This paper investigated how different types of automation affect tactical driving behavior, depending on trust in the system. Previous research indicates that drivers wait for automation to act, delegating the monitoring of traffic situations. This would be especially true for those who have more trust in automation. Behavioral and gaze data from 30 participants driving an advanced simulator were recorded in four driving conditions, namely, manual driving, intentional car following, adaptive cruise control (ACC), and ACC with adaptive steering. Measures of trust in the systems were recorded with a questionnaire. Three fairly common traffic events requiring a driver response were analyzed. Trust in automation was high among the participants, and no associations between trust levels and behavior could be found. Drivers seem to make informed choices on when to let the automation handle a situation and when to switch it off manually or via the vehicle controls. If drivers did not expect the system to be able to handle the situation, they usually resumed control before the automation reached its limits. If the automation was expected to be able to deal with the situation, control was usually not resumed. In addition, situations were dealt with in a tactically different manner with automation than without. Controlling the car with automation systems is thus accepted by drivers as being a different undertaking than driving in manual mode.
|Tidskrift||IEEE Transactions on Intelligent Transportation Systems|
|Status||Published - 2014|