Development of Human Factors Guidelines for Advanced Traveler Information Systems and Commercial Vehicle Operations: Comparable Systems Analysis
Displaying Information in a Vehicle
General
A question in the TravTek post-experiment questionnaire assessed the general usability of the TravTek visual displays. The mean question responses are shown in figure 21. As shown, subjects rated the usability of the TravTek visual displays quite positively.
Route Guidance
Two of the primary issues addressed by the TravTek evaluation were the performance and preference of subjects using the various display configurations. As was discussed in the usability section above (see figures 18, 19, and 20), navigation performance didn't vary much among the TravTek configurations, with the exception of turn errors committed while using the route map without voice.
Subjects' perceptions regarding the value and usability of the maps and screens, as measured by questionnaire responses, are shown for the route map in figure 22, and for the guidance screen in figure 23. As shown in the figures, both maps were highly rated by the subjects. In fact, very few differences between comparable questions were found. A more informative finding with respect to the relative value of the route guidance screens is depicted in figure 24. This figure represents an objective and unobtrusive assessment (through data telemetry from the vehicles) of what the drivers actually used for the Navigator and Navigator Plus conditions.
Visiting drivers chose to use the guidance screen more than 85 percent of the time that TravTek was activated. Local drivers chose to use the guidance screen more than 70 percent of the time. Although the initial route guidance system default state was the guidance screen, the drivers had the capability to "swap maps" at the press of a steering wheel button at any time. It is clear from these results that the majority of drivers preferred the turn-by-turn guidance screen over the full-route moving map. However, it is also clear that some drivers in some circumstances found the route map more useful. In addition, there are a number of cases, including when the driver is off-route for either a planned or unplanned reason, for which a moving map may be the only way to effectively display required information. Therefore, the TravTek strategy of allowing the driver to access both types of maps appears to have been the most effective choice. This interpretation is supported by the questionnaire responses evaluating the "swap map" feature on the steering wheel (figure 25). As shown, the swap map feature was rated highly in general and as an aid to way finding in particular.
The use of voice to provide route guidance information turned out to be quite successful in the TravTek system. As previously shown in figures 18, 19, and 20, the addition of the voice generally provided improved navigation performance. Another informative result regarding voice is shown in figure 26. Rental drivers chose to have the voice activated while using TravTek more than 90 percent of the time. Note that drivers could deactivate the voice at any time by pressing a steering wheel button. Figure 27 shows that the local users used the voice more than 65 percent of the time. One lesson learned from this result is that voice is perceived as being valuable, particularly by drivers who are unfamiliar with an area. Perhaps a more surprising finding is that the local users continued to use the voice at a fairly high rate. Note that the local users drove TravTek cars daily for 2 months. Therefore, apparently voice is still considered to have value to users who are both familiar with the area and familiar with the vehicle and system.
A global rating of the TravTek voice intelligibility is shown in figure 28. Although the voice clarity, message understandability, and function were rated positively by subjects, these ratings are not as high as many of the other ratings. The "lesson learned" from this finding may be that the state-of-the-art in low-cost synthesized speech still leaves something to be desired. However, as previously described, the TravTek voice messages would have to be highly modified and would contain less information in order to use digitized speech, so the digitized vs. synthesized trade-off is still somewhat in question. Note, however, that the TravTek voice design and implementation was successful as indicated by these results and that technological advances will continue to improve the quality of synthesized speech at an ever–lowering cost.
Figure 29 summarizes the preferences for the various TravTek display options. As shown, preferences for the route map and guidance screen were almost identical. Note that consistent with the voice data described in the preceding paragraph, voice was preferred in conjunction with either visual display. This is in contrast to the rating of voice alone, which was still positively rated, but not as highly. Figure 29 provides a good summary of the "lessons learned" for route guidance. Drivers prefer to have the option in type of visual display as well as a voice feature. Another route guidance feature provided by the TravTek system was to offer a faster route if one became available while in transit due to changing traffic conditions. An item, If the TravTek system offered you new routes due to traffic conditions, how often did you accept the new routes?, was included in the post-experiment questionnaire to assess driver willingness to change routes to save time. The response to this item was 4.3 on a 1 = "None of the time" to 6 = "All of the time" scale. This result indicates that many drivers are willing to change routes to save time while en route.
Navigation Assistance
The TravTek system allowed access to a moving map even when a destination was not entered. This screen, called an "information map," allowed the driver to get proximal orientation and navigation information, as well as traffic congestion and incident information, without entering a destination. The perceived usability of this screen was addressed with an item in the post-experiment questionnaire. The subjects' responses to this item are shown in figure 30. As with the other visual display responses, subjects' opinions of this display were positive. The "lesson learned" from this finding is that providing drivers who are perhaps somewhat familiar with an area and their destination with supplemental orientation and traffic information is viewed as a positive feature.
Services Information
The services provided by the TravTek system are described in detail in the System Description section. In addition to the features associated with route guidance and navigation, such as entering a destination, the major services consisted of a services and attractions menu (including a "yellow pages" database of selected Orlando businesses), emergency services, and on-line system tutorial.
The subjects' ratings of the services and attractions directory are shown in figure 31. As shown, the features of the directory were rated positively by the TravTek users. There are similar results for the emergency services feature shown in figure 32. Although these ratings are quite positive, questions regarding the timeliness of the emergency services response were rated somewhat lower. These data may provide insight on features and aspects that have the most room for improvement in future ATIS
systems.
The perceived effectiveness of the on-screen instructions is shown in figure 33. As with the other services features, this instruction technique was rated positively by the subjects.
Cautions and Warnings of Hazards
The TravTek system provided warnings about congestion or an incident ahead on either visual route guidance display configuration. Figure 34 shows the questionnaire responses to the display of this information on a route map. As shown, several of the items received essentially neutral responsesðCsome of the lowest ratings received by the system. Of particular interest are the categories "Provided Believable Information" and "Provided Timely Information." A number of subjects apparently questioned the reliability and timeliness of the traffic information provided by the TMC. In future production systems, it is apparent that such information will have to be accurate in order to optimize driver acceptance and route-changing behavior.
In contrast to the somewhat lower ratings of traffic and incident information provided on the route map, the traffic information provided on the guidance screen was highly rated (figure 35). Note that the information provided for the route map and the guidance screen came from the same sourceðCthe TMC. Therefore, the reliability of the information was not different. The higher rating for the guidance screen may have occurred since less information was presented that could be evaluated by the subjects.
|
