Handover Assist in Highly Automated Vehicles: How Vocal Communication Guides Visual Attention

Abstract

Automated vehicles that require human intervention will inevitably require the transition of control and responsibility between driver and automation. These ‘handovers’ represent a vulnerability in the driving system due to factors such as reduced situation awareness. As a solution, handover assistants have been proposed to alleviate these drawbacks and facilitate better communication between vehicle and driver. We present findings from a vocal-handover task between two drivers, conducted in a driving simulator, to explore how visually scanning the environment can be encouraged using different vocal interactions. The data revealed trends such as how mentioning location may encourage more efficient visual gaze. Conversely, no vocal interaction may result in little-to-no visual gaze towards certain areas of the driving environment. Further study could explore how vocal interaction can work in conjunction with visual displays to guide visual attention during the handover task.