• Federal Highway Administration >
• Publications >
• Research Publications >
• 03065 >
• Altext1.Cfm >
• In-Vehicle Display Icons and Other Information Elements: Volume I

In-Vehicle Display Icons and Other Information Elements: Volume I

PDF files can be viewed with the Acrobat® Reader®

508 Captions - Final August 18, 2004

Figure 1-1.   Graphic. Format Used in the Preliminary Design Guidelines. The graphic shows a sample of the two-page format used to present each design guideline, with arrows pointing to the standard parts of the format. On the left side, arrows point to the guideline title at the top of the page, beneath which is the introduction paragraph. Under the introduction is the design guideline, which uses a bar scale rating to indicate whether designs should be developed based on experimental data or expert judgment. Below the design guideline is the figure, table, or graphic. At the center of the bottom of the page is the page number. At the top of the right side, arrows point to the header, which actually is found on both pages. The header includes the abbreviated handbook title, the abbreviated chapter title, and the revision date. Below the header on the right side is the discussion section, followed by the design issues, cross references, and references sections. At the center of the bottom of the page is the page number.

Figure 2-1.   Flowchart. General Design Process for In-Vehicle Icons. This decision tree shows a process for developing in-vehicle icons. The first step is to identify the in-vehicle message requiring an icon. Next, review the general issues in icon design (see chapter 2). After reviewing these, determine if a standard icon exists (see chapter 15). If one does exist, then conduct analyses to determine if the icons conform to the design principles and guidelines for legibility, recognizability, and interpretability. See chapters 3, 4, and 5 of this report for more information.

If a standard for the icon does not exist, determine if candidate icons have been developed and or if they are in use (see chapter 8). If candidate icons have been developed, then conduct analyses to determine if the icons conform to the design principles and guidelines for legibility, recognizability, and interpretability. See chapters 3, 4, and 5 of this report for more information.

If candidate icons have not been developed, then use production tests to generate ideas for icons (see chapter 7). Then, conduct analyses to determine if the icons conform to the design principles and guidelines for legibility, recognizability, and interpretability. See chapters 3, 4, and 5 of this report for more information.

After conducting the analyses to determine if icons conform to the design principles and guidelines for legibility, recognizability, and interpretability, conduct appropriateness ranking and comprehension recognition tests (see chapter 7). Based on these tests, determine if there is a poor response to the icons. If the response is poor, perform the production tests; the analyses to determine if the icons conform to design principles and guidelines for legibility, recognizability, and interpretability; the appropriateness ranking tests; and comprehension recognition tests again until the response to the icons is no longer poor.

After the response to the icons improves, the testing may be complete. Proceed as necessary or complete testing by conducting matching tests (see chapter 7). When needed, redesign the icons and repeat the evaluations. Consider alternate approaches to evaluations. See chapter 7 of this report for more information.

Figure 2-2. Table with Graphics. Examples of the Appropriate Use of Icons. This figure begins with a bar scale. The left side of this scale reads "Based Primarily on Expert Judgment", the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line on the bar scale is shaded five-sixths of the way to the right side of the scale. This indicates that determining when to use icons primarily is based on experimental data, although some expert judgment might also be required.

The table shows examples of when it is appropriate to use certain types of icons. If quick and accurate recognition of a message is necessary, then an example would be a diamond-shaped sign with an orange background showing a black silhouette of a person shoveling. The inappropriate example is a diamond-shaped sign with a yellow background that only has "Road work ahead" printed on it.

To display visual or spatial concepts, an example is given of an icon used to depict that the road curves ahead. This is a diamond-shaped icon with a yellow background with a black arrow attached to a wavy line. The inappropriate example is a diamond-shaped icon with a yellow background that only has "curves ahead" printed on it.

To present a set of alternatives, an example is given of four square icons with brown backgrounds displayed together, with two on top and two below. Clockwise from the top left, the icons are: the sign for women's and men's restrooms, which show a female silhouette and a male silhouette in white with each silhouette drawn inside of a box; the sign for a gas station, which is a white silhouette of a gas pump; the sign for a telephone, which is a white silhouette of a telephone handset; and the sign for lodging, which is a white silhouette of a person lying in a bed. The four respective inappropriate examples given are four square signs with brown backgrounds that only have "restrooms," "gasoline," "telephone," and "lodging" printed on them.

Figure 2-3. Flowchart. Information Processing Elements Used to Identify Icon Requirements. This figure begins with a bar scale. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded one-third (two of the six sections) of the way from the left of the scale, indicating that how icons are used to facilitate driving tasks is based almost equally on expert judgment and experimental data, although expert judgment might be slightly more useful than the experimental data.

This figure shows nine information processing elements (IPE) that define the information requirements of a driver interacting with an IVIS device. These elements comprise a decision cycle represented by the circle. Between each IPE, phrases in italics identify a knowledge state that acts as the input to one element and the output of the previous element. Together the nine elements describe the range of information processing activities supported by IVIS messages. The decision cycle consists of four quadrants, with the IPEs within each quadrant serving a common purpose. The first quadrant, Attention, involves detecting disturbances and deviations and directing attention toward the disruption. The second quadrant, Interpret, builds upon this to classify and understand attended inputs. The third quadrant, Selection, uses this interpreted information to identify an appropriate course of action. The fourth quadrant, Action, carries out the course of action. Each quadrant helps identify general design requirements for supporting driver decisions and the IPEs within each quadrant identify specific requirements.

Figure 2-4. Table with Graphics. Types of Icons. This figure begins with a bar scale. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded one-third (two of the six sections) of the way from the scale, indicating that determining types of visual icons is based almost equally on expert judgment and experimental data, although expert judgement might be slightly more useful than experimental data.

This table demonstrates the types of icons to use. It has three columns. The first column is icon type, followed by key examples in the second column, and other examples in the third column. If the icon type is image-related, the key example is an icon showing "Fasten seatbelt." The icon is round with a black background and a white silhouette of a seatbelt about to be snapped together. Other examples include the "Telephone ahead" icon, which is a square icon with a brown background and a white silhouette of a telephone handset, and the "Gas station ahead" icon, which is a square icon with a brown background and a white silhouette of a gas pump.

If the icon type is concept-related, a key example is the icon used to show the flash function on a camera or a high voltage symbol in a power plant. The icon is a black lightening bolt. Other examples are the "Curve signs," which is a diamond-shaped sign with a yellow background and a black arrow attached to a curved line, and the "Elevator" icon, which is a square icon with a blue background and a white silhouette of a man and woman in a box with up and down arrows drawn above the box.

If the icon type is arbitrary, the key example is the addition symbol, first-aid symbol, or international symbol for the American Red Cross, all of which are depicted with two red bars of equal length crossing in the middle, and the entire symbol is positioned inside a rectangle. Other examples are the icon used by the U.S. Postal Service, which is a blue silhouette of a bald eagle with a red bar under it, and the symbol of the medical profession, which is a caduceus, or two snakes entwined around a staff with wings at the top.

Figure 2-5. Graphic. Key Components of an Icon. This figure begins with a bar scale. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded two-thirds (four of six sections) of the way from the left to the right side of the scale, indicating that determining the composition of an icon is based almost equally on expert judgment and experimental data, although the experimental data might prove slightly more useful than the expert judgment.

This graphic shows the key components of an icon. The warning icon is used as an example, which is a square icon with a gray background, bordered by a black line. Inside the square is a yellow triangle with a black border. Inside the triangle is a black exclamation point. The key components of an icon are the border, which is the black line around the icon, the background or the area behind the triangle, the symbol, which is the triangle in this case, the element, which is the actual picture or image inside the symbol that pictorially displays the message of the icon, and the text label which is any wording used to describe the icon, its meaning, or show instructions.

Figure 2-6. Decision Tree. Sequence of Icon Comprehension and Use. This figure begins with a bar scale. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded two-thirds (four of six sections) of the way from the left to the right side of the scale, indicating that considering icon comprehension and determining whether icons are legible, recognizable, and interpreted correctly depends almost equally on expert judgment and experimental data, although the experimental data might prove slightly more useful than the expert judgment.

The sequence of icon comprehension and use starts with determining the icon's legibility. The questions to ask are "Can the driver see the icon?" "Is it legible at various distances?" and "Can it be seen under both nighttime and daytime lighting conditions?" See chapter 3 of this report for more information. After determining legibility, determine whether the icon can be recognized. The questions to ask are "How well do the parts of this symbol relate to one another?" "Does the construction of the symbol support accurate recognition?" and "Is it easily confused with other symbols?" See chapter 4 of this report for more information. Next, determine how the icon will be interpreted. The questions to ask are "How well does the icon represent the message?" "Will it be understood when presented in the appropriate context?" and "Does it require special knowledge particular to a culture, language, or drive age?" See chapter 5 of this report for more information.

Figure 15 Graphic. Measuring Luminance Nonuniformity. This figure begins with a bar scale. To determine the appropriate luminance uniformity within an icon, the design guidelines specify no more than 33 percent element nonuniformity. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that determining the percent of luminance uniformity primarily will be based on experimental data, although some expert judgment might be required.

Figure 3-1. Graphic. Measuring Luminance Nonuniformity. This graphic shows how to measure luminance nonuniformity. It contains a library icon, which is a square icon with a green background and a white silhouette of a person holding a book. An arrow labeled "Luminance A" points to the person's head. Another arrow, labeled "Luminance B," points to the book. Below the icon is printed, "Within a segment of an icon, measure at two locations using a photometer with a spot size small enough to fit inside the segment or element."

Also within the graphic is equation 1 for determining the percent element nonuniformity.

Equation 1. Percent Element Nonuniformity. Percent element nonuniformity is the absolute value of the minimum luminance minus the maximum luminance, which is then divided by the maximum luminance. Minimum luminance is equal to the smaller luminance value and maximum luminance is equal to the greater luminance value.

Equation 2. Contrast Ratio. The contrast ratio equals the luminance maximum divided by the luminance minimum. The luminance maximum is equal to the luminance emitted by the area or element of greatest intensity, and the luminance minimum is equal to the area or element of least intensity.

Figure 3-2. Graphic. Example of Measuring Contrast. This figure begins with a bar scale. These design guidelines indicate that a three-to-one ratio is the minimum symbol contrast, and that a seven-to-one ratio is the preferred symbol contrast. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that determining the appropriate contrast will be based almost primarily on experimental data, although some expert judgment also might be required.

Above the figure is printed "The figure below may be used to aid contrast measurements." Note: The spot size of the photometer used to take luminance measurements must be small enough to fit inside the icon elements being measured. Inside the figure is the library icon, which is a square icon with a green background and a white silhouette of a person holding a book. There are two sets of arrows. The first set of arrows indicate the three points where measurements should be taken to find minimum luminance, including an area slightly below the left corner of the silhouette, an area slightly to the right of the corner of the silhouette, and in a larger area closer to the right side of the silhouette. After taking the measurements, calculate the average of the measurements to obtain the minimum luminance. The second set of arrows indicate the three points where measurements should be taken to determine the maximum luminance, including an area on the head of the silhouette, on the body of the silhouette, and on the book. After taking the measurements, calculate the average of the luminance values to determine the maximum luminance.

Equation 3. Visual Angle (first equation). This is an equation for determining the visual angle if the distance and symbol height are known. The visual angle equals the arc tan of the symbol height divided by the distance. The symbol height is equal to the height of the symbology. Distance is equal to the distance from the viewer's eyepoint to the display. The visual angle equals the angle in degrees.

Equation 4. Visual Angle (second equation). This is another equation for determining the visual angle if the distance and symbol height are known. The visual angle equals the 3,438 multiplied by the height divided by the distance. The symbol height is equal to the height of the symbology. Distance is equal to the distance from the viewer's eyepoint to the display. The visual angle equals the angle in degrees.

Equation 5. Symbol Height. This is an equation for determining the height of the symbol if the distance and visual angle are known. The height of the symbol equals the tangent of the visual angle multiplied by the distance. The symbol height is equal to the height of the symbology. Distance is equal to the distance from viewer's eyepoint to the display. The visual angle equals the angle in degrees.

Equation 6. Distance. This is an equation for determining the distance when the visual angle and symbol height are known. The distance is equal to the symbol height divided by the tangent of the visual angle. The symbol height is equal to the height of the symbology. Distance is equal to the distance from the viewer's eyepoint to the display. The visual angle equals the angle in degrees.

Figure 3-3. Drawing. Relationship Between Viewing Distance, Symbol Height, and Visual Angle. This figure is related to a bar scale at the top of the page. The design guidelines indicate that the optimum visual angle of symbols contained in an icon is 1.43 degrees or 85 arcmin. The minimum visual angle is 0.69 degrees or 41 arcmin. The optimum visual angle of text labels contained in an icon is 0.40 degrees or 24 arcmin. The minimum visual angle of text labels is 0.27 degrees or 16 arcmin. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that determining the appropriate visual angle primarily will be based on experimental data, although some expert judgment might be required.

The figure shows the relationship between viewing distance, the symbol height, and the visual angle. On the left side of the figure is the eyepoint, which is represented with a drawing of an eye. On the right side of the figure is the symbol, which in this case is the warning triangle. The horizontal line between the eyepoint and the bottom of the symbol represents the distance from the viewer to the symbol. The height of the symbol from top to bottom is the symbol height. The diagonal line drawn between the top of the symbol and the eyepoint is used with the horizontal line to form the visual angle.

Figure 3-4. Graphic. Schematic Examples of Text Labels. This figure begins with a bar scale. The design guidelines indicate that when designing effective text labels, they should have the following attributes.

• Text labels should be brief, no more than two to three words.
• The type should be at least 0.27 degrees of visual angle or 16 arcmin and 0.40 degrees or 24 arcmin at most.
• Use a clear and simple sans serif typeface such as Helvetica. Avoid using boldface, italics, underlining, or differences of color to emphasize words.
• The space between lines should be at least one-thirtieth the line length.
• Use both uppercase and lowercase letters rather than all of one or the other.
• Character width-to-height ratios should be at minimum 0.6 to 1 and 1 to 1 at most.
• Wider characters should be used as criticality increases.
• If text must be compressed, compress the space between characters rather than the characters.
• Use initial capital letters for multiword labels not incorporated in the symbol.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded two-thirds (four of six sections) of the way from the left to the right side of the scale. This indicates that determining how to design effective text labels will be based almost equally on both expert judgment and experimental data, although slightly more experimental data might be required.

This figure is divided into two parts. The top part of the figure shows the ineffective use of text labels, while the bottom shows the effective use of text labels. Two icons are shown in each part. These icons are diamond-shaped road signs with a yellow background and has an arrow at the top pointing up and an arrow at the bottom pointing down. Between the two arrows, it has the numbers representing seven feet, six inches. In the portion of the figure showing the ineffective use of text labels, the text label below the icon reads, "Overheight vehicles take another route around restricted clearance." The word "restricted" is italicized. Three arrows pointing to this text label and make the following suggestions: the text label type should be greater than or equal to 0.27 degrees of the visual angle; the space between the lines should be at least one-thirtieth the line length; and avoid using italics to emphasize words. In the part of the figure showing the effective use of text labels, the words "Low Clearance Ahead" are written in larger type below the road sign icon. Three arrows pointing to the text label make the following suggestions: use a clear and simple font; use initial capital letters when the label is not part of the symbol; and keep text labels brief.

Figure 3-5. Equation for Determining Color Contrast. (This figure is also equation 7.) This figure begins with a bar scale. The design guidelines indicate that when choosing colors for an icon, the following should be considered for legibility:

• Any reasonably visible color may be used to create icons as long as guidelines for symbol height and contrast are followed and population stereotypes are not ignored. However, highly saturated blue (i.e., approximately 450 nanometers) should be avoided.

• If colored lines are shown against a colored background, the color contrast between the elements should be a minimum of 100 E (CIE Y the first derivative of lowercase U the first derivative of lowercase V) distances.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that determining colors for an icon primarily should be based on experimental data, although some expert judgment could be required.

The equation is: Delta E parenthesis CIE Y, first derivative of U, first derivative of V, close parenthesis equals the following: (167 times the delta of the first derivative of V all squared) plus (367 times the delta of the first derivative of U all squared) plus (delta Y divided by Y times M all times 155) and then all squared. After that number is calculated, find its square root. In the equation, Delta E parenthesis CIE Y, first derivative of U, first derivative of V, close parenthesis is equal to the color contrast metric. Delta Y is equal to the difference in luminance between text or symbology and background. Y times M is equal to the maximum luminance of text or symbology or background. The delta of the first derivative of U is equal to the difference between the first derivative of U coordinates of the text or symbology and the background (per the 1976 CIE UCS; see note below). The delta of the first derivative of small V is equal to the difference between the first derivative of small V coordinates of the text or symbology and the background (per the 1976 CIE UCS; see note below). Note that the constants 155, 367, and 167 in equation 7 are empirically derived weights (Reference 1).

Figure 4-1. Table with Graphics. Five Levels of Realism. This figure begins with a bar scale. When designing an icon or symbol, consider the following:

• For general or abstract concepts, less detailed symbols such as caricatures or silhouettes are most appropriate.

• When several symbols have the same general shape or profile, detail is necessary to make them distinct from one another. In this case, they may be best portrayed using a simplified drawing.

• For small, familiar symbols with a distinct profile, use an outline. However, when the symbol is too thin to recognize in this format, a silhouette is preferred.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded one-third (two of six sections) of the way from the left to the right side of the scale. This indicates that determining how to design an icon or symbol using levels of realism will be based almost equally on expert judgment and experimental data, although slightly more expert judgment might be required.

The table is divided into four columns: example, (this column includes the icons), level of realism, design style, and when to use. In the example column, in the first row is a black-and-white drawing of a tractor-trailer truck. The level of realism is a simplified drawing. The design style is a simplified drawing with distinct interior details. This style should be used for presenting complex symbols with small significant parts, especially when objects have similar profiles, such as mechanical or electrical devices. The second row or example is a square icon with a green background, with a cartoon-like drawing of a car with five people sitting in it. The level of realism is caricature. The design style is exaggeration of crucial details. This style should be used for presenting symbols that have a small, crucial feature or for simplifying complex details. In the third row, the example is an outline drawing in black and white of a person wearing a seatbelt. The level of realism is outline. The design style is outline with only prominent details. This style should be used for presenting small objects that represent a familiar object with a distinct profile. In the fourth row, the example is the lodging symbol, a square icon with a blue background and a white silhouette of a person lying in a bed. The level of realism is silhouette. The design style is shape filled with solid color contrasting with background. This style should be used for presenting symbols that are too thing to show in outline format and for symbols that have a very distinct profile and do not require detail for recognition.

Figure 4-2. Graphic. Example of Appropriate Level of Detail.This figure begins with a bar scale. When determining the level of detail in an icon or symbol, consider the following:

• Design symbols on a 20 by 20-unit grid, making sure that no significant detail is smaller in size than 1 square unit.
• Lines and other continuous aspects of the symbol do not need to span one grid square.
• Significant details within a symbol should subtend, at a minimum, 3 degrees of visual angle.
• Line thickness for a significant detail should subtend, at a minimum, 2 degrees of visual angle.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded one-third (two of six sections) of the way from the left to the right side of the scale. This indicates that determining how to design an icon or symbol using level of detail will be based almost equally on expert judgment and experimental data, although slightly more expert judgment might be required.

The figure shows a square sign with rounded corners and a dark black border overlaid over on a 20 by 20-unit grid, which is then overlaid on another grid that is set at a 45-degree angle. The road sign is a bicycling symbol; a silhouette of a person riding a bicycle. Three arrows point to different parts of the sign. One arrow points to the symbol and is labeled, "Determine the appropriate level of realism (e.g., silhouette)." Another arrow points to the gridded area that is the background of the sign and is labeled, "No significant detail should be smaller in size than 1 square unit." The third arrow points to the outside of the sign and is labeled, "Design symbols on a 20 by 20 unit grid."

Figure 4-3. Table with Graphics. Design Guidelines. The table shows the icon guidelines for designing the visual characteristics of icons. It has four columns labeled icon design parameter, recommendation, do this..., and ...not this. Under the figure/ ground relationship icon design parameter, the recommendation is to "emphasize a clear, stable, and solid relationship between the elements of the symbol and its background." The example symbol in the "do this" column is a square, no-left-turn road sign where a black, left-turn arrow is set behind the red "do not" circle or a red circle with a red 45-degree slash from the top left to bottom right, so the red line crosses over the left-turn arrow. The example in the "not this" column is the same image, but the black, left-turn arrow is set in front of the red "do not" circle so that the diagonal red slash crosses behind the black left-turn arrow. The recommendation for the figure edges design parameter is that "relatively solid shapes are better than thin or dotted-line edges, unless the element in question depicts action or movement." The example symbol in the "do this" column is a black, solid arrow pointing left in a red, solid "do not" circle. The example in the "not this" column is the same arrow, but the "do not" circle has a dotted-line border and a white background. For the closure design parameter, the recommendation is to "use closed figures without discontinuous lines, outlines, or disjointed elements that can result in a fragmented figure." The "do this" example is a black silhouette of a school bus. The "not this" example is an outline of a school bus where the numerous lines are not joined together. If the icon design parameter is simplicity, the recommendation is that "icons should be simple with only the necessary detail included. Removal of these details should result in low recognition." The "do this" symbol is a square traffic sign with a black silhouette of the front of a car with the word "Taxi" written in bold, uppercase letters above it. The "not this" symbol is a square traffic sign with a very detailed drawing of the side view of a taxi. If the icon design parameter is unity, the recommendation is that "all parts of the symbol should be enclosed within a single boundary." The "do this" symbol shows a square traffic sign with a "no parking" symbol, which is a black letter P with the red "do not" circle around it. Below the no parking symbol within the same border are two arrows pointing to the left and right. The "not this" symbol shows the "no parking" symbol, with the words "No Parking" written below it in small, uppercase text, and arrows pointing to the right and left enclosed in their own box. There is no border around the entire symbol.

There is a bar scale with a line underneath this table. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded all of the way from the left to the right side of the scale. This indicates that determining the visual characteristics of the icon will be based entirely on experimental data.

Figure 4-4. Graphic. Schematic Examples of the Appropriate and Inappropriate Use of Flashing Icons. This figure begins with a bar scale. When determining the flash rate for a signal, consider the following:

• Flashing lights or icons should be reserved only for emergencies, since they have the potential to distract the driver.
• Flash rates should be 3 to10 per second (although, 4 is best) with equal light to dark intervals.
• Flash duration should be at least 0.05 second.
• Avoid having more than one signal flash at a time.
• Keep the background steady when using flashing signals.
• Do not use flash rates as a visual coding method, as only 2 levels can be discriminated on an absolute basis under optimum conditions.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that determining the flash rate primarily will be based on experimental data, although some expert judgment might be required.

This figure displays two icons that show the appropriate and inappropriate use of flashing icons. The top icon shows the appropriate use. It is a white rectangle with a black border. In the sign is a silhouette of a snow plow and a snowflake and the text label, "Caution Snow Removal." An arrow points to the icon labeled flashing signals, should be reserved for emergency situations for which the driver's immediate attention is necessary, such as when presenting vehicle safety and warning information. The bottom icon shows the inappropriate use. It is a blue square sign with a black border and a white silhouette of a gas pump. An arrow points to the icon and is labeled use static signals to convey information that the driver does not need immediately, such as motorist services type information.

Figure 4-5. Table with Graphics. Examples of Alternative Methods for Indicating Prohibition. This figure begins with a bar scale. When designing a prohibition symbol, consider the following:

• Care should be taken not to obscure small details of the symbol with the slash or cross. When necessary, the symbol's placement or orientation may need to be changed. Modification of the slash may also be necessary to ensure the display of all important features. See the table below for alternative methods for indicating prohibition.

• Use full-length, solid slashes rather than partial or transparent ones. Slashes that appear in front of or behind the pictorial are preferred.

• Use a standard circle and a 30-degree to 60-degree slash to indicate prohibition. The circle and slash should be colored red for saliency.

• Use pictorials that do not contain too many small details since these are likely to become obscured by the circle or slash, making the pictorial ambiguous.

• Avoid using pictorials that show a negative consequence of an action. With the slash removed, the pictorial should portray a positive course of action.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that determining the use of prohibitive symbols primarily will be based on experimental data, although some expert judgment might be required.

This table shows five alternatives to indicate prohibition. It has two columns, labeled method and example. The first alternative method is changing the slash angle to be 30 to 60 degrees off the vertical line. Its example shows a no-turn sign, which is a black letter "T" with arrows on the ends of the horizontal line. A "do not" circle, a red circle with a slash drawn from the top left to bottom right surrounds the symbol. The angle of the slash is less than 45 degrees. The second alternative method is reversing the direction of the slash. Its example is a no-left-turn sign with the slash going from top right to bottom left, rather than top left to bottom right. The third alternative method is reversing the direction of the symbol inside of the red "do not" circle. Its example is a "no smoking" sign showing a black silhouette of a lighted cigarette. Its lighted tip is on the left rather than the right. The fourth alternative method is displacing the slash and placing it to either the right or left side of the center of the symbol. Its example is a no trucks sign, which is a black silhouette of a truck inside the red "do not" circle with the words "No Trucks" written outside and below the circle. The red slash across the circle is set higher than usual, making it off center. The fifth alternative method is using a cross rather than a slash. Its example is a sign indicating that no one should ride on the forklift. The sign has a black silhouette of a forklift with the driver in the appropriate position, but also with someone standing on the lift and someone sitting on the back. Both the person riding on the lift and the person on the back have red Xs drawn over them.

Figure 4-6. Table with Graphics. Examples of General and Specific Icons for Key In-Vehicle Message Categories. This figure begins with a bar scale. When determining whether general or specific icons should be used, consider the following:

• To minimize driver memory requirements and system complexity, general icons should be used as long as they do not negatively impact driver acceptance or driver performance. Well-designed general icons will be acceptable to most drivers under most driving circumstances.

• The exception to this seems to be safety-related messages such as collision avoidance icons. For safety-related messages, specific icons will provide higher levels of driver acceptance than general icons.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded one-half (three of six sections) of the way from the left to the right side of the scale. This indicates that determining whether general or specific icons should be used will be based equally on experimental data and expert judgment.

This table provides examples of general versus specific icons for 12 key in-vehicle message categories. This table has three columns: the message that the icon is trying to display, general icon examples, and specific icon examples. The first in-vehicle category is if the message is related to trip navigation and is telling drivers to reduce their speed, an example of a general icon is a black exclamation point inside of a white triangle with a black border with the word "Slow" written below it. The background is a yellow square. Three specific icons might be a black silhouette of men at work, which is illustrated as a man shoveling with the word "Slow" written below it; a black silhouette of a car riding on a bumpy road with the word "Slow" written below it; or a black silhouette of a truck entering an intersection with the word "Slow" written below it, all in the background of a yellow square.

The second in-vehicle category is a trip navigation message related to blocked lanes. An example of a general icon is a birds-eye drawing of two lanes separated by a dotted line. A green arrow pointing forward is drawn in the left lane and a red X is drawn in the right lane. The background is orange. Three specific icons include the same two lanes with the green arrow on the left and the red X on the right all inside the background of an orange square. The first drawing is of something blocking the right lane. The second specific icon is a drawing of two cars colliding in the right lane. The third specific icon is a car drawn in the right lane and a red mark above it indicating something is blocking the road.

The third in-vehicle category is a trip navigation message related to closed roads. A general icon is a drawing of two road lanes separated by a dotted line with another road intersecting the right lane noting it is open. The lane for the alternate route has a green arrow drawn in it. Two red Xs, noting lane closure, are drawn in both of the main lanes after the cutoff for the alternate route. The background is an orange square. All examples of three specific icons include the two same main lanes with the Xs, alternate road with a green arrow inside an orange background. The first specific icon is a drawing of something blocking the two lanes. The second specific icon is a drawing of a two-car accident blocking the two lanes. The last specific icon is a drawing of a truck with a skull and crossbones drawn in the two lanes, noting a hazardous vehicle spill.

The fourth in-vehicle icon category is a trip navigation message about en-route emergency vehicles. A general icon example is the black silhouette of a front of a police car. The background is an orange square. Three specific icons could be of the side of a police car; a side of an ambulance; or a side of a fire truck. All four icon emergency vehicles portray flashing red lights and are inside a square icon with an orange background.

The fifth in-vehicle icon category is a trip navigation message about route guidance. A general icon example is a sign with a green background and white text that reads, "next exit" and an arrow pointing diagonally to the right. A specific icon including the same green sign but also has an exit name, such as "Oak Street," written on it to provide more guidance.

The sixth in-vehicle icon category is the in-vehicle collision avoidance message noting an imminent crash. An example of a general icon is a white silhouette of an explosion on a red, square background with the words "Crash Warning" written in white above the silhouette. Three examples of specific icons include white silhouettes of a car with an explosion coming from the rear, the side, or the front. Each of these three icons are also is set on a red, square background with the words "Crash Warning" written at the top.

The seventh in-vehicle icon category message is vehicle condition monitoring of an urgent mechanical problem. An example of a general icon is the warning symbol, which is a black exclamation point drawn inside a white triangle with a black border. Drawn above it is a black wrench. Both drawings are set on a red, square background. An example of a specific in-vehicle icon is the same white warning symbol but with a black car battery showing its positive and negative nodes set on a red, square background.

The eighth in-vehicle icon category message is vehicle condition monitoring of vehicle maintenance requirement. A general icon example is a white wrench on a yellow, square background. An example of a specific icon is a black-and-white drawing of a flat tire on a yellow, square background with the word "Low" written below it.

The ninth in-vehicle icon category message is ATIS motorist services about lodging. A general icon example is a white silhouette of a person lying in bed inside a brown, square background. A specific icon shows the logo for Motel 6; with a red six inside a blue, square background.

The tenth in-vehicle icon category message is ATIS motorist services about food. A general example is a square icon with a brown background and a white fork and knife. A specific icon shows the logo for Burger King; the words "Burger King" in red, sandwiched between two yellow hamburger buns.

The eleventh in-vehicle icon category message is ATIS motorist services about gas. A general example is a white gas pump against a brown, square background. A specific icon shows the logo for British Petroleum (BP), which include the letters "BP" in yellow, inside a green shield with a yellow line border, all inside a green, square background.

The twelfth in-vehicle icon category message is ATIS motorist services about water recreation activities. A general example is a square icon with a brown background with curved lines drawn on it to look like waves. The words "Water Sports" are written above it. Two specific icon examples include square icons with brown backgrounds and white silhouettes of a person swimming or of people canoeing.

Figure 5-1. Table with Graphics. Examples of Icons that Benefit from Text Labels. This figure begins with a bar scale. Text labels should be considered when:

• Icons are abstract and have no conventional or broadly understood meaning.

• The icon represents a message that is particularly important or warns of a particularly hazardous situation.

• The driver can safely spend one to three seconds interpreting the icon.

• The task is complex.

Keep text labels concise (two to three words) and use them sparingly. Text labels reduce the space available for the icon, making them less interpretable.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded two-thirds (four of six sections) of the way from the left to the right side of the scale. This indicates that determining when text labels are needed will be based almost equally on expert judgment and experimental data, although experimental data might be slightly more useful.

This table has three columns to show five examples of how text labels can be helpful. The first column shows the icon without a text label, the second shows the icon with a text label, and the third discusses the added benefit of the label.

In the first example, the first column shows a picture of a black decline drawn inside of a warning triangle with a red border. The second column features the same icon, but with the words "10 percent" written above the incline as the text label to describe the slope of the decline. The added benefit of the label is that the synergy of the icon and the embedded text label quantifies the degree of hazard and clarifies the potentially ambiguous image.

In the second example, the first column shows the front of a black car with a small rectangle on its roof. The second column features the same car with a text label above it reading "Taxi" in bold, black, uppercase letters inside an orange rectangle. The benefit of the text label is that the label embedded in the icon clarifies an ambiguous icon by clearly identifying the type of vehicle.

In the third example, the first column shows a rectangular icon with a brown background. Inside the rectangle, from top to bottom, are a white, uppercase letter P on a square background, a white addition sign, and a black silhouette of a male and female hiking. The second column has the same three icons, but with a text label below it that reads "Trail Parking." The benefit of the text label beneath the icon is that it clarifies the meaning of a complex combination of icons.

In the fourth example, the first column features a white square icon with a drawing of two columns of black cars with their brakes applied. The second column has the same icon, but with a cursor pointing to the cars and the words "Congestion Ahead" written next to the cursor. The benefit of this type of text label is that it is provided only on demand when the cursor placement shows the meaning of the icon. This approach helps people who are unfamiliar with the system without cluttering the monitor for more expert users.

In the fifth example the first column shows a square icon with a gray background and a white telephone handset. The second column has the same icon, but with the words "Cell Phone Inactive" superimposed over the handset. The benefit is that the semitransparent label provides a redundant indicator of system status.

Bar Scale with Line. Design Guidelines. (On page 5-4, associated with table; not numbered as figure.) When composing text labels, consider the following:

• Use explicit icon descriptions. Explicit text labels improve comprehension. They also are perceived as conveying a more hazardous and salient notification than non-explicit labels. In addition, precise descriptions allow the driver to judge when or where a situation will occur and thus aid in response preparation.

• Use non-technical, common vocabulary. Users will not know how to respond to a warning if they do not understand the text verbiage.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded two-third (four of six sections) of the way from the left to the right side of the scale. This indicates that determining the right text label will be based almost equally on expert judgment and experimental data, although experimental data might be slightly more useful.

Figure 5-2. Table with Graphics. Schematic Examples of Conveying Action. This figure begins with a bar scale. To convey a feeling of action within icons, consider the following:

1. Show consequences of action with arrows, speed lines, or the ghosting of images.
2. Use sequence of images to convey simple transformations.
3. Consider animated icons for pre-drive applications to show complex multistep consequences.

Because of the distracting nature of animated icons, they should be used only rarely and with great caution.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded one-half (three of six sections) of the way from the left to the right side of the scale. This indicates that determining how to convey a feeling of action within icons will be based equally on expert judgment and experimental data.

This table shows four schematic examples of icons conveying action. This table has two columns: icon characteristic and examples. To show the icon characteristic of arrows, one example provided is a drawing of a street grid pattern in gray and white with a black arrow drawn through some of the streets where a vehicle has traveled. Another example is a drawing of a large, black arrow, with a white shopping bags drawn inside the arrow.

To show the icon characteristic of speed lines, one example provided is a black-and-white icon of a piece of mail with speed lines of varying length drawn on the left side. Another example is a white outline of an ambulance on a black background with speed lines drawn coming from the back of the ambulance.

To show the icon characteristic of sequencing of images, one example provided is a blue silhouette of two houses, placed side by side, both divided into four quadrants. In the house on the left is a white silhouette of a person in bed in each of the quadrants except for the one in the lower right corner. The house on the right is similar, but a silhouette is drawn in all of the four quadrants as though lodging will become full. A black arrow is drawn between the two houses pointing to the house on the right. Another example is a drawing of two brown file folders, placed side by side. On the left folder, a yellow piece of paper is drawn above the folder along with a red arrow pointing into the folder to look as though the paper is being withdrawn from the folder. The folder on the right appears empty. A black arrow pointing toward the folder on the right is drawn between the two folders.

If the icon characteristic is ghosting, one example is a black silhouette of a telephone with four handsets drawn as fanning out around the base of the phone. Each handset is drawn in an increasingly lighter shade of gray as the handset gets closer to the phone's base. Another example is a side view silhouettes of three cars in a row, each one a darker shade gray from left to right.

Figure 5-3. Table with Graphics. Schematic Examples of Ways to Identify Icons as Part of a Group. This figure begins with a bar scale. When identifying icons as part of a group, consider the following:

• Group icons based on the tasks that the user performs, rather than the architecture of the system.
• Put related icons close together, especially those that are almost the same and those that are opposite.
• If an icon fits into more than one group, duplicate it for each group.
• As a last resort, label groups of icons or individual icons.
• Put groups into separate boxes or windows.
• Put borders or extra space between groups of icons.
• Use a common color for background or icon elements to distinguish groups.
• Within a group of icons, use similar style, level of realism, and level of detail.
• Combine, transform, and include elements from other icons in the same group.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that determining how to group related will be based primarily on experimental data, although some expert judgment might be useful.

This table provides four examples on how to group icons. This table has two columns: a description of the icon group and example icons.

The first description to identify icons as a group, is common boundary, identify as a group of icons conveying warning information. The example shows a group of three triangular warning signs, each with thick, red borders. The first warning sign shows a black silhouette of a person walking in a crosswalk. It warns drivers of this crosswalk. The second icon shows a black silhouette of a rear of a car driving half on the road and half in a rocky area on the side of the road. It warns drivers not to drive too close to the side of the road. The third icon shows two thick, black lines with one merging into the other. It warns drivers about an upcoming lane merge.

The second description to identify icons as part of a group, is icons, identify them by a common level of detail and realism. Three examples are a black silhouette of a snowflake, a white silhouette of a sun on a black, square background, and a white silhouette of an umbrella, on a black, square background.

The third description to identify icons as part of a group is to use a common style. Three examples are a cartoon drawing of a white snowflake with a blue border inside a square icon with a light blue background, a yellow sun on a white background with no border, and a brown umbrella in a square icon with a blue background and a black border. They are drawn with bold lines, are not symmetrical, and have extra detail showing movement; the falling of the snowflake, the rays of the sun, and rain falling on the umbrella.

The final description to identify icons as a group is icons grouped by border, background, color, and style. Three examples of icons including a restaurant icon, which is a blue silhouette of a fork and knife on a white plate surrounded by a blue, square background; a picnic area icon with a white silhouette of a picnic table and pine tree against a blue, square background; and a gas station icon with a white silhouette of a gas pump against a blue, square background. The icons are grouped side by side, and surrounded by a border that looks like a picture frame.

Figure 5-4. Table with Graphics. Examples of Conveying Status. This figure begins with a bar scale. When trying to convey system status with icons, consider the following:

• Use a uniform decrease in contrast or a change of solid to dashed lines to convey changes to on or off or active versus inactive.
• Use an easily recognizable element and project changes upon it to convey system status that is described by multiple categories.
• Represent status changes associated with changes in magnitude by increasing the number or size of well-differentiated icon elements, rather than changing color or contrast.
• Consider using text or numbers to represent changes of magnitude or categories.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded one-half (three of six sections) of the way from the left to the right side of the scale. This indicates that determining how to convey system status will be based equally on expert judgment and experimental data.

This table provides four descriptions and related examples of in-vehicle icons conveying a system's status. This table has two columns, description and examples. The first description reads that an on/off status is indicated by using decreased contrast and gray or dashed lines. It is paired with examples of two icons that are the lodging symbol, which is a white silhouette of a person in a bed on a dark, square background. The icon illustrating that lodging is available has a brown background: the icon illustrating that lodging is unavailable has a gray background.

The second description reads that changes of categorical values are represented by changes projected on an easily recognizable element. One example shows that diesel fuel is available at the next exit by using a square icon with a blue background and a white gas pump with a blue, uppercase letter D in the middle of the pump. Another example shows regular gasoline is available at the next exit by using a similar square icon with a black background and a gas pump, but does not include the letter D on the pump.

The third description reads that changes in magnitude are represented by increasing the number of well-differentiated icon elements. The example shows high versus low speaker volumes by using a round icon with a blue background and a white silhouette of a megaphone. On the megaphone illustrating high volume, more sound wave lines radiate from the megaphone than from the megaphone showing low volume.

The fourth description reads that changes in magnitude are represented by the addition of a numeric label. The example shows that different slope declines by using a warning triangle with a thick, red border and a drawing of black incline. To show a grade of 5 percent, insert a 5-percent text label over the slope. To show a 10-percent grade, insert a 10-percent text label over the slope.

Figure 5-5. Table with Graphics. Schematic Examples of the Use of Color in Icons. This figure begins with a bar scale. When enhancing an icon with color, consider the following:

• Use color coding only when well-established conventions exist, such as temperature, dangerous situations, and permissiveness.

• Use red to indicate highly urgent messages, yellow to indicate cautionary information, and green to indicate normal operations or safe conditions.

• For temperature, use red for hot and blue for cold.

• Arbitrary codes using color to convey meaning are likely to induce errors.

• Color, such as red, can be used to enhance the relative salience of icons. Therefore, color choices should be made with respect to the relationship between messages, such as the relative urgency of messages, and the relationship between messages and the background upon which they are presented.

• Small spots of intense saturated color can convey information effectively. This requires a conservative use of these colors and the use of less saturated colors for backgrounds. See Reference 1 for more information.

• Use shades of gray, rather than color, for showing quantity. See Reference 2 for more information.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded two-thirds (four of six sections) of the way from the left to the right side of the scale. This indicates that determining how to use color to enhance an icon should be based almost equally on experimental data and expert judgment, although the experimental data might be slightly more useful.

This table provides four examples of using color to enhance icon interpretation. This table has two columns. The first column has an explanation of the color enhancement and the second column has the icon example. The first example's explanation states that red enhances the relative salience of this icon and the strong population stereotype indicates a hazard. The icon in the first example is a triangular warning sign with a thick, red border. Inside the triangle are three black cars.

The second example's explanation states that yellow is often used to convey information such as caution or wait. The icon in the second example is a "slippery when wet" sign. The icon is diamond-shaped with black border and a yellow background and a black car drawn inside it. Curved lines on the road drawn near the rear of the car to indicate that the road is slippery.

The third example's explanation states that green is often used to indicate normal operations or to indicate permissible behavior." The icon in the third example shows a parking sign. The icon is a rectangle with a green border and a white background.

The fourth example's explanation states that small spots of intense color greatly enhance the effectiveness of this icon in identifying heavy traffic. The icon in the fourth example is a square with a black border and a white background. Two columns of the rears of cars are drawn in black, except for the break lights, which are red.

Figure 5-6. Table with Graphics. Schematic Examples of Conveying Urgency. This figure begins with a bar scale. To increase the perceived urgency of an icon, consider the following:

• Increase the font size of text labels to identify icons of greater urgency.
• Increase the white space around the label.
• Use red lettering or a red background.
• Increase the line weight of the border.
• Use blinking or flashing to draw attention to the icon.
• Increase the relative size of the high urgency icon.
• Always position urgent warnings within 30 degrees of the operator's normal line of sight.
• Pair the icon with an auditory cue.
• Show consequence of not responding.
• Do not use blue or green coloration, as those convey low urgency.

For time critical situations, such as collision avoidance, icons should be positioned such that they attract the driver's attention to the appropriate part of the vehicle or environment.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded two-thirds (four of six sections) of the way from the left to the right side of the scale. This indicates that determining how to illustrate urgency in an icon primarily will be based on experimental data, although some expert judgment could be useful.

This table provides two examples icons conveying urgency. The table has two columns; one showing the original icon and the other showing how the icon conveys a sense of urgency. The first icon shows a square icon with a thin, black border, a gray background, and a white silhouette of one car crashing into another car. White lines radiate from the point of impact. In the second column, the same icon is shown but it has a red background and a thicker, black border. Three arrows point out how this icon shows a sense of urgency. One arrow points to the two cars and explains that the cars crashing together shows the consequence of not responding to hazards and other cars on the road while driving. Another arrow points out that the red background shows urgency. The third arrow indicates that the increased line weight of the border shows a sense of urgency.

The other icon example shows a black silhouette of a snowplow and snowflake on a white background. The icon has a thin, black border. The icon in the second column has the same two icons but it is slightly larger, with a thicker border. The words "Caution Snow Removal" are written in red beneath the snowplow and snowflake. Arrows point to the red lettering, the thick border, and larger size of the icon, each of which helps convey a sense of urgency.

Figure 5-7. Locations of Icons Within the Visual Field (Adapted from Reference 2). This figure graphically shows that urgent or high-priority symbols should be placed within 15 degrees of either side of the line of sight. Lower priority symbols should be placed outside of the 15-degree mark on either side.

Figure 5-8. Table with Graphics. Schematic Examples of the Use of Shape in Icons. This figure begins with a bar scale. When trying to enhance an icon interpretation using different shapes, consider the following:

• Icons shaped like standard traffic signs convey the respective hazard level of the message.
• Use octagon-shaped icon borders for high hazard conditions or situations.
• Use diamond or inverted triangle-shaped icon borders for medium hazard conditions or situations.
• Use circle, square, or rectangular-shaped icon borders for low hazard conditions or situations.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths (five of six sections) of the way from the left to the right side of the scale. This indicates that determining how to use shapes to enhance an icon primarily will be based on experimental data, although some expert judgment could be useful.

This table shows three examples of how shapes can be used to interpret an icon's hazard level. In the first example, an octagon-shaped icon is used to show a high-level hazards,including dangerous road conditions, vehicle equipment malfunctions, and road obstructions. An example is an octagon-shaped icon with a white background and black silhouettes of a car driving over the edge of land and into water. The icon has a thick, black border.

The second example indicates that a diamond or inverted triangle shape could be used for icons conveying medium-level hazards, such as general road hazards, traffic congestion, or construction warnings. An example would be a yellow diamond-shaped "turn ahead" icon that features a black arrow drawn as a vertical line and then turning right 90 degrees.

The third example indicates that a circle, square, or rectangle could be used for icons conveying low-level hazards including both tourist activities, such as food, lodging, gas signs, and navigation instructions. An example of this type of icon would be a picnic area marked by a rectangular sign with a black border.

Bar Scale with Line. Design Guidelines. (On page 6-2, associated with tables; not numbered as figure.) When augmenting icons with auditory information, consider the following:

• Use the auditory modality for presenting high priority alerts and warnings, with additional contextual information being presented visually.

• Use auditory prompts when a previously static visual display changes.

• Use auditory prompts when high priority information is automatically displayed.

• Use a combination of visual and auditory for repeating low complexity messages.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded four-sixths (four of six sections) of the way from the left to the right side of the scale. This indicates that determining how to augment icons with auditory information will be based almost equally on expert judgment and experimental data, although the experimental data might prove slightly more useful.

Bar Scale with Line.Design Guidelines. (On page 6-4, associated with table; not numbered as figure.) When determining the appropriate auditory signal, consider the following:

• Use simple tones and auditory icons when an immediate response is required.
• Earcons should be used when it is important for the driver to know pieces of information are related.
• Auditory icons are effective for use in collision warning applications (i.e., horn or skidding tires).
• Use speech messages when a high degree of message flexibility is required.
• Use speech messages when a high degree of message detail is required.
• Use speech messages when the meaning of tones or other sounds may be forgotten under stress.
• Use speech messages when the auditory message deals with a future point in time for which there must be some preparation, such as time or distance to turn.
• Speech message displays should not be used for time critical tasks.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded one-half (three of six sections) of the way from the left to the right side of the scale. This indicates that determining the appropriate auditory signal will be based equally on expert judgment and experimental data.

Bar Scale with Line.Design Guidelines. (On page 6-6, associated with table; not numbered as figure.) When designing simple tones to augment the visual presentation of a message, consider the following:

• Appropriate loudness levels are 15 to 25 decibels above the predicted masked threshold.

• Auditory warning signals should be less than 30 decibels above the masked threshold to minimize operator annoyance and the disruption of communication.

• The pitch of warning sounds should be between 150 and 1000 hertz.

• Avoid continuous tones because they are usually high pitched and aversive, they prevent communication if they are loud, and they are easy to habituate because they never change.

When more than one tone is used:

• Avoid tones with the same on/off ratio.
• Avoid tones that share the same temporal pattern.
• Avoid tones that begin in the same way, such as with a long tone.
• No more than 6 simple tones should be used.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded two-thirds (four of six sections) of the way from the left to the right side of the scale. This indicates that determining the use of simple tones will be based almost equally on expert judgment and experimental data, although the experimental data might prove slightly more useful.

Figure 6-1. Graphic. Temporal Parameters of Auditory Signals-Pulse, Burst, and Sound Parameters Defined Graphically. This figure begins with a bar scale. When designing complex tones, consider the following:

• The amplitude envelope of the initial pulse should include a 20-millisecond onset to reduce startle effects.

• The pulse should be composed of multiple frequency components, such as formants or harmonics, to mitigate masking due to background noise.

• The temporal pattern of auditory signals should be as distinct as possible, otherwise confusion is likely even it the spectral content is substantially different.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded two-thirds (four of six sections) of the way from the left to the right side of the scale. This indicates that determining the use of complex tones will be based almost equally on expert judgment and experimental data, although the experimental data might prove slightly more useful.

This figure illustrates how four different signals could sound. First, the high-density, low-speed signal is illustrated with four thick, black columns, all of the same height, drawn close together above a horizontal base line representing elapsed time. The space between each column is equal. A low-density, low-speed signal is illustrated using the same base line, but with four thinner black columns, all of equal height, spaced further apart with equal space between each column. A high-density, high-speed signal is illustrated by having 11 columns, each spaced apart equally above a base line. A gradual onset, gradual offset signal also is represented by 11 columns. The column height at the ends of the line are shorter than the ones in the center. The column height gradually increases in size from the left side to the center of the line and then decreases in size from the center to the right end. The figure also shows how a set of pulses composes a burst, and set of bursts composes an auditory signal. On a time line representing 100 milliseconds, 4 acoustic waveforms are shown, which make up one pulse. On the next time line, which represents 500 milliseconds, 4 sets of pulses are shown, which make up one burst. One the third time line, which represents 2,000 milliseconds, 5 sets of bursts are shown, which make up the auditory signal.

Bar Scale with Line. Design Guidelines. (On page 6-10, associated with table; not numbered as figure.) When designing earcons, consider the following:

• Use synthesized musical timbres that are subjectively easy to tell apart, such as organs and brass timbres.
• Do not use pitch alone to distinguish between tones unless there are very significant differences. Some suggested ranges for pitch are a maximum of 5 kilohertz and minimum of 125 hertz to 150 hertz.
• Use tones that are three or more octaves apart.
• Make the rhythm as different as possible. Putting a different number of notes in each rhythm is effective.
• Some suggested ranges for intensity are a maximum of 20 decibels above threshold and a minimum of 10 decibels above threshold.
• When playing combinations of multiple earcons, a gap of 0.1 seconds should be between them so that the user can tell where one finishes and another starts.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded one-third (two of six sections) of the way from the left to the right side of the scale. This indicates that determining the design of earcons will be based almost equally on expert judgment and experimental data, although the expert judgment could prove slightly more useful.

Figure 6-2. Bar Graph. Brake Reaction Times for Different Warning Sounds (from Reference 2). This figure begins with a bar scale. When designing auditory icons, consider the following:

• By definition, auditory icons must be identifiable as having relevance or conveying some inherent meaning. See above for examples of iconic, metaphorical, and symbolic auditory icons.
• Auditory icons should be detectable: 10 to 20 decibels above the masked threshold.
• No more than six auditory icons should be used in an auditory icon set.
• Auditory icons should strive to attract the attention of the driver without generating a startle reaction. Special attention should be paid to the perceived urgency associated with different candidate auditory icons.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded one-half (three of six sections) of the way from the left to the right side of the scale. This indicates that determining the use of auditory icons will be based equally on expert judgment and experimental data.

This graph shows the mean brake reaction time for four different warning sounds; horns, tire skids, speech, and tones. Warning sounds are placed on the horizontal axis, and mean braking reaction time is recorded on the horizontal axis. Mean reaction times are displayed for warning sounds that indicate both when a stationary vehicle is ahead and when a vehicle is pulling out from a side road. The mean braking time when a horn is heard is approximately 0.725 seconds when a stationary vehicle is ahead and 0.75 seconds when a vehicle is pulling out from a side road. The mean braking time when a tire skid is heard is approximately 0.775 when a stationary vehicle is ahead and 0.74 when a vehicle is pulling out from a side road. The mean braking time when speech is used as a warning sound is approximately 0.775 when a stationary vehicle is ahead and 0.90 when a vehicle is pulling out from a side road. The mean braking time when a tone is used as a warning sound is approximately 0.825 when a stationary vehicle is ahead and 0.80 when a vehicle is pulling out from a side road.

Bar Scale with Line. Design Guidelines. (On page 6-14, associated with tables; not numbered as figure.) When designing speech messages, consider the following:

• If speech must be used in a time-critical application, such as a warning, the message should be kept to a single word or a short phrase with the fewest number of syllables possible.
• Messages that are not urgent or for which a response may be delayed can be a maximum of seven units of information in the fewest number of words possible. If the information cannot be presented in a short sentence, the most important information should be presented at the beginning and/or the end of the message.
• Navigation instructions should be limited to three or four information units, such as accident ahead, merge right or turn right in half of a mile.
• Do not try to make the voice sound too human. A machine should have a machine voice in order to cue its identity when it speaks.
• Provide a means for repeating speech messages.
• Provide a redundant visual presentation of the information being presented aurally.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that determining how to use speech messages primarily will be based on experimental data, although expert judgment might also prove useful.

Figure 6-3. Graph. Example of Using Steven's Power Law for Producing Urgency Exponents (see references 2 and 3 for more detailed explanations and examples). This figure begins with a bar scale. When trying to increase the perceived urgency of an auditory signal, consider the following:

• Use faster auditory signals.
• Use regular rhythms.
• Use a greater number of units (4).
• Use auditory signals that speed up.
• Use high fundamental frequencies.
• Use a large pitch range.
• Use a random pitch contour.
• Use an atonal musical structure.

When trying to decrease the perceived auditory signal,consider the following:

• Use slower auditory signals.
• Use irregular rhythms.
• Use fewer number of units (1).
• Use auditory signals that slow down.
• Use low fundamental frequencies.
• Use a small pitch range.
• Use a down or up pitch contour.
• Use a resolved musical structure.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that determining how to change the perceived urgency of the auditory signal primarily will be based on experimental data, although some expert judgment might prove useful.

The graph shows an X and Y axis. The X axis represents the base 10 logarithm of the objective value of a sound parameter. The range of the X axis is 2.6 to 3.2. The Y axis represents the base 10 log of the subjective value of the perceived urgency rating. Note that the range of the Y axis is 1.5 to 1.85. On the graph, a line of best fit is drawn. The intercept of the best fit line with the Y axis is known as K. M is equal to the slope of the line of best fit. Marks on the graph are found at the following X, Y coordinates: approximately 2.65 and 1.56, 2.87 and 1.63, 2.88 and 1.66, 2.93 and 1.67, 3.0 and 1.76, and 3.05 and 1.75.

Equation 8. Steven's Power Law. S equals K multiplied by O to the power of M, where S is the subjective value of perceived urgency rating, K equals the intercept of the line of best fit, O is the objective value of a sound parameter, and M is the slope of the line of best fit.

Bar Scale with Line. Design Guidelines. (On page 6-18, associated with table; not numbered as figure.) When using automatic speech recognition (ASR) devices consider the following:

• For IVIS applications, ASR devices should be used to aid complex tasks that involve high visual, cognitive, or manual requirements.

• Vocabulary sets for ASR devices should reflect natural language conventions as much as possible, avoid similar-sounding words or phrases, and be small enough so that drivers can recall command words rapidly and with few or no errors.

• The microphone for an ASR device should be located on the forward portion of the vehicle headliner, right in front of the driver (see Reference 3).

• Drivers should be provided with immediate feedback, such as error correction and input confirmation, of the recognition results or the system's response to the speech input. Changes in the visual display itself provide a good form of feedback, but require driver head or eye movements to verify. Although any feedback will improve the driver's performance with the system, size limits on IVIS displays in the in-vehicle environment, as well as concerns about visual overload, suggest that auditory feedback should be used.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded two-thirds (four of six sections) of the way from the left to the right side of the scale. This indicates that determining use of ASR devices will be based almost equally on expert judgment and experimental data, although the experimental data might prove slightly more useful.

Figure 6-4. Figure. Equations for Determining the Appropriate Timing of an Instruction. This figure begins with a bar scale. When determining the timing of auditory navigation information, consider the following:

• For maneuvers defined as leaving the current route, such as turning onto a side road, the timing of the auditory guidance instruction can be based on the equations provided below.
• It may be advisable to implement the equation for preferred maximum distance. An instruction given slightly too early is preferable to one given too late.
• When the distance between two subsequent maneuvers is less than the minimum preferred distance for that speed, the instructions are stacked or given during a single message.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded one-half (three of six sections) of the way from the left to the right side of the scale. This indicates that determining the timing of auditory navigation signals will be based equally on expert judgment and experimental data.

This figure illustrates an example of the suggested timing for navigational information. In the figure, a car is shown approaching an intersection. A dotted line and arrow indicates that at the intersection, the car will turn right. Text around the diagram indicates that the vehicle is 150 meters from the intersection and traveling at 48 kilometers per hour. The preferred minimum distance for the navigational information is 93 meters from the intersection. The ideal distance is 116 meters from the intersection and the preferred maximum distance is 144 meters from the intersection. Equations 9, 10, and 11 are imbedded in the figure.

Equation 9. Preferred Minimum Distance. The preferred minimum distance for the appropriate timing of an instruction is equal to 14.799 added to the sum total of the speed multiplied by 1.637. Speed is in kilometers per hour and distance is in meters.

Equation 10. Ideal Distance. The ideal distance for the appropriate timing of an instruction is equal to 21.307 added to the sum total of the speed multiplied by 1.973. Speed is in kilometers per hour and distance is in meters.

Equation 11. Preferred Maximum Distance. The preferred maximum distance for the appropriate timing of an instruction is equal 37.144 added to the sum total of the speed multiplied by 2.222. Speed is in kilometers per hour and distance is in meters.

Figure 6-5. Graphs. Appropriateness Depends on Perceived Annoyance for Benign Situations (such as E-mail Notification), Whereas Appropriateness Depends on Perceived Urgency for Highly Critical Situations (such as Collision Avoidance). This figure begins with a bar scale. When determining the perceived annoyance with auditory signals, consider the following:

• For signals to be perceived as appropriate, highly urgent sounds should be used for highly critical situations.
• For signals to be perceived as appropriate, low annoyance sounds should be used for benign situations.
• Sound characteristics of pulse duration, burst density, sound type and speed all increase perceived urgency more than perceived annoyance.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded one-half (three of six sections) of the way from the left to the right side of the scale. This indicates that determining the perceived annoyance of auditory signals will be based equally on expert judgment and experimental data.

This figure has two graphs to demonstrate the appropriateness of the annoyance of in-vehicle auditory signals for two different scenarios. These graphs visually display the annoyance trade-offs between an annoying signal noise and a situation's level of urgency. The X-axis of the two graphs represents related level of urgency/annoyance. The range of the X-axis is 30 to 80. The Y-axis represents related level of appropriateness. The Y-axis has a range of 20 to 90. In each of the two graphs there are two lines of best fit: all four lines of best fit have 15 coordinates each. The two lines of best fit for urgency appropriateness are solid: the two lines of best fit of annoyance appropriateness are dashed. For the e-mail scenario, the lines of best fit show that an e-mail auditory must sound less annoying as the situation is benign and does not require a driver's immediate response. For the collision avoidance scenario, the lines of best fit show that the collision avoidance auditory signal can be much more annoying to trigger the driver's immediate reaction to an urgent situation.

Figure 7-1. Flowchart. Overview of Procedures for Evaluating In-Vehicle Icons. The following procedures can be used to evaluate in-vehicle icons:

Are candidate icons developed and or in use (see chapter 8)?If yes, follow these steps:

1.   Conduct analyses to determine that the icons conform to design principles and guidelines for legibility (see chapter 3), recognizability (see chapter 4), and interpretability (see chapter 5).

2.  Conduct appropriateness ranking tests (see page 7-6). This is done by reproducing production test icons, randomizing presentation order to test icons, rank ordering the test icons, calculating the scale value of the icons, and selecting candidates for further study.

3.  Conduct comprehension and recognition tests (see page 7-8). This is done by preparing the test materials, providing subjects with instructions, testing candidate symbols with a representative group of subjects, analyzing the data, and interpreting the results.

4.   Determine if there is a poor response to the icons. If there is a poor response, use production tests to generate ideas for icons, by identifying messages for the icons, generating candidate symbols and icons, evaluating candidate symbols and icons, and identifying candidate icons for further testing. Proceed again with steps 2 and 3, and then reassess whether there is a poor response to the icons. Repeat this process until the response is no longer poor.

If the response to the icons is not poor, testing may be complete; proceed as necessary to step 5.

5.   Conduct matching tests (see page 7-10). This is done by preparing test materials, providing instructions to the subjects, conducting the test, analyzing the data, and interpreting the results.

6.  After step 5, testing may be complete. As necessary, redesign the icons and repeat the evaluations. Consider alternate approaches to evaluations (see page 7-12).

If candidate icons have not been developed or are not in use, use production tests to generate ideas for icons (see page 7-4) by identifying messages for the icons, generating candidate symbols or icons, evaluating candidate icons or symbols, and identifying candidate icons for further testing. Following the production tests, perform the rest of the steps, starting with step 1.

This figure ends with a bar scale. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that the evaluation of icons will be based primarily on experimental data, although some expert judgment might be useful.

Figure 7-2.Flowchart. Production Test. The following steps should be taken to conduct a production test: identify messages for the icons, generate candidate symbols and icons, evaluating the candidate symbols and icons, and identify candidate icons for further testing. When identifying messages for the icons, select messages (functions, conditions, etcetera) for which an in-vehicle icon is needed and no standard icons or symbols are available. When generating candidate symbols or icons, ask groups of designers, as well as representative drivers, to generate candidate icons or symbols or verbal descriptions of icon requirements. Key goals of this step include using a wide and diverse range of subjects in the test; identifying meaningful, simple, and understandable icons for the messages; and asking subjects to consider typical and comprehensible associations and metaphors that come to mind when considering the message. When evaluating candidate symbols or icons, key questions that can be used to guide the analysis include whether there were any icons or elements or features within the icons that were consistently produced; if different subject groups, such as design engineers versus real-world drivers, produced consistently different images for the same message; if these different images can be combined or reconciled in some way; and if the verbal descriptions generated by the subjects can be used to generate new candidate icons. When identifying candidate icons for further testing, use the icons and verbal descriptions that were produced earlier, select a broad range of ideas, and keep in mind that drawings or ideas may be categorized into different categories based on consistencies in image content, and that one or two images can be selected or generated for additional tests from each category of similar images. It is important to note that this production test will not result in a final icon selection. It is used to generate candidate symbols or icons only.

The figure ends with a bar scale. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that conducting a production test and generating candidate icons and symbols will be based on experimental data, although some expert judgment might be useful.

Figure 7-3. Flowchart. Appropriateness Ranking Test. The following steps should be taken to conduct an appropriateness ranking test: reproduce production test icons, randomize the presentation order of the test icons, rank order the test icons, calculate scale values, and select candidates for further study. When reproducing production test icons, produce a set of cards for each message being considered. Each card should show one of the candidate icons developed during the production test or identified using other means. When randomizing the presentation order of the test icons, randomize the set of cards containing the candidate icons prior to testing for each subject. If multiple messages are being tested and there is more than one set of cards, the order in which subjects rank order each set of cards should be counterbalanced across subjects. When rank order testing the icons, ask representative subjects to sort the cards within each card set according to the degree to which each candidate icon is perceived to be an appropriate one for the message under consideration. The cards can be sorted by placing the card with the most appropriate icon first in the set and placing the least appropriate icon last in the set. Multiple card sets can be used to test a group of subjects at the same time. When calculating scale values using an approach such as Torgerson's Categorical Scaling Procedure (Reference 3), calculate scale values for each of the candidate icons. Tutorial 1 provides a step-by-step description of the calculations required to calculate scale values from rank orders. This provides not only a general rank order, but also some indication as to how much the candidate symbols differ along an interval scale. When selecting candidates for further study, using the interval scale values, select approximately three candidates for each message for further study. Use the following criterion for selecting these candidates: "When differences between scale values are minimal, symbol variants with different image content are chosen, ignoring graphic detail." Thus, the top three candidates are not necessarily the ones chosen for further study. Other selection issues include the relative differences between the scale values and the uniqueness across the candidates. Thus, in a close field of candidates with smaller differences in scale values, those candidates with acceptable appropriateness ranks and some heterogeneity, in terms of icon symbol features, might be selected. The appropriateness ranking test will not result in a final icon selection. It is used to provide a preliminary evaluation of candidate symbols/icons only.

This figure ends with a bar scale. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that the appropriateness ranking test to screen the candidate symbols generated during the production tests will be based on experimental data, although some expert judgment might be useful.

Figure 7-4.   Flowchart. Comprehension/Recognition Test. The following steps should be taken to perform a comprehension or recognition test: prepare the test materials, provide subjects with instructions, test candidate symbols with a representative group of subjects, analyze the data, and interpret the results. When preparing test materials, place candidate symbols on separate sheets of paper, slides, or computer screens; randomize presentation order across subjects; separate different candidates for the same message into distinct test sets; prepare and provide subjects an example sheet with a common icon like a fuel pump to indicate a fuel gage and its meaning written beneath the graphic. When providing subjects with instruction, indicate the context in which the icon will be used, either verbally or co-located with the icon. Subjects then write down the action, condition, activity, and location associated with the icon. When testing candidate symbols with a representative group of subjects, present test subjects with candidate icons or symbols and asks them to write down the action, condition, activity, and location that they believe the icon or symbol represents. When analyzing the data, have a panel of judges independently categorize responses along a scale according to well-defined criteria that identify the likelihood that an individual response indicates correct comprehension of the icon. That is, the perceived meaning should be compared to the intended meaning. For example, use an 8-category rating scheme with a 1 meaning that the response matches the intended meaning exactly; a 2 meaning that the response captures key informational elements of the intended meaning, but missing one or more minor elements; a 3 means that the response captures some aspects of the intended meaning, but is missing one or more key elements; a 4 means that the response does not match intended meaning, but captures one or more key informational elements; a 5 means that the response does not match intended meaning, but is somewhat relevant; a 6 means that the subject's response is in no way relevant to the intended meaning; a 7 means that the subject indicates no understanding of the icon; and an 8 means no answer. For each candidate, convert the total number of responses in each category into percentages. When interpreting the results, keep in mind that decisions regarding minimum percent correct rates for individual icons should reflect designer's needs, as well as the consequences associated with selecting a cutoff that is too high or too low. Reference 2 notes that ISO requirements for an acceptable symbol have, in the past, been a minimum 67 percent correct comprehension level that combines categories 1 and 2 from above.

This figure ends with a bar scale. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that the comprehension or recognition test, which is used to determine which of a number of candidate icons or symbols designed to illustrate a concept is understood best, will be based on experimental data, although some expert judgment might be needed.

Figure 7-5.Flowchart. Matching Test. The following steps should be taken to perform a matching test: prepare the test materials, provide instructions to subjects, conduct the test, analyze the data, and interpret the results. When preparing test materials, select or create a symbol for each message within a message set based on the results of a comprehension or recognition or comparable test. Place the complete set of symbols for a function or group of related functions in a matrix on a sheet of paper, slide, or computer screen. Include 4 to 8 symbols in each matrix. The medium selected for information presentation, such as paper, slide, or computer screen, should be consistent with the expected in-vehicle conditions with respect to parameters such as resolution, color, luminance, contrast, and size. Repeat for additional function or groups of icons and randomize presentation order of symbol sets across subjects. When providing instructions to subjects, tell subjects the context in which the symbol set will be used, being careful that such information does not create any bias. The subjects' task will be to select the icon that best represents a particular driver message. When conducting a test, give the subject the meaning or driver message associated with one of the icons in the symbol set and ask them to indicate which of the icons in the matrix represents that meaning. Repeat test procedure with additional symbol sets. To maintain independent choices, individual subjects can be tested only on one icon per symbol set. Thus, while individual tests require little time, many subjects are required. When analyzing the data, calculate the percentage of correct responses for each symbol or message combination. Construct a confusion matrix with icons in the set as columns and messages as rows, data in the cells should reflect the frequency with which an individual symbol was matched with a message. When interpreting the results, use the percentage correct for each icon, as well as the confusions across icons, to develop ideas for new icons or for the redesign of easily confusable icons.

This figure ends with a bar scale. The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded five-sixths of the way from the left to the right side of the scale. This indicates that the matching test, which is used to determine how well a symbol will work within a set and whether the many symbols within a set can be discriminated from one another without confusion, will be based on experimental data, although some expert judgment might be useful.

Bar Scale with Line. Design Guidelines. (On page 7-12, associated with table; not numbered as figure.) The design guidelines indicate that the evaluation technique selected to assess the effectiveness of candidate icons should be consistent with the goals and constraints of a particular system development effort. The table below the bar scale summarizes a range of alternative evaluation approaches, gives the advantages and disadvantages associated with each, and provides references for designers to go to for more detailed information.

The left side of this scale reads "Based Primarily on Expert Judgment," the middle of the scale reads "Based Equally on Expert Judgment and Experimental Data," and the right side of the scale reads "Based Primarily on Experimental Data." The scale is divided into six sections of equal size. The line is shaded all of the way from the left to the right side of the scale. This indicates that determining which additional evaluation techniques to use will be based primarily on experimental data.

Note: Page numbers are provided throughout chapter 8 for identification purposes, because these icons are not numbered as figures.

Page 8-2

Icon. This is a lodging icon and is identified by number 382. The icon is a square icon with a brown background and a white silhouette of a person lying in a bed.

Page 8-3; Augmented Signage; Roadway Notification Sign Information; General

Icon. This is an added lane sign icon and is identified by number 119. The icon is yellow and diamond-shaped. In the diamond, two arrows are drawn side by side. The arrow on the right is pointing straight ahead. The arrow on the left curves to indicate that the lane on the left is being added next to the right lane. The line separating the two lanes is solid before the lane is added, and then is dotted after the lane is added.

Icon.This is the divided highway (road) icon and is identified by number 264. The icon is diamond-shaped with a white background and has two curved, black vertical arrows pointing in opposite directions separated by a wedge shape.

Icon. This is a drawbridge icon and is identified by number 96. The icon is triangle-shaped with a black silhouette of a drawbridge opening on a white background. The triangle has a thick, black border.

Icon. This is a lane reduction transition icon and identified by number 354. The icon is a yellow diamond with a black border. Two black lines are drawn vertically down the middle. The line on the left is straight, while the line on the right starts set apart from the other line and then curves to merge closer to the other line.

Icon. This is a merge sign icon and identified by number 262. The icon is a yellow diamond with a black border. A black vertical arrow pointing upwards is drawn in the center. The tail of the arrow is split in two so that one part of it fans to the right.

Icon. This is a pedestrian crossing ahead icon and is identified by number 271. The square icon has a brown background with a white border and a white silhouette of a man walking in a crosswalk.

Icon. This is a percent of grade icon and is identified by number 65. It is a triangle-shaped with a thick, red border around the triangle. The background inside the triangle is white. In the triangle is a black silhouette of a hill with the words "10 percent" written on the slope of the hill.

Icon. This is a railroad crossing icon and is identified by number 366. The icon is a circle and has a black border and a yellow background. Two black lines cross diagonally in the center of the circle. A black letter R is written on both side of the lines.

Icon. This is the sharp curve ahead sign and is identified by number 42. The rectangle-shaped icon has a black border and a yellow background and a thick, black arrow pointing to the right drawn in the center.

Icon. This is a steep downgrade icon and is identified by number 14. The diamond-shaped icon has a black border and yellow background. A black silhouette of a truck travels down a hill. Below the silhouette of the truck and hill are the words "6 percent."

Page 8-4; Augmented Signage; Roadway Notification Sign Information; General

Icon. This is a tight ramp or intersection icon and is identified by number 183. The diamond-shaped icon has a black border and a yellow background and a black silhouette of the back of a truck, which looks like it is tipping over. Above the truck, a curved arrow pointing to the right is drawn to indicate that the road curves sharply to the right. Below the truck, the number 35 indicates the maximum speed limit and that a truck going over 35 miles per hour is in danger of tipping over.

Icon. This is the turn signs icon and is identified by number 351. The diamond-shaped icon has a yellow background and a black arrow drawn curving to the right.

Icon. This is the winding road ahead icon and is identified by number 139. The diamond-shaped icon has a black border and a yellow background. A black arrow curves several times to illustrate a winding road.

Augmented Signage; Roadway Regulatory Sign Information; General

Icon. This is a no trucks icon and is identified by number 104. The icon is square with a white background. The icon has a black "do not" circle symbol, which is a circle with a slash cutting through it at 45 degrees from top left to bottom right. Inside the circle is a black silhouette of a truck.

Icon. This is the rock spray icon and is identified by number 87. The triangle-shaped icon has a thick, red border, a white background, and a black car with rocks spraying to the side from its tires.

Icon. This is a speed limit 50 icon and is identified by number 414. The icon is a rectangle with a black border and a white background. The words "Speed Limit 50" is written in black, uppercase letters. Text under the box reads "Do not exceed posted speed."

Icon. This is a stop ahead icon and is identified by number 268. The diamond-shaped icon has a black border and a yellow background with a red octagon drawn in the center to represent a stop sign. Above the octagon is a black arrow indicating a stop sign is ahead.

Icon. This is the yield icon and is identified by number 228. The icon is an inverted triangle with a thick, black border and the word "Yield" written in black, uppercase letters on a white background.

Icon. This is a yield ahead icon and is identified by number 176. This diamond-shaped icon has a black border and a yellow background. Drawn inside is an inverted white triangle with a red border. Above the triangle is a black arrow pointing ahead.

Page 8-5; Automated/Adaptive Cruise Control Devices

Icon. This is an adaptive cruise control engaged icon and is identified by number 195. This square icon has a white background and a black silhouette of a car. Below the car an arrow points to a speedometer.

Icon. This is an adaptive cruise control engaged icon and is identified by number 216. This rectangular icon with a black background is composed of six rows. In each row is a picture of two small, turquoise cars, each facing to the right. Between the cars are spacing bars noting how much space is between them. In the top row, the two cars are spaced close together. In each subsequent row, the cars are spaced further and further apart. When this icon is illuminated, it warns a driver that he/she needs to increase the travelling space with the vehicle ahead.

Icon. This is an adaptive cruise control engaged icon and is identified by number 31. This small rectangle-shaped icon has a black background. It shows two small, turquoise cars side by side and facing toward the right. In the space between the cars are six, small gray boxes. When this icon is illuminated, it warns a driver that he/she needs to increase the travelling space with the vehicle ahead.

Icon. This is an adaptive cruise control failure icon and is identified by number 8. This square icon has a white background and a black car. Below the car is a drawing of a speedometer. Beside both the car and speedometer is a large, black exclamation point.

Icon. This is an adaptive cruise control not engaged icon and is identified by number 114. This rectangle-shaped icon is composed of six rows. In each row is a picture of two small, turquoise cars, each facing to the right. In the top row, the two cars are spaced close together. In each subsequent row, the cars are spaced further and further apart. In the space between the cars is a series of turquoise radar wave lines indicating that the space is growing from row to row, with fewer lines appearing in the top rows than in the bottom rows. The icon has a black background.

Icon. This is an adaptive cruise control not engaged icon and is identified by number 393. This rectangle-shaped icon features two small, turquoise cars side by side facing toward the right. In the space between the cars are six turquoise radar wave lines. The icon has a white background; the symbols are in a black rectangular bar.

Icon. This is a brake service required icon and is identified by number 206. The rectangle shaped icon has a dark red background and is an overheated drum brake with its brake pads applied. Above the brake element is a warning symbol; a white triangle with a black exclamation point in it.

Icon. This is a brake service required icon and is identified by number 68. The rectangle shaped icon has a dark yellow background. It shows a drum brake with applied break pads. The brake padding is stripping off and falling below.

Icon. This is an imminent collision warning icon and is identified by number 341. The icon is a square with a dark red background. The words "Crash Warning" are written inside the square in white, uppercase letters along with a silhouette drawing of a star shaped explosion.

Page 8-6; Collision Avoidance Information

Icon. This is an imminent collision warning icon and is identified by number 281. The icon is a rectangle covered with a black and red stripes drawn diagonally across it.

Icon. This is an imminent collision icon and is identified by number 399. The icon is a horizontal bar with a red background with thin, black stripes drawn at equal distances vertically across it.

Icon. This is an imminent collision warning icon and is identified by number 332. The icon is a vertical column with a red background with thin, black stripes drawn at equal distances horizontally inside it.

Icon. This is an imminent collision warning icon and is identified by number 53. The icon is a black square with a large inverted, isosceles, red triangle in it.

Icon. This is an imminent collision warning icon and is identified by number 172. The icon is a black square with a large, red circle in it. The circle's circumference touches all four sides of the square.

Icon. This is a traffic collision icon and is identified by number 303. The icon is square with a white background. The icon shows a picture of two gray cars, one facing front and one approaching from the side. The car approaching from the side has collided into the other car. A spray of red marks is drawn at the collision point.

Icon. This is a forward crash vehicle warning icon and is identified by number 85. The icon is square with a red background. The words "Crash Warning" appear in the top of the square in white, uppercase letters. A silhouette drawing of a vehicle and impact marks on the front of the vehicle represent the possible crash.

Icon. This is the forward crash vehicle warning icon and is identified by number 326. The icon is square with a white background. This icon shows a black car facing left. A star-shape at the car's front represents the impact of the car crashing into something.

Icon. This is the forward crash vehicle warning icon and is identified by number 162. The icon is square with a white background. This icon is a black silhouette of the front two thirds of a car, with a star shape drawn at the front to represent the impact of the car crashing into something.

Page 8-7; Collision Avoidance Information; Forward Collision Avoidance

Icon. This is a forward crash vehicle warning icon and is identified by number 293. The icon is square with a white background. This icon shows the front half of a black car. A star-shape drawn at the car's front represents the impact of the car crashing into something.

Icon. This is the icon for a forward crash vehicle warning system failure and is identified by number 260. The icon is square with a white background. This icon shows a black silhouette of a side-facing car. A star-shape at the car's front represents the impact of the car crashing into something. Overlaying the car silhouette is a black exclamation point.

Icon. This is an icon for a forward crash vehicle warning system failure and is identified by number 373. The icon is square with a white background. This icon shows two-thirds of a side-facing black car. A star shape at the car's front to represents the impact of the car crashing into something. A black exclamation point is between the car and the star shape.

Icon. This is an icon for a forward crash vehicle warning system failure and is identified by number 59. The icon is square with a white background. This icon shows the front half of a side-facing black car. A star shape at the car's front represents the impact of the car crashing into something. To the right half of the car is a black exclamation point.

Collision Avoidance Information; Rear-End Collision Avoidance

Icon. This is an alert level icon and is identified by number 363. The icon has four black squares arranged vertically in a column. In the top box is a small, green silhouette of a car. In the second box from the top is a yellow silhouette of a car, which is larger than the car in the top box. The third box from the top has a red silhouette of a car, which is larger than the car in the second box. In the fourth and bottom box, there is the largest silhouette of a car that is also red. At the front of this car is a yellow star shape that represents a collision. These cars symbolize the distance of a vehicle approaching from the rear.

Icon. This is an imminent collision warning icon and is identified by number 385. The square icon has a crimson background with white, uppercase letters that read "Crash Warning." Below this is a view of a white car facing vertically. A star shape is drawn at the bottom of the car warning the driver of a rear-end collision.

Icon. This is an imminent collision warning icon and is identified by number 44. This square icon has red background and an overhead view of two white cars involved in a rear-end collision.

Icon. This is a lane departure warning icon and is identified by number 386. This square icon has a white background and shows three vertical lines converging at some point in the distance. The two outer lines are solid black and the middle line is dotted. The lines represent a highway or other roadway.

Page 8-8; Collision Avoidance Information; Side Collision Avoidance

Icon. This is an imminent collision warning icon and is identified by number 296. This square icon has a black background, an overhead view of a white car, and a red star shaped drawing to the left of the car to represent a collision on the car's left side.

Icon. This is an imminent collision warning icon and is identified by number 106. This square icon has a black background, an overhead view of a white car, and a red star shaped drawing to the right of the car to represent a collision on the car's right side.

Icon. This is an imminent collision warning icon and is identified by number 37. The square icon has a crimson background and white, uppercase letters that read "Crash Warning." Below is an overhead view of a white car facing vertically. A star shaped mark on the right side of the car to represents a side collision.

Icon. This is an imminent collision warning icon and is identified by number 116. This rectangular icon has a black background and an overhead view of a white car on the right just ahead of a red car on the left. The icon warns the driver of the white car that the red car is in his or her vehicle blind spot.

Icon. This is an imminent collision warning icon and is identified by number 349. This rectangular icon has a black background and an overhead view of a white car on the left just ahead of a red car on the right. The icon warns the driver of the white car that the red car is in his or her vehicle blind spot.

Icon. This is an imminent collision warning icon and is identified by number 416. This rectangle icon has a black background. On the top two-thirds of the rectangle are red diagonal stripes. Below the stripes, "Caution Right Side" is written in white letters.

Icon. This is a no left turn icon and is identified by number 256. The square icon has a black background. A red "do not" circle covers a green arrow centered inside the circle. The arrow points to the left.

Icon. This is a no right turn icon and is identified by number 275. The square icon has a red "do not" circle with a green arrow pointing to the right inside the circle.

Icon. This is a side obstacle warning icon and is identified by number 55. The icon is square and has a white background. The icon shows an overhead view of a white car with a black triangle radiating from the left side of the car where the side view mirror is located. The triangle represents that the car's side radar has detected an object in the driver's blind spot.

Icon. This is a side obstacle warning icon and is identified by number 317. The icon is square and has a white background. The icon shows an overhead view of a white car and white-striped triangle radiating from the left side of the car where the side view mirror is located. The triangle represents that the car's side radar has detected an object in the driver's blind spot.

Page 8-9; Collision Avoidance Information; Side Collision Avoidance

Icon. This is a side obstacle warning icon and is identified by number 306. The icon is square and has a white background. The icon shows an overhead view of a white car with a black triangle radiating from the right side of the car where the side view mirror is located. The triangle represents that the car's side radar has detected an object in the driver's blind spot.

Icon. This is a side obstacle warning icon and is identified by number 407. The icon is square and has a white background. The icon shows an overhead view of a white car with a black-and-white-striped triangle radiating from the right side of the car where the side view mirror is located. The triangle represents that the car's side radar has detected an object in the driver's blind spot.

Icon. This is a cautionary warning icon and is identified by number 235. The icon is a horizontal, yellow bar with four black, vertical stripes spaced evenly inside the bar.

Icon. This is a cautionary warning icon and is identified by number 74. This rectangle icon has black-and yellow-diagonal stripes sloping down 45 degrees from left to right.

Icon. This is a cautionary icon and is identified by number 299. The icon is a black square with an inverted, isosceles, yellow triangle drawn inside of it.