The following taxonomy was developed on the basis of materials, data and information collected for this study. It reflects a very diverse and comprehensive review of available wireless communications technologies and associated equipment (e.g., third party mounting hardware, hands-free conversion, etc.), how it is used, and how specific design features and implementation strategies might be related to the potential for distraction and hence crash risk. Many issues are as yet unresolved but, as a general rule, the greater the number of features requiring user manipulation/observation, the longer the duration of manipulation/observation required, and the more difficult a device is to manipulate (e.g., hold securely, one hand vs. two hands vs. no hands), the greater the risks associated with its use in a moving vehicle.

The consideration and application of good human factors practice in the design of in-vehicle communications systems are seen as having a significant potential for reducing the risk of a crash while using these systems. In addition, an understanding of the relevance of these features to safety within the general public may also provide an industry incentive to design the "best" possible systems, since it can serve as a basis for purchasing preferences. In this regard it is recommended that industry generated human factors design guidelines be developed to provide informed guidance to equipment manufacturers on developing ergonomically sound systems.

Driver Tasks Associated with Cellular Phone Use While the Vehicle is in Motion:

Cellular telephone use while driving can be characterized by the tasks that make up such use. While phones with more advanced features (e.g., hands-free) may modify the nature of, or eliminate some of these tasks entirely, the majority of users currently own portable phones and at one time or another are faced with the following tasks:

 
• Handset Pickup and Stowage (answering a call, placing a call, hanging up after a call)
• Dialing (including recall of a stored number)
• Voice Communications (usually dialogue, listening, and talking)
• Associated Tasks (e.g., taking notes, referring to a calendar, reading a map).

The first may be considered trivial, especially if the location of the handset is well learned or the unit is of the hands-free variety. It is not trivial in the case of a hand-held phone that is out of reach. The dialing and voice communications tasks, on the other hand, are not trivial. These have been the focus of most of the published research in this area. As suggested in the body of this report, the communications aspect of phone use may play the dominant role in crashes, at least in the United States.

In Japan, where usage patterns may be different, research findings indicated that responding to a call is the dominant factor (see Chapter 3). With regard to associated tasks, less is known. Anecdotal information and common sense suggest that such activities are particularly risky since they would likely require the use of two hands and provide considerable opportunity for inattention to the driving task. Nevertheless, observations of drivers taking notes or checking maps while using their phones confirms the willingness of some to take these risks.

Regardless of the task being considered, it is clear that the application of good human factors design principles can reduce the adverse influence these tasks might have on a person's ability to drive safely. The following discussion of cellular telephone "features" highlights a number of areas where such considerations can provide significant safety benefits.¹

General Features

Hand-held vs. Hands-free : Hands-free systems have the obvious benefit of allowing conversation while the driver has both hands on the wheel. Variations in the actions required of users of these systems (i.e., voice answer, one button press, etc.), however, suggest that the best implementation strategies for these systems are yet to be determined.

Road noise may serve to mask the communications with hands-free systems. Thus, there may be a need to enhance the ability of the driver to hear the conversation through the use of an earphone (single). Such use is illegal in most states and may itself have distraction potential if the earphone has the potential to fall or if the earphone cord is cumbersome. Cordless earphones might solve the latter problem.

Voice activated vs. Non-voice features: These features are primarily related to hands-free operation and define the extent to which the user must interact with the system. A completely hands-free system requires no manipulation on the part of the driver, while less capable systems may require one or more buttons to be manipulated. Voice activated control of the volume may also be an alternative solution. Voice-activated control should reduce or even eliminate the demands of manipulation on the driver. It may nonetheless impose high cognitive demands (e.g., listening to a spoken menu of options) or induce unintended driver behaviors (e.g., taking eyes off the road scene to look at the device while speaking commands).

Mounting kit/base vs. no mounting kit/base: Any implementation that minimizes the need to hold or "pocket" the phone, whether in-use or parked in a cradle is preferable since it reduces handling related distractions. Nevertheless, placement of the mounting bracket from the standpoint of convenience and length of receiver cord is also critical to ensure normal driving posture can be maintained. In addition, crashworthiness considerations must also be addressed. Future vehicles will likely move towards a uniform cellular telephone/PCS bus interface in vehicles that will allow "plug and play" hands free capabilities, perhaps integrated with ITS technology.

Cord vs. Non-cord: As a general rule hand-held phones are typically operated in vehicles without being connected to the vehicle itself, although users have the option of connecting an external antenna, operating off vehicle power, or using a "cradle" which may provide these capabilities along with hands-free operation. To the extent that cords are used, they should be of sufficient length and be placed so as not to interfere with the task of driving. For "car phones" permanently installed in a vehicle, the accessibility of the handset and length of the cord are of primary concern. Because of short cords, drivers have been observed holding a handset in conversation while leaning over and crouched at the height of the dashboard.

¹ Many of the human factors issues discussed in this appendix are equally relevant to a number of ITS technologies (e.g., in-vehicle information systems) currently being developed, and studies are either ongoing or planned which address many of the same considerations. Information generated by these efforts could provide useful data to support the development of human factors design guidelines for wireless communications systems.

---

Antenna Configuration: (i.e., retractable vs. "pop-up" antenna vs. "antenna on case" vs. non-retractable ["stubby"] antenna vs. vehicle-mounted external antenna) As with other features requiring user manipulation, the need to manually extend an antenna on making or receiving a call is potentially distracting and may influence controllability of the vehicle, particularly when the effort requires two hands. Use of an external antenna or units with "pop-up" antenna will minimize distraction in this regard.

Flip-phone vs. non-flip Phone: In recent years flip-phones have become very popular and have allowed the overall size (collapsed) of the phone to be reduced while maintaining display and keypad size. While such features facilitate storage and usability, they do require the added task of opening or "flipping" the phone for use, a task that typically requires two hands. Here again, the additional distraction and the increased manual demand have the potential to increase risk.

Various Logics for speed- or auto-dial (e.g., <RCL> <4><SEND> vs. <RCL> scroll down to Option 4 in stored Phone Directory, then <SEND>): As illustrated by the examples, keypad strategies for "speed-dial" or "auto-dial" options can vary considerably. Efforts to simplify this task to minimize time required are worthwhile to the extent that they reduce both the time on task and opportunity for error.

Power (watts) of unit affecting ability to maintain communications and avoid dropouts: Variations in "cell" locations, physical obstructions, and cellular telephone transmitting power capabilities can result is signal drop out, an event more frequently encountered with hand-held phones since they are generally of lower power, have a built-in antenna and are used within the metal shell of the vehicle. Observations indicate that some drivers attempt to compensate for intermittent signals by adjusting position of the hand-held phone, shifting the unit from one hand to the other or even holding the unit partially outside the vehicle window.

All of these behaviors have the potential to divert attention from the driving task and possibly interfere with control of the vehicle, particularly under emergency conditions. This problem may best be managed by the use of vehicle mounted antennas, where the process of connecting to the antenna is convenient and non-interfering. Alternate strategies of lower power hand-held units with on-board "repeaters" may also solve the problem with no hard connection to an external antenna.

Stowing of cell phone: (e.g., search in pocket, vs. cradle vs. loose on car seat): The results of the Japanese study highlight the potential consequences of responding to calls while driving. Although driver actions/responses associated with the crashes cited varied, the findings are consistent with crash data in the United States. Clearly, reaching or searching for the phone can be very distracting, particularly when reaching results in movement of the steering wheel in combination with reduced attention to the driving task. This suggests significant benefits of having a convenient cradle available for holding, as well as using the phone.

"Hold-ability" (Based on size, texture, curvature, etc.): Crash data also indicates that incidents can occur when a driver drops a cellular telephone. Dropping a cellular telephone usually results in a natural response to retrieve the unit, an activity that may clearly put the driver at risk. Ergonomic design consideration here can play a significant role in mitigating the potential for such an event.

Visual Display Features

The visual display features outlined below all interact to define visual and cognitive demands on the driver and serve to determine the readability of the display and the degree of attention required to obtain the displayed information. These design features in particular are in need of human factors consideration insofar as multifunction displays may require considerable attention to displayed information. It is important to optimize both the display and the presentation of information to minimize demands on the driver. Human factors research is clearly needed to establish optimal design trade-offs for a given device architecture.

Large screen-area display vs. small screen-area display: Trends toward miniaturization have the potential to shrink the size of screens and hence reduce their readability. Careful attention to text characteristics (e.g., font, color, size), presentation style (e.g., lower case vs. caps) and format (e.g., number of lines, line length) can be particularly beneficial for small screen displays.

Luminous display (e.g., LED) vs. backlit display (e.g., LCD): Considerable variation in ambient light conditions in vehicles, particularly bright sunlight, can significantly impair the ability of a driver to read a cellular telephone display. Since difficulty in reading a display may precipitate behavior (e.g., changing phone orientation, extended periods of inattention) which is contrary to safe driving, optimizing the readability of the display from this standpoint is also of great importance.

1-line vs. Multi-line displays: As wireless services increase, there is likely to be an evolution of cellular telephone display formats that incorporate more than one line of text. Multi-line presentations may result in longer glance times and more glances per transaction, potential concerns from the standpoint of attention. In addition, text characteristics, presentation style and display format all must be considered together to ensure that readability of the information is not compromised to achieve a more comprehensive presentation of information. As a possible solution, this suggests that cellular telephones have two modes of operation, one for stationary use and one for road use, designed specifically to limit the need for extensive driver attention or manipulation.

Text font, style and presentation: As suggested earlier, font selection and case (e.g., all upper case vs. mixed upper and lower case) can both influence display readability. Similarly, presentation of the information in terms of the manner in which consecutive lines are displayed (e.g., scrolling vs. paging) can also influence readability as well as the need for user refresh.

Keypad Features

Keypad button size and spacing: As cellular telephones get smaller the size of keypad buttons and/or their spacing is likely to be reduced. Attention to this feature is important so as to minimize eyes-off-road time while keying, to avoid the need for reentry of information, and to reduce the need to frequently verify entries where errors are perceived to be likely.

High clutter vs. low clutter on keypads (more vs. fewer keys): Here again, as wireless devices get more sophisticated and offer more capabilities it is likely that keypads will get more cluttered, a situation that can lead to increased difficulty in accurately making key selections while driving.

Multifunction keypads are not necessarily a straightforward solution since some keying logics may place high demands on the driver.

Stimulus-response compatibility (e.g., up-down arrow keys arranged side-by-side are not S-R compatible; up and down arrow keys arranged one above the other are S-R compatible): Layout and function of keys should be logical and follow user expectations to ensure that errors are kept to a minimum.

Keying feedback absent or present (e.g., some cellular telephones provide auditory feedback in the form of a beep upon button depression, some have the buttons designed to provide an audible "click", and some have buttons designed to provide tactile feedback): Regardless of the approach used, some positive feedback is necessary on key presses to reduce the need for input verification.

Auditory Display Features

Clarity of receiving (and sending): Based on the quality of the audio features of the unit, the intelligibility of the communications can have significant implications for the cognitive loading of the driver. Poor audio quality is likely to require very focused attention on the part of the driver. Unfortunately, audio quality is a consequence of the combination of units communicating and therefore is not wholly under the control of the mobile user. Nevertheless, clarity of speech is an important consideration.

Startle potential of incoming call announcement (e.g., various types of "ring," voice announcement; "vibration" feature in flip-phones): The startle created by a ringing cellular telephone has been identified as a possible problem by some drivers. This potential suggests that the manner in which drivers are notified of an incoming call should be examined carefully. In particular, special in-vehicle alerts should be investigated which are less
likely to startle a driver. Use of vibration, a gradual increase in ring volume, or use of voice alerts may be appropriate alternatives.

Key press feedback (informational): While the need for keypad feedback has been addressed above, such feedback offers the opportunity to provide the user with distinguishing information as to the keys pressed through the use of codes (e.g., number of beeps) or differentiating tones. Such feedback may reduce the need for verification of keypad entries.

Associated Tasks

In the course of a call, particularly a business call, a driver may find it necessary to carry out associated tasks. These tasks might include, for example, taking notes, retrieving information from a calendar or notebook, or scanning a map. Such behavior has obvious risks associated with it. For those drivers who link their computers or faxes to their cellular telephones, the risks are equally obvious. It is not clear what technology can do in this arena short of enhancing the driver interfaces, supporting communications with ITS technologies, and perhaps building into cellular telephones a voice recording/data recording capability for later access.

Other Considerations: Content, Environment and Human Variability

Along with the design considerations discussed above, there are a number of other factors that may interact with design so as to facilitate or impair the users' ability to operate wireless systems safely. The content or nature of communications has already been mentioned as having potential for "capturing" a driver's attention and influencing situational awareness. Other factors or interest include traffic, roadway, weather (visibility), ambient light (day/night) presence of passengers/children in the vehicle, noise, etc., all of which can influence the workload imposed on the driver and provide additional opportunities for distraction. Predicting the influence of these environmental factors is further compounded by individual differences in ability to timeshare between the task of driving and other secondary tasks such as speaking on a cellular telephone.

Individual differences associated with circadian rhythm (i.e., effects of time of day on attention) also can play a role in one's ability to drive and use a wireless system. These issues highlight the importance of individual responsibility in selecting a "safe" wireless system and using it responsibly within their capabilities and under appropriate circumstances.

Summary

The discussion above highlights the potential importance of human factors design considerations to the safety of using wireless communications systems in moving vehicles. New light weight digital wireless systems on the immediate horizon, capable of a variety of data services, voice memo capabilities, built in phone directories capable of holding as many as 250 entries, and using color active matrix multi-line displays, have the potential to place significant demands on users. While the responsibility for safely using these systems in a moving vehicle lies with the driver, careful attention to their design by manufacturers along with careful selection of the devices and their implementation by the user will help reduce the risk of a crash.

Nonetheless, it must be emphasized that while attention to these design issues can enhance the safe use of a given wireless system, to the extent that such considerations (by facilitating usability) increase the use of the systems (e.g., frequency or duration of calls), there is a potential for a net increase in crashes due to increased use and, hence, increased exposure. It is the uncertainty of how this will play out in the future with more demanding technologies and societal pressures, that highlights the need for more accurate data collection and a greater emphasis on human factors research.