Stewart Katz, Jill Fleming, Paul Green, David Hunter,
and Daniel Damouth
Two experiments examined the safety and usability of the Siemens Ali-Scout navigation system. The in-vehicle interface provides turn-by-turn visual and voice guidance. In the first experiment, 54 drivers varying in age drove to four destinations, twice using the Ali-Scout and once using experimenter verbal guidance. Subjects were tested in the afternoon, at rush hour, and in the evening.
There were no crashes or near misses using the Ali-Scout, but there were four critical incidents where drivers changed lanes in response to navigation voice instructions without checking traffic. Excluding the turns into destinations, the turn error rate was 8 percent, and uncertainties occurred at an additional 13 percent of the turns. Most of the errors and uncertainties occurred in autonomous mode in which only the distance and direction to the destination are shown, not turn-by-turn guidance. Most longitudinal control measures (trip duration, mean and standard deviation of moving speed, overall mean speed, mean throttle position, and the standard deviation of throttle position) reflected significant experimental differences.
Drivers rated the interface as reasonably safe for themselves but not as safe as the Rockwell PathMaster or the simulated UMTRI interfaces, primarily due to mistiming of the voice guidance.
In the supplemental experiment, eye fixation data at night for the Ali-Scout was collected for 10 drivers. An additional three drivers used the PathMaster. There were no crashes, near misses, or critical incidents associated with the PathMaster. In contrast to the first experiment, measures of lateral control (lane position) were most sensitive to experimental differences.
|Graphical Abstract (.pdf)|||||UMTRI-96-32 Full Report (.pdf)|
Close This Window