Visual and cognitive demands of manual and voice-based driving mode implementations on smartphones

Abstract

Mobile phone apps and operating systems are increasingly adopting driving mode functions that attempt to reduce driver visual and cognitive demand by limiting functionality, using larger buttons and icons, and adding voice-based interactions. The present study assessed the visual and cognitive demands and the subjective level of distraction from two driving mode implementations (voice or manual) on an Android™ mobile phone using Google Assistant™, compared to a typical mobile phone operating system experience. While driving on a test track, participants performed several trials of five tasks on each of three interfaces: A mobile operating system interface, a manual driving mode interface, and a voice driving mode interface. Visual demand was measured with eye-gaze recordings, cognitive load was measured with the detection response task, and a Likert scale was used to rate the perceived level of distraction. The voice driving mode resulted in the lowest visual attention demand and lowest subjective ratings of distraction. The manual driving mode condition also reduced visual demand and subjective ratings of distraction relative to the mobile operating system condition. The cognitive load results were inconsistent across the task and interaction mode conditions. Overall, the results of this study provide positive evidence in support of voice-based driving mode implementations for reducing visual demand and subjective levels of distraction from mobile devices while driving. Moreover, the results suggest that manual driving mode implementations also have the potential to reduce visual demand and subjective levels of distraction, relative to the mobile operating system condition.