diagrammatic signs that attempted to provide a driver's eye with perspective of the upcoming interchange. Although the diagrammatic signs did not perform significantly better than conventional signing in most cases, they significantly improved lane choice selections when C-D roads were present, a secondary split occurred on a ramp, and a major split occurred in the highway. Driver preference studies showed that drivers preferred diagrammatic signs with plan views over all other types of signs. The details of the study merit review because of their influence on the current project and current standards.
The study focused on graphical characteristics that would most effectively communicate roadway-interchange and route-guidance information to the driver. The researchers identified several interchange characteristics associated with traffic flow and accident rate. The existence of two or more of these characteristics occurring at an interchange warranted the use of a graphic guide sign. These interchange characteristics included the following:(31)
The interchange types that typically had two or more of these characteristics included the following:(31)
The laboratory study was divided into four parts based on measures of effectiveness: (1) lane choice, (2) subject confidence ratings, (3) guide sign interpretation, and (4) guide sign preference. The researchers used a dual-projection tachistoscopic method consisting of two slide projectors with timer-controlled shutters to measure subject response. One projector displayed the roadway scene with through-the-windshield images of a sign location. The second projector was fitted with a tachistoscopic shutter to project the image of a guide sign onto the roadway scene, overlaying the sign location. The shutter was timed for a 1-s exposure.
Prior to starting the test, subjects were given a destination and instructed on how to indicate lane choice and confidence level. A total of 102 people participated in this portion of the study. An example of the roadway scene shown to the participants is in figure 6 for the without guide sign information and in figure 7 for the with guide sign information. The researchers compared results of the graphic signs and found that a single sign type did not perform better than the other types across the interchange types. Testing the conventional signs against the graphic signs showed that graphic signs performed better with C-D interchanges, close-choice point interchanges, and major fork interchanges.
In addition to the timed comprehension testing, a preference test was conducted with the lane choice and confidence test. The subjects were shown a line drawing of an interchange and a list of sign types. The subjects were asked to pick the sign types that they liked best and least. The conventional signs were the least preferred (p < 0.05).
©National Academy of Sciences reproduced with permission
of the Transportation Research Board (TRB) from
Highway Research Record 414, Figure 2, pg. 27.
Figure 6. Photo. Roadway scene shown to subjects without guide sign information.(30)
©National Academy of Sciences reproduced
with permission of the TRB from Highway
Research Record 414, Figure 2, pg. 27.
Figure 7. Photo. Roadway scene shown to subjects with guide sign information.(30)
Further experiments of this project tested additional design elements to determine how well graphic signs convey information about roadways, such as safe exit speed, distance between exits, and location of the driver's exit. The researchers used the curvature of the arrow graphic and the distance between exits on the graphic as variables in this test. Subjects were asked to estimate the safe exit speed (miles per hour) and the distance (miles) between two exits. Two interchange designs were chosen, and for each interchange, four signs were tested: three graphic signs plus a conventional sign. There were 48 test subjects. The tachistoscopic method was used in the test as well. The researchers found that a curved exit arrow was understood to mean a lower safe exit speed. The second graphic sign concept had twice the spacing than the other two. Drivers judged the distance between exits on the first graphic sign concept as being greater than the other two signs. The conventional signs had the highest estimate of exit speed. A significantly greater percentage of subjects correctly identified their exit with the graphic signs than the conventional signs.
Based on the results of the three tests, the researchers determined that graphic guide signs can help improve lane position for closely spaced exits, C-D interchanges, and major fork interchanges. The exit arrow can be used to provide information on exit speed and the distance between the exit ramps.
FHWA conducted a follow-up laboratory study to the NHTSA diagrammatic sign study. This study modified the NHTSA study procedures by testing each subject individually and testing both destinations shown on the study signs.(31) A total of 60 test subjects viewed a series of slides with diagrammatic or conventional signs for six interchanges on I-495 in Washington, DC. Lane choice, reaction time, and driver preference for each type of sign was evaluated. The slide exposure time was controlled by the subjects, who pressed a button when they felt they understood the sign. This study found that drivers generally performed better at lane selection and had shorter reaction times with conventional signing. The conventional signs were also preferred by a larger number of test subjects than the diagrammatic signs. It is possible that greater driver familiarity with conventional signs than the then-experimental diagrammatic signs may have influenced these results.
Another study examined the relative effectiveness of using diagrammatic signs rather than conventional guide signs.(32) A total of 120 participants viewed a series of slides. They indicated which lane they would travel in to reach a predefined destination, and the correctness and latency of the responses were recorded. This study found that there was no significant difference between the use of diagrammatic and conventional guide signs. The findings showed that subjects responded more quickly to conventional guide signs and generally seemed to prefer them to diagrammatic signs.
In general, these evaluations of diagrammatic signs did not show conclusive evidence that the diagrammatic signs outperformed conventional signs. The first large laboratory study showed strong preference for graphic signs and corresponding gains in performance, but the two subsequent studies showed that conventional signs performed better than graphic signs. These results may be biased, however, since the studies were conducted at a time when diagrammatic signs were not familiar to many drivers. It is possible that results would be different if such a study were conducted today. The research did identify specific geometric situations where the diagrammatic signs performed better than the conventional signs, but they did not show a widespread superiority over conventional guide signing across a range of conditions.
Recently, researchers made several recommendations to alter diagrammatic signs to improve the understanding of older drivers.(33) The authors recommend using a modified form of diagrammatic signing using a separate lane assignment arrow to indicate lane use on a freeway. The number of arrow shafts on the modified diagrammatic sign should be the same as the number of lanes on the freeway. The report notes that this configuration is not approved by the MUTCD and requires FHWA permission before it can be used. These recommendations were derived from a 1990 TTI project where Skowronek examined the use of different guide sign formats at freeway interchanges in Houston, TX.(34) He conducted a driver survey and tested conventional signing, diagrammatic signing, and modified diagrammatic signing. The major findings of Skowronek's study are as follows:(34)
Another study focused on lane choice at exit direction signs.(35) Driver surveys were used and produced the following findings:
A recent study sponsored by NCHRP addressed two-lane freeway exits with one optional and one exit only lane.(4) The signs that were tested are similar to those shown in figure 8 through figure 10 but used a standard exit only plaque for the far right lane. The researchers also tested conventional text-only signs with pull-through (down) arrows and conventional diagrammatic signs. These signs were compared to various arrangements of down arrows and exit plaque arrows slanted up and to the right. A total of 96 participants drove in a driving simulator and were asked to follow signs to a particular destination. Measures of effectiveness of the various signs included path deviations (i.e., swerving) and lane changes. This flexibility is one advantage to using a dynamic driving simulator or on-road test. Rather than a discrete choice of lane as used in most surveys of sign comprehension, dynamic tests allow lane indecision to be assessed by examining the path of the driver. Overall, this study showed that about one-third of drivers made unnecessary lane changes, demonstrating their poor understanding of optional lane exits.
Reproduced with permission of the TRB.
Figure 8. Illustration. Lane designation signs.(4)
Reproduced with permission of the TRB.
Figure 9. Illustration. Advance guide sign located approximately 0.5 mi in advance of exit and centered over the four approach lanes.(4)
Reproduced with permission of the TRB
Figure 10. Illustration. Advance guide sign located 1 mi in advance of exit and centered over four approach lanes.(4)
Another study conducted at TTI evaluated driver comprehension of diagrammatic freeway guide signs and their text alternatives through a multiphase human factors study.(7) Signing for four different types of interchanges was tested: (1) left optional exit, (2) left lane drop, (3) freeway-to-freeway split with optional center lane, and (4) two-lane right exits with optional lanes.
The strong effect of the addition of an exit only plaque on an advanced guide sign was one of the most striking discoveries made throughout this research. Drivers did not fully grasp the meaning of the exit only plaque and thus consistently made incorrect decisions about the meaning of signs displaying this plaque.(4) The addition of the exit only plaque tended to have the effect of increasing unnecessary lane changes for the exit route but reducing the unnecessary lane changes for the through (or mainline) route. The exit plaque tended to pull drivers planning to exit all the way over to the lane marked with the exit only plaque. From this result, researchers inferred that drivers tend to believe "exit only" means that the lane over which the sign is displayed is their only option to exit the main roadway. In the TTI research, the same trend was apparent, with the majority of drivers incorrectly assuming they must be in the exit only lane.(7)
However, the misunderstanding of the plaque did not always result in negative outcomes. In a lane drop scenario, more drivers correctly assumed that the left lane was their only option to exit. Although this assumption may again illustrate the misunderstanding of the plaque, it did increase correct responses when participants were attempting to follow the through route.
Based on these findings, future research should focus on using exit only plaques for optional lane situations, including multilane exits and splits. While the driving habit of going to the outside lane "just to be sure" could promote safety by reducing lane changes at the gore, it would reduce the capacity of the interchange.
In the text versus diagrammatic portion of the study conducted using the driving simulator, the results were consistent with the trend of the sign sequences resulting in the fewest unnecessary lane changes also receiving the shortest lane change distance. Researchers attributed this phenomenon to the fact that text-based signs may be easier to understand across the entire driver population, but they are not necessarily visible from as long a distance as the diagrammatic or modified diagrammatic signs. The longer viewing distance may lead drivers to guess the lane configuration based on the large arrows before they can read the destination names. This can lead to more unnecessary lane changes due to chance; however, when these guesses are accurate, the early changes result in longer lane change distances.
The modified diagrammatic signs that were tested were originally based on those that contained only route shields because these signs are relatively straightforward and uncluttered. This sign design performed very well in the early phases of the project. However, in later phases, additional elements were added to the modified diagrammatic signs to equate the amount of information present on the other sign types. This information included route shields, cardinal directions, and destination city names. When the additional elements were added, the signs became crowded and visually complex, resulting in relatively poor performance in the final phases of the research. Additionally, if this sign format is applied to larger freeway interchanges with more lanes represented, the visual complexity will increase. More research is needed to refine the design of these signs and test their application at complex interchanges.
Several studies tested diagrammatic signs in the field. Roberts and Klipple examined the use of diagrammatic freeway guide signs in New Jersey by implementing diagrammatic signs at an interchange.(36) The researchers collected data on erratic maneuvers and traffic volumes at the interchange when conventional signing, diagrammatic signing, and diagrammatic signing with lane lines were used. In general, the diagrammatic signs performed better than the conventional signs. The number of erratic maneuvers dropped when diagrammatic signs were implemented and was reduced further when lane lines were added to the diagrammatic signs.
The research was performed at the request of FHWA. The location selected for evaluation was the interchange of I-287 and US-22 in Somerville, NJ. The study had the following three parts:
The researchers conducted before and after studies for each sign change and included after studies for the initial use of diagrammatic signs. The initial before study was performed in July and August 1969, and the final after study was completed in May 1970. The I-287 northbound (NB) to US-22 westbound (WB) exit was chosen as a study site. The eastbound (EB) exit to US-22 had a low traffic volume and was not included in the study. Researchers recorded the number of unusual or erratic maneuvers at the exit gore. Researchers collected data using automatic traffic counters and video recorders, including through and left exit volumes. Traffic was videotaped as it approached the exit gore 400 ft upstream from the gore. All lanes were recorded, and data were collected between 2 and 7 p.m.
The researchers found no significant differences (95 percent confidence level) in the rate of unusual maneuvers between the original signs and the modified signs. A significant reduction was found when the signs were changed to diagrammatic signs. This reduction may be attributable to the uniqueness of the diagrammatic signs (commanding greater attention) and the fact that drivers may have felt that the change in sign type indicated a need for greater attention. A comparison of the after and long-term after studies for the diagrammatic signs showed an increase in the rate of unusual maneuvers. The researchers felt this could be attributed to changes in the traffic makeup and the 6-month span between data collection periods. After the addition of lane lines to the diagrammatic signs, researchers noted a significant decrease in the number of unusual lane changes.
The Virginia Highway Research Council conducted a diagrammatic sign field study in 1970 that examined traffic volumes and the number of erratic maneuvers at the site in Washington, DC.(37) The number of erratic maneuvers increased after the diagrammatic signs were installed, but the researchers noted that data for diagrammatic signs were collected during late spring and early summer. During these months, the proportion of drivers not familiar with the area increases on the highways around Washington, DC. The researchers hypothesized that these non-local drivers were responsible for the increase in erratic maneuvers.
The study evaluated erratic maneuvers using time lapse photography. The variables included the following:
Researchers chose the exit 1 interchange on the Capital Beltway south of Alexandria, VA, as the study site. This location exhibited sight distance restrictions and an unusual geometric layout.
The 85th percentile speed at the study location was determined to be 45 mi/h during the morning and afternoon peak times and 65 mi/h during the off-peak times. The main lanes at the exit had a volume of 81,000 vehicles per day. An accident analysis showed that over a 26‑month period prior to the study, there were 240 accidents, including 4 fatalities and 136 injuries.
The researchers used the comparative erratic maneuver method for their analysis. They divided the study area into zones and recorded erratic vehicle movements in each zone. The erratic maneuvers identified included the following:
Traffic volumes and erratic maneuvers were recorded at random times during the day for 30-min intervals. The before data were collected during fall 1970 and early spring 1971. Diagrammatic signing replaced the standard guide signing. The diagrammatic signing used 20‑inch route name letter heights and 36-inch route shields. The sign itself measured 14 by 19.5 ft.
The before period covered 19 days, and 56,326 vehicles were observed over 47 30-minute intervals. The after period observed 91,423 vehicles over 73 30-minute intervals. The research compared the before and after traffic volumes and found no evidence that tourist traffic had a significant effect on traffic volume, suggesting a similar mix of familiar and unfamiliar drivers during the study periods. The researchers determined that after the installation of the diagrammatic signing, fewer motorists were weaving across the gore. The researchers also noted an increase in the amount of weaving traffic across the solid line pavement marking in advance of the gore area, which indicates that drivers were making lane decisions earlier. The researchers also noted that while the use of the diagrammatic signs reduced gore area weaving, it increased the number of hesitations and partial weaves. The number of stopping and backing maneuvers also decreased.
A 1972 report by Mast and Kolsrud examined the use of diagrammatic signs on controlled access highways.(38) The objective of this research was to develop warrants and standards for the use of diagrammatic guide signs. The field studies used an instrumented vehicle equipped with an in-vehicle sign display system. Subjects were required to navigate the test route using the information supplied by the in-vehicle signs for destination and direction. The routes used real highway facilities and interchanges open to normal traffic. The researchers measured the drivers' sign information interpretation time, vehicle speed control, incidence of hazardous maneuvers, and exiting errors.
As a result of the field studies, the researchers discovered the following three general findings:
The researchers concluded that diagrammatic guide signs should be used in advance of left exit interchanges.(38) These interchanges include major forks where the through traffic uses the right fork and exiting traffic takes the left fork, interchanges where there is a single left exit in combination with a right exit, and all single left exit interchanges. The researchers also recommended four cases where diagrammatic guide signs should not be used, as the use of diagrammatic guide signs in these cases provides no benefit to the driver and in some instances may reduce driver performance:
In addition to the application warrants, Mast and Kolsrud developed design standards for diagrammatic signs as part of this study.(38) These design standards still form the foundation of current designs in the MUTCD. The warrants and general design standards were developed from approximately 20 study sites in eight States: Arizona, Connecticut, Illinois, Michigan, New Jersey, Virginia, Wisconsin, and Wyoming.
The researchers identified the following 19 general design standards:
Another study used a laboratory eye tracker to examine eye scanning of day and nighttime roadway scenes.(39) Participants were seated in front of a computer monitor with their heads in a chinrest. Photographs of roadway scenes were digitally manipulated to produce different levels of clutter and luminance. While this method allows for very exact eye tracking, it is not practical to use a chinrest arrangement in a vehicle on the road.
Another study using eye tracking examined lane change behavior.(40) This method, coupled with vehicle instrumentation, allowed for fine-grained analyses of driver attention and decisionmaking while making lane change choices.
Situations where a lane is dropped at a freeway interchange have the potential to violate driver expectancy and can cause confusion among drivers. This confusion can result in high-speed variability, erratic maneuvers, and driver frustration, all of which negatively impact safety. A variety of research has been performed to assess the effectiveness of different ways of signing lane drops at interchanges.
A study was conducted in the mid-1970s to assess the effectiveness of interchange lane drop signing standards.(41) This study examined left- and right-side exits for single lane drops. After reviewing the literature, surveying State agencies, and performing some limited driver surveys, the researchers developed the following recommended treatments for signing interchange lane drops:
In a 1996 TxDOT project, Somers et al. evaluated alternative treatments for right-side multilane exits.(8) First, the researchers evaluated innovative ways to sign an optional exit lane for a multilane exit. They tested the supplemental messages "Exit OK" and "May Exit" for use on the optional exit lane. They also examined the use of a divergent arrow over the optional lane to indicate lane usage. The divergent arrow was tested by itself as well as in conjunction with the "Exit OK" and "May Exit" messages. The researchers hypothesized that this additional guidance would improve driver understanding of the use of the optional lane.
These alternatives were examined by surveying 548 participants and evaluating their lane choices and comprehension of the messages. This survey produced the following results:
The researchers then examined methods for signing a multilane exit with an optional lane exit followed by a secondary ramp split. This study evaluated treatments that utilized the "May Exit" supplemental message and modified standards from Ohio and Texas. This study showed that the differences between the "May Exit" and modified Texas standard were not as large as the earlier survey results indicated. None of the methods provided a significant improvement over existing methods for signing a multilane exit followed by a secondary ramp split.
The review of research literature and applicable signing standards helped the research team identify signing problems and potential solutions. The main problems identified in the literature review are as follows:
