Abstract
Research and /or Engineering Questions/Objective:
On in-vehicle information system (IVIS), the manufacturers are developing the human-machine interface (HMI) in view of reducing driver distraction as well as improving the usability. However, recently, multi-modal HMIs such as gesture or speech dialogue have increased. The objective of this study is to develop an easy-to-use bench method for assessing multi-modal distractions and to make clear the capability and limitation of the method.
Methodology:
We combined two types of simple tasks. One is a pedal-tracking task (PT) which participants maintain the size of a square on LCD in front of them by using a gas pedal. The other one is a detection response task (DRT) which participants step a brake pedal as soon as possible after a visual stimulus are presented. Participants are required to maintain the square size and to respond to the visual stimuli simultaneously (PT-DRT). We think these tracking and detection response task are the simplification of the actual driving operation. The tracking task simulates keeping headway distance in car following situation, and detection response task simulates the evasive braking to sudden brake of a leading vehicle. To verify this method can detect the differences of driver distraction, visual-manual tasks and auditory-vocal tasks were conducted with performing the PT-DRT. Also, we compared the PT-DRT results with the driving simulator (DS) experiment.
Results:
First, we found significant differences of PT-DRT performances between workload levels of the secondary tasks. These results were shown in both visual-manual tasks and auditory-vocal tasks. Second, we also found the similar tendencies in the DS results, which are headway keeping and detection response performances. Third, we found a strong correlation between response time of PT-DRT and response time in DS experiment. And also, a medium correlation between PT performance and headway keeping performance of DS was found, but it was not statistically significant.
Limitations of this study:
The conditions of the PT-DRT method and DS protocol in this paper were not optimized for the highest sensitivity in terms of difference in level of workload. Therefore, we might not find a significant correlation between PT performance and headway keeping performance in DS experiment. Also, limited secondary tasks were used in this study. For example, we did not conduct tasks which combine visual-manual and auditory-vocal interfaces.
What does the paper offer that is new in the field in comparison to other works of the author:
This method is a new easy-to-use bench method for assessment of driver distraction of both visual-manual and auditory-vocal tasks. And the results indicate that PT partly reflects the headway keeping of actual driving and DRT partly reflects the detection of some stimuli and response maneuver in actual driving.
Conclusion:
We showed that the PT-DRT is capable of evaluating the impact on driving operation from the workload of visual-manual tasks and auditory-vocal tasks. These results indicate that this method can be used for simultaneous evaluation of impacts on tracking tasks and detection response tasks. We think these two tasks are important maneuvers in actual driving situation. This method also enables evaluation with equipment that is convenient to use and readily obtainable, and so, it would be useful to test in actual product development of IVISs.
KEYWORDS – visual distraction, cognitive distraction, bench test, pedal tracking, detection response task