We're just a few years away from a revolutiona dramatic shift in technology similar in its implications to that of the introduction of personal computers on office desktops during the 1980s. Powerful computer hardware, software, networks, and wireless communication capabilities are being embedded in vehicles that transport people everywhere in the world. Vehicle user-interface (VUI) design, including, specifically, information visualization, sonification, and haptic technology, are fundamental challenges to developers. In addition to the traditional human factors challenges, and the more recent challenges of branding, cultural diversity, and emotional appeal, there is one additional, inescapable factor: life and death are at stake.
Another significant difference from desktop and mobile device contexts is that VUIs represent a context in which the user experience is significantly controllable. Good VUIs can improve both performance and preference, but VUI developers must give their tasks of planning, research analysis, design, implementation, evaluation, documentation, and training careful thought and consummate skill. The context of communication will be not only information, but also persuasion (advertisements) and entertainment.
The computer-human interaction (CHI) community, representing user-centered, user interface design, has an important role to play, but CHI professionals are going to have to be politically and socially sensitive to other professional communities as all attempt to provide their skills to multidisciplinary challenges. At CHI 2003 and at HCII 2003, I organized panels on vehicle user-interface design to explore some of the differing viewpoints and potential contributors to successful solutions. Here are some key stakeholders.
- Vehicle manufacturers: They are concerned, are sometimes uncertain about what to do, and depend heavily on system suppliers and the industry's traditional service providers: human factors specialists and product designers. They are also used to traditional cycles of product development, which may take six years, a scenario quite different from consumer electronics, which has taken a few years, or the now constantly emerging products and services cycles of the mobile phone, personal digital assistant, and Internet industries. Some vehicle manufacturers are attempting to shorten development time significantly, but even roll-outs in half the time are still much longer than for the components provided by others.
- Vehicle system suppliers: They include manufacturers of sound systems, instrumentation, navigation, and media communication. Examples include Aline, Bose, Bosch, Harmon Kardon, Fujitsu-Ten, and Qualcomm. Some suppliers may or may not have user interface (UI) development groups with the requisite skills required for newer, more complex vehicle telematics data (such as tire pressure, brake condition, fuel consumption efficiency, preventive maintenance scheduling, electrical or electronics system status, or road conditions) or for the rich amounts of phone, Internet, e-mail, video, and other audiovisual persuasion or entertainment sources that already exist in some vehicles but will soon be in almost all.
- Human factors specialists: These skilled analysts have worked in the automotive industry for decades. They have amassed impressive compendia of knowledge related to safety and ergonomics in general, that is, in many areas of usability, but some may not have a background in usefulness and appeal.
- Product designers: These designers and analysts have a long history of automotive styling, interior design, and some experience with dashboard design. However, as with many product designers working in the user-interface realm, they may have less experience with the complex communication, interaction, and storytelling of large collections of content or "industrial-strength" functions that are typical of advanced software applications or multimedia content collections such as those found on Web sites, CD-ROMs, or DVDs. Some firms have proposed prototype designs modeled on video games, arguing that for vehicles to appeal to a youth market, the instrumentation should look-and feel like video games. One hopes that making the designs more game-like has positive performance results and is not a catalyst to emotions and behavior that prevail in today's games that emphasize unreal, extreme action, and often revolve around death and destruction.
- User-experience designers, analysts, and engineers: A wide range of disciplines and professional groups stand ready to provide guidance for "the user experience," however that is defined. Terms, concepts, and methods differ among professions, but inter-professional, interdisciplinary groups have formed at CHI, the American Institute of Graphic Arts (AIGA), and elsewhere. Among others, graphic designers, interaction designers, information designers, information visualizers, and, in general, user-interface designers and analysts are all able to provide significant expertise and experience. One ought to mention as well the Usability Professionals' Association (UPA), whose flagship publication's name, User Experience, already signifies a set of concerns that can be applied to vehicle matters.
- Usability analysts: In general, these researchers and analysts stand ready to evaluate vehicle systems with methods that are similar to human factors specialists, but also different. Their focus groups, user profiles, use scenarios, user testing, and heuristic evaluation methods have matured in the realm of consumer applications, entertainment media, search engines, as well as radio knobs and door handles.
- Documentation and training specialists: One version of a driver manual for an advanced vehicle user interface is approximately 300 pages long. This is only the beginning of the increasingly complex online and offline training, documentation, and help that drivers and riders will require (in addition to simpler, more effective user interfaces), if there is any hope at all for any of us to become competent in the new systems about to be installed in our vehicles. Documentation and training specialists may be able to help.
As you can see, many other disciplines may join CHI professionals to collaborate on or compete for overlapping, desired or required skill sets. To become more effective, CHI professionals may need to learn more about the standards, specifications, issues, and concerns of the vehicle manufacturers as well as attend some of the showcase automotive conferences that take place each year in Detroit, Tokyo, and elsewhere. In addition, knowing basic sources of information is helpful. One place to start is the activities and publications of the University of Michigan's Transportation Research Institute (UMTRI), which for decades has produced vital studies and recommendations on vehicle systems, telematics, and human-computer communication and interaction. One of its steadfast researchers is Paul Green, who has worked in this field for many years.
In 2000, BMW asked my firm to investigate much of the public literature about the driver experience (as opposed to the rider experience), and kindly gave us permission to publish a summary. Some of these key ideas may sound familiar and seem almost self-evident. However, remember that in the automotive vertical market, user-centered design has sometimes taken a back seat to technology and marketing, or has focused on a narrower range of concerns. Bear in mind, also, that many UI designers do not have to consider safety and life-and-death circumstances as part of their typical decision-making matrix of issues.
Here are the basic maxims we synthesized from the material we studied with a brief description:
- Design for safety: Displays must be readable at a one-second glance. Most tasks should be completed in two glances. Users should be able to complete most other tasks in a 15-second period during conditions when the vehicle is stationary. All must be designed to avoid driver distraction, which means most tasks must be broken up into small steps. The use of a cell phone increases the risk of a crash by factor of five. The design implication: Simplicity must be emphasized to increase clarity and driver confidence. For example, a small detailed map diverts more cognitive focus than simple, direct navigation instructions.
- Reduce complexity: Developers first should offer helpful, rather than powerful, features. They should emphasize output leading to vehicle actions and displays, services, and automation, rather than the user's input and control. Rather than offer users full control, the VUI should emphasize useful defaults, for which the default choices cover the most frequently (approximately 80 percent of the time) desired results. The designs should focus foremost on driving-related functions. Developers can accomplish this objective by using task analysis to eliminate "function bloat," determine the most useful features, and concentrate on providing simple VUI solutions first for most drivers, then enable greater complexity for those who desire or require it.
- Use a graphical user interface only when necessary: Graphical UIs are less direct, less dedicated. For example, many clicks in a graphical UI may be required for simple tasks, like selecting desired music. Graphical UIs should be used for dynamic content and can be powerful means to summarize complex situations through charts, maps, or diagrams. The right choices require careful consideration of trade-offs.
- Use physical controls appropriately: Physical controls dedicated to a single function are often more direct, and similar actions produce similar results (like brake and accelerator pedals). Muscle memory, physical access, and multimodal input often may be the most powerful attributes that create a safe, efficient, and satisfying experience. Developers should retain physical controls for urgently needed and frequently needed functions. Other functions need to be evaluated for physical vs. graphical UI optimization.
- Avoid cognitive and sensory overload: The point of diminishing returns is being reached in VUIs, because attention is a limited resource and cognitive costs are cumulative. In addition, age has a dramatic effect. Consider that many applications may not have special versions for both teenagers or seniors yet, but VUIs must immediately provide for seniors and their special needs, in which older users require 60 to 120 percent greater time for learning and decision making. Developers must consider workload management and overall cognitive capacity, in which strategic tasks (for example, choosing destinations and routes), maneuvering tasks (reactions to drivers, conditions), and control tasks (basic, automatic processes), must all be performed well. Testing designs with users is likely to be more extensive than for office desktop systems or home entertainment systems. Safety-centered design focuses on reducing distractions and finding an optimal balance of sensory, cognitive, and time-critical demands. Remember: Most drivers will begin as novices with VUIs but must begin performing well almost immediately.
- Allow customization of information: Customization of the VUI itself is risky but can be considered if that customization enables drivers to successfully accomplish their tasks in shorter time. It is also likely to be a consumer demand. To know how to customize, developers must follow a driver-centered design process, especially as cognitive issues become more important than physical. Remember: Users are quick to abandon difficult software if alternatives are available.
- Follow a user-centered design process: Much of the literature points to the need for a user-centered UI development process:
- Determine user profiles
- Determine goals
- Define scenarios
- Design prototypes early
- Evaluate alternatives
- Iterate the design
This is not exactly hot news for the CHI community, but for some vehicle and system manufacturers, it places different emphasis on what has been often a technology- and marketing-driven context. These maxims suggest that the CHI community has a strong educational challenge ahead. The preceding has emphasized a traditional human-factors orientation and traditional UI development process, which emphasizes that VUI development centers should consider the following:
- Adopt a user-centered design process.
- Work with UI and information-visualization experts to develop centers of competency.
- Plan systematic series of projects to improve UIs.
- Test existing (and competitive) systems.
- Test actual performance, not only focus groups.
- Conduct detailed UI audits, observe drivers and riders, analyze tasks.
- Design prototypes based on alternate controls and displays, such as heads-up display, touch screen, multiple displays, auditory cues.
- Develop specific vehicle UI rapid-prototyping tools.
While all of the foregoing focusing on human factors and ergonomics is extremely important for the success of VUIs, other challenges are emerging for developers, just as they are emerging in other fields such as desktop and Web applications, information appliances, and mobile devices. Among other new issues are the following:
- Intelligent or "smart" vehicles will have significantly new hardware and software. For example, large LCD or multiple-LCD display screens enable dynamic, unconventional displays quite different from the electromechanical displays of past vehicle systems. Sensors and wireless Web connectivity imply the possibility of fundamentally new content.
- In addition to usability, usefulness and appeal have increased in importance for drivers' and riders' experiences. Among other human factors analysts, Jordan expresses this concept in the title of his recent book: Designing Pleasurable Products: An Introduction to the New Human Factors.
- As in many realms, the content for vehicle displays will include not only informational, but also persuasive and aesthetic, communication. Especially important are the issues of branding, and more likely, co-branding of information providers.
- With increased commuting times, vehicles will become a temporary mobile work and home place for both drivers and riders, with additional communication and interaction of the following kinds:
- Information about one's own identity
- Relationship-building and maintenance (communication with others)
- Self-enhancement (including education and health monitoring)
- Access to information resources (travel information, references, and other databases)
- For example, riders and drivers may experience virtual presence in other spaces, play games with other vehicles' drivers and passengers, review status or instruction concerning driving and parking maneuvers, repairs, or parts reordering.
- Globalization and localization of product development will require increased attention to cross-cultural communication and other dimensions of difference of groups as well as individuals. As I have commented elsewhere, beyond culture, issues of trust, persuasion, intelligence, and cognition itself may cause vehicle UI developers to consider new kinds of metaphors, mental models, testing techniques.
Innovative VUI thinking, cross-cultural analysis and design, and exploration of human factors and visualization issues beyond the traditional concerns of usability will need to be considered more integrally in vehicle UI development. Developers will need new kinds of checklists and guidelines to assist them in their work. Having a better understanding of fundamental human factorsthe relationship of VUIs to culture and other issuesdesigners can make better decisions about usability, usefulness, aesthetics, and emotional experience. Ultimately, better, more powerful, and more appealing vehicle user interfaces and information display will emerge. The road ahead is a wide and challenging one.
7. Marcus, A. User-interface design, culture, and the future." In Proceedings of the Working Conference on Advanced Visual Interfaces (M. De Marsico, S. Levialdi, and E. Panizzi, eds.), Trento, Italy, May 22-24, 2002, pp. 15-27.
Aaron Marcus, President
Aaron Marcus and Associates, Inc. (AM+A)
©2004 ACM 1072-5220/04/0100 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2004 ACM, Inc.