Evaluating Multimodal Driver Displays of Varying Urgency

Previous studies have evaluated Audio, Visual and Tactile warnings for drivers, highlighting the importance of conveying the appropriate level of urgency through the signals. However, these modalities have never been combined exhaustively with different urgency levels and tested while using a driving simulator. This paper describes two experiments investigating all multimodal combinations of such warnings along three different levels of designed urgency. The warnings were first evaluated in terms of perceived urgency and perceived annoyance in the context of a driving simulator. The results showed that the perceived urgency matched the designed urgency of the warnings. More urgent warnings were also rated as more annoying but the effect of annoyance was lower compared to urgency. The warnings were then tested for recognition time when presented during a simulated driving task. It was found that warnings of high urgency induced quicker and more accurate responses than warnings of medium and of low urgency. In both studies, the number of modalities used in warnings (one, two or three) affected both subjective and objective responses. More modalities led to higher ratings of urgency and annoyance, with annoyance having a lower effect compared to urgency. More modalities also led to quicker responses. These results provide implications for multimodal warning design and reveal how modalities and modality combinations can influence participant responses during a simulated driving task