Evaluating attentional demands of in-vehicle interfaces

Abstract

Touchscreens have become widely adopted in vehicles because they are flexible interfaces that enable easy updates and accommodate advancing technology. However, in-vehicle touchscreens may have detrimental effects on driver attention.  As a result, incorporating some number of physical controls into in-vehicle interfaces may better support driver attention to the driving task. However, there is currently no unified framework for determining how to optimally allocate functions (e.g., navigation control) between different types of inputs like touchscreens or traditional buttons. The allocation of functions between touchscreen and button inputs has substantial implications for how drivers interact with in-vehicle interfaces (i.e., the interaction model). Interaction models are influenced by factors like input type (touchscreen vs. button), input feedback (visual, haptic), and display behavior which may be highly dynamic (where display content may change, like on touchscreens) or primarily static (where information format and location are consistent, like on many vehicle dashboards). To support the design of next-generation vehicle interfaces, we conducted a systematic literature review of studies that collected empirical data about driver interactions with in-vehicle interfaces. We classified the studies based on the characteristics of the interaction models used, specifically examining input type, input feedback, and output behavior. We identified specific measures of attentional demand and driver performance collected within each study. Finally, we identified characteristics of interaction models associated with higher attentional demand and the resulting impact on performance. We used those findings to identify research gaps, make recommendations for future research, and apply these insights to the driving environment.