U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
Publication Number: FHWA-RD-03-063
Date: September 2004
In-Vehicle Display Icons and Other Information Elements: Volume II
PDF Version (1.07 MB)
PDF files can be viewed with the Acrobat® Reader®
CHAPTER 2: SUMMARY OF PROJECT TASKS
This section of this report summarizes the following project tasks:
The purpose of task A was to conduct a review of relevant symbols and research, including the use of symbols by manufacturers and after-market vendors for existing and planned in-vehicle systems. The methodology employed included examining articles collected as part of the previous guideline development efforts, conducting extensive database searches, and accessing the Internet to gather information regarding the most current use of symbols in existing and future in-vehicle information systems. More than 200 articles, several books, and more than 100 Web sites were found via this methodology.
Once all of the literature was gathered, it was reviewed to determine its relevance to the current project. Specifically, attention was given to including information relevant to:
The review comprehensively captured the status of icon/symbol research and applications and, most importantly, provided a solid foundation for subsequent tasks in this project. The following conclusions emerged from the review and analyses of the literature for icon and symbol research and current applications:
The lack of guidelines and standards for icons has resulted in design by consensus, a lack of scientific rigor in icon development, and multiple icons for the same messages. A number of sources for existing transportation symbols and icons were found during this literature review. They provided symbols and icons for numerous transportation applications (i.e., road signs, traveler information, warnings, etc.). However, during our investigation of these sources, it became clear that the process of developing and choosing icons is very subjective. Icons are frequently incorporated into system designs on the basis of consensus, opinions, and aesthetic preferences of the system development team. While such a process can result in an effective icon (as evidenced by the many effective icons in use in in-vehicle devices), it also runs the risk of producing unclear and ineffective icons.
Existing literature and standards provide little guidance for the design of new icons for IVIS devices. The majority of literature relevant to the design of icons can be placed into one of two categories: (1) a general discussion on the development of symbols (i.e., what they are, why they should be used, and how they work), or (2) proposed methods for the evaluation of symbols and demonstrations of these evaluations using existing symbols. Although the existing literature helps provide background information necessary to understand how we derive meaning from icons and symbols, it does little to aid in the development of design guidelines for new icons.
General design principles for icon design are sufficient to avoid development of a "bad" icon, but are not specific enough to support development of the "best" icon. The review and integration activities suggested that a number of general principles for icon design are available to IVIS developers. These include the structure, shape, and color of icons. However, for several reasons, these guidelines are not sufficient to support the development of optimum, or "best," icons. First, these principles, like many human factors guidelines materials, are not specific enough. For example, they identify how icon shape can affect comprehension. But how do designers select a shape to begin with? How does a designer start with "givens," such as a driving context and driver information needs, and identify a shape that matches these "givens?" Second, it is difficult for designers to know how to apply the principles in any given situation. For example, when is the structure of an icon more important than its shape? Are there times when the conspicuity of an icon (such as a collision avoidance warning) is more important than the details of its physical design? How do color and shape interact to impact the driver's interpretation of an icon or symbol? Third, existing principles do not provide adequate guidance on issues such as how to "match" an icon with its associated message. In this regard, available principles for icon design do not generally address the importance of information elements (the purpose of the icon, such as alert, inform, plan, and decide) to the driver's accurate interpretation and effective use of icons. Thus, there are still considerable gaps between the needs of icon developers and the availability of human factors design information.
Development of new icons and symbols for in-vehicle devices will require iterative testing and evaluation; existing test and evaluation methods provide sufficient scientific rigor for future evaluations of icons and symbols. The interpretation and ultimate utility of icons and symbols depend on the relationship, or "match," between the message and the graphic elements selected to convey the message. Unfortunately, there is no immediate or obvious method of determining this "match," given the variability associated with IVIS devices, IVIS messages, and drivers. Therefore, good icon design requires development of a range of candidate icons and, equally important, iterative testing and evaluation of these candidate icons.
Despite industry concerns over the utility and relevance of human factors design guidelines, rigorous and proven methods for design guideline development exist and will be used in tasks C and F of this project. . Designers of advanced automotive displays have criticized many existing human factors reference materials for being too wordy, too general, and too hard to understand, and have requested guidance that is concise, specific, and clear.(1) In particular, three challenges are associated with the development of human factors design guidelines for in vehicle icons and symbols: (1) the lack of human factors design criteria; (2) the development of selection criteria for data sources used to produce guidelines; and (3) variability in the user population of human factors design guidelines.(1,2)
Despite these challenges, a number of successful design guidelines for ATIS, CAS, and other in vehicle devices have been developed.(3,4) The general procedures used in these efforts were used to guide design guideline development activities in tasks C and F of the current project.
Task B in this project served two very useful purposes. First, it identified credible procedures, heuristics, and principles for the joint use of visual, auditory, and tactile information to present in-vehicle messages. The task B report documented the underlying rationale for selection of display modality by reviewing the relevant literature and assessing the current state of knowledge. Second, task B defined message characteristics that should guide symbol design. Defining these characteristics and their interactions helps to identify design tradeoffs and provides the basis for future design guidelines and tools. In summary, task B provided design guidance for the joint use of visual, auditory, and tactile information presentation and provided a foundation for future design tools that could assist designers in specifying icon design for in-vehicle information technologies, particularly as they relate to ATIS.
The process used to identify design requirements of in-vehicle icons and IVIS messages in task B included seven basic steps:
Figure 3 shows the order in which these steps were completed and their interrelationships. From this flow diagram, we are able to see how the results of each step were used to develop the final product design requirements and tradeoffs for categories of IVIS messages.
One of the first steps completed during task B was to review the literature for information related to the development of rules for selecting display modes. An examination of these rules led to the design of several decision aids that would assist designers in the selection of a sensory modality for displaying different pieces of in-vehicle information. Each decision aid was tested using several candidate information elements until a final viable approach could be determined. The final approach was then refined through additional informal testing and analysis. The final design tool can be seen in figure 4.
The results of applying the design tool suggest that: (1) the visual modality should be used to display more complex information that does not require the driver's immediate response and may need to be referred to at a later time; (2) the auditory modality should be used to present simple information that is extremely urgent or critical messages that require the driver's attention; (3) a combination of the auditory and visual modalities should be used to present information that is both complex and relatively urgent but is too complex to be presented via simple tone or verbal message; and (4) the tactile modality did not appear to be a viable option for presenting any IVIS messages that had been identified.
The next step was to provide a more solid basis for design by defining the messages according to their contextual characteristics and the information processing elements they supported. Once this was complete, a cluster analysis was conducted; it identified 12 unique clusters of IVIS messages. To organize these clusters for interpretation, a further analysis identified 4 groups of clusters based on the center of each of the 12 clusters. Table 1 summarizes each of the message groups and the design requirements that they support.
Devising these design tools and analyzing the current list of relevant IVIS messages resulted in the following conclusions in task B:
A review of existing literature regarding visual, auditory, and tactile information presentation provided numerous general principles for modality selection, which was the basis for an effective sensory modality design tool. A review of both general human factors research and more recent research directly related to ATIS and Collision Avoidance System (CAS) displays provided a number of general principles and heuristics regarding different display modes (visual, auditory, and tactile). Summarizing these rules and categorizing them according to the design decisions they supported led to a design tool that would direct designers toward the most appropriate sensory modality choice.
Results of applying the sensory modality design tool indicated that the visual modality should be used for presenting complex messages that are less urgent and critical and that the driver may need to refer to at another point during the drive. Auditory messages were identified as those that have some type of alerting property: They provide the driver with urgent and critical information that is simple enough to be presented via an auditory tone or a brief verbal message. A combination of the visual and auditory modalities should be used for messages that require the driver's attention but are too complicated to be presented by an auditory message or will be referred to again later in the drive. The tactile modality was not identified as appropriate for displaying any of the 273 candidate IVIS messages. However, it is important to note that in a few instances tactile displays have been shown to be useful (i.e., the shaker stick on an aircraft); therefore, they should not be ignored as a potential display modality.
Classifying IVIS messages according to ITS technologies and general functions is not sufficient for providing effective design guidelines. Classifying IVIS messages according to general IVIS capabilities and functions does catalog the range of messages and show similarities based on the IVIS capabilities they are meant to support. However, this organizing approach does not reflect several important characteristics of IVIS messages that can influence design guidelines. Effective design guidelines and design tools require a description of IVIS messages that reflects message characteristics that influence driver comprehension and response. Defining messages according to their driver-relevant characteristics provides a more solid basis for design.
Understanding the driving context under which IVIS messages are presented is critical for successful design guideline development. Successful presentation of IVIS messages using icons depends on creating a message appropriate to its driving context. This report defines the context of IVIS messages using four dimensions that capture key elements of how context aids the interpretation of messages. Specifically, message urgency and criticality identify the consequences of not responding to a message in a timely manner. In contrast, dimensions such as the link to the driving task and the independence of the message identify opportunities to enhance the interpretation of a message by providing additional cues. Grouping the messages according to these four dimensions provides a first step in defining the requirements for integrating IVIS messages into a coherent set.
The Information Processing Elements (IPEs) associated with an IVIS message can be used successfully to develop the design guidelines that consider the perceptual, memory, and motor control limits of the driver. This report identifies nine different IPEs: alert, identify, search, evaluate, plan, decide, coordinate, control, and monitor. Together, these nine elements describe the range of information processing activities supported by IVIS messages. Each element supports a different set of design requirements that complement those identified by contextual characteristics. Identifying the elements associated with each individual message informs the designer about design decisions and tradeoffs that will need to be made for several different design parameters.
The cluster analysis technique provides a powerful tool to focus future analyses on a meaningful subset of possible combinations of contextual characteristics and IPEs. The cluster analysis proved to be a very effective technique in the preliminary assessment of visual symbols. The original four contextual characteristics (with five levels within each), combined with the nine IPEs, yield 5,626 unique combinations. This presents designers with a dizzying array of tradeoffs to make when designing in-vehicle icons and other information elements. This approach uses a tradeoff analysis that serves to focus future design guideline development efforts. Using statistical clustering techniques, the preliminary analysis identified four general message groups that describe 12 message clusters. These groups and their corresponding clusters identify important combinations of contextual characteristics and IPEs that describe the range of IVIS messages. Preliminary consideration of these groups and clusters suggests that each cluster and group has unique design requirements for in-vehicle messages. The initial description of these design requirements and their associated tradeoffs provides the basis for more specific design guidelines and practical design tools.
The tools and decision aids developed as part of task B provided the project team with a solid analytical foundation to begin guideline development in task C of this project. Combining the information obtained by identifying:
The initial description of these design requirements and associated tradeoffs provided the basis for more refined design guidelines developed as part of task C of this project.
A key challenge associated with task C would be to integrate the information provided in the task B report and develop clear, relevant, and easy-to-use design guidelines for in-vehicle icons. The task B report established some important relationships among IVIS messages, display modality, the driving context, and IPEs of the IVIS messages. Understanding these relationships is necessary, but not sufficient, to support the development of clear, relevant, and easy-to-use human factors design guidelines for in-vehicle icons and other information elements. During task C, the project team would need to integrate the information presented in this report and the task A report with specific design options for icon design such as background, symbol, border, symbol elements, and text labels.
The overall goal of task C was to produce a set of preliminary human factors design guidelines for in-vehicle visual symbols. Toward this end, task C activities included three interrelated subtasks:
Each of these task C subtasks is discussed in more detail below.
Developing a workplan for the preliminary guidelines. The development of the Human Factors Design Guidelines for In-Vehicle Symbols and Other Information Elements was directed toward answering the following general question: What is the relationship between various symbol design parameters and a driver's ability to effectively and comfortably use automotive displays-given variations in operating conditions, driving tasks, and driver demographics? In the human factors community, it is widely acknowledged that the majority of human factors/human performance research findings cannot be easily generalized to complex, real-world systems. This is often because such research is situation-specific (i.e., the results are highly dependent upon the environment, context, experimental task, subject demographics, etc. that identify a given study). Nonetheless, legitimate and much-needed generalizations of the data are possible through careful and thoughtful integration of existing research findings.
Specifically, development of the guidelines was expected to include several essential elements: (1) a comprehensive database of existing information sources; (2) knowledge of the automotive display design environment, including tradeoffs and constraints; (3) analytical activities, including a clear definition of the criteria that will be used to determine the quality and applicability of data sources; (4) a willingness to apply experience and judgment to the development of human factors design guidelines; and (5) methods of presenting the design guidelines in a manner compatible with the user-community's needs and desires for the guidelines.
Figure 5 shows the overall design guideline development process presented in the task C workplan, while figure 6 shows the process for developing individual guidelines.
The preliminary design guidelines were expected to contain a similar content structure as that used in the Advanced Traveler Information Systems/Commercial Vehicle Operations (ATIS/CVO) design guidelines produced for the Federal Highway Administration (FHWA). The final content was expected to be determined after consultation with FHWA and the project working group members. Figure 7 summarizes the proposed table of contents for the preliminary guidelines.
For the preliminary design guidelines, a two-page presentation format was chosen to satisfy the conflicting requirements for having simple graphics versus more complex, detailed descriptions of "real world" design situations. A schematic example of the two-page format presented in the original workplan is shown in figure 8. The left-hand page presents the graphic data and simple supporting text, while the right-hand page provides the more detailed information that a human factors practitioner is likely to require in performing his or her design tasks.
The contractor's previous experience in developing human factors design guidelines provided us with a number of heuristics regarding "ideal" presentation formats. These heuristics included:
Developing the preliminary guidelines. The task C preliminary guidelines were subsequently developed according to the steps outlined in the workplan described above and distributed to both FHWA and the project working group members. The actual table of contents from these preliminary guidelines is shown below in figure 9.
Developing a working paper that explored information and symbology usage issues. The final task C activity was to develop and present an overview of key theoretical, development, empirical, or design issues associated with in-vehicle icons. This subtask provided an opportunity to consider the work that has been performed during previous project activities and to identify areas where additional analysis, empirical study, or guideline development was needed. In the task C5 report, ten key issues associated with in-vehicle visual symbols and information elements were identified and briefly discussed. These issues were: