Methodology to Analyze Adaptation in Driving Simulators

Abstract

Adaptation to a driving simulator is one of the necessary conditions for the validity of almost all driving simulator studies. Learning how to control a simulated vehicle requires practice, which will put some mental load on drivers and can distract them from their main task (i.e., driving). Such distraction during the learning phase affects drivers’ reactions to different situations compared with what they would have done in the real world; therefore, having a tool to confirm that adaptation occurred is necessary to have accurate data. This methodology needs to be sensitive to the diversity of driving styles and applicable to a variety of driving tasks and performance measures. A comprehensive review of the literature indicated a general deficiency in the methodologies being used. Common approaches include having participants drive for a predefined time or driving distance or drive until the participants report that they feel comfortable. Such approaches do not ensure that adaptation has indeed occurred. This paper proposes a methodology to evaluate adaptation by analyzing driver performance measures. The methodology is based on the concept of experience curve effect and is intended to complete and further develop the idea of the learning curve effect that was used in the authors’ earlier research. The methodology was tested for adaptation to acceleration and braking tasks using UBCDrive driving simulator. The results indicated that experience and learning curve effects can be used to identify adapted, adapting, and nonadapting participants.