…the flexible open source driving simulation

Reichel, S., Sohn, J., Ehrlich, U., Berton, A., & Weber, M. (2014, June). Out-of-Domain Spoken Dialogs in the Car: A WoZ Study. In 15th Annual Meeting of the Special Interest Group on Discourse and Dialogue (p. 12).

Mobile Internet access via smartphones puts demands on in-car infotainment systems, as more and more drivers like to access the Internet while driving. Spoken dialog systems (SDS) distract drivers less than visual/haptic-based dialog systems. However, in conversational SDSs drivers might speak utterances which are not in the domain of the SDS and thus cannot be understood. In a Wizard of Oz study, we evaluate the effects of out-of-domain utterances on cognitive load, driving performance, and usability. The results show that an SDS which reacts as expected by the driver, is a good approach to control incar infotainment systems, whereas unexpected SDS reactions might cause severe accidents. We evaluate how a dialog initiative switch, which guides the user and enables him to reach his task goal, performs.

The open-source driving simulator OpenDS provides a driving environment and extensive logging facilities (Math et al., 2012). As explained in Section 2, it is essential to keep the driver occupied at a constant level all the time. Therefore, we used the ConTRe task (Mahr et al., 2012), which consists of a continuous steering task and a reaction task.

The movement of the yellow cylinder and the appearance of the traffic light can be controlled by manipulating OpenDS’ control variables. We used the “hard driving” condition as described by Mahr et al. (2012).

Therefore, we used four kinds of measurements (see Table 2): objective driving performance logged by OpenDS, subjective driver distraction with DALI questionnaires, usability scores measured by SASSI questionnaires, and dialog performance.