In the Loop of Safe Driving: An Assessment of HMI Strategies Enabled by Intelligent Driver Monitoring Systems with Daily Drivers

Leveraging the driver state classification performed by state-of-the-art intelligent driver monitoring systems, new multimodal Human Machine Interfaces (HMIs) strategies can be designed to support safe driving. With the purpose of engaging drivers in safe driving behaviors by keeping them aware of their driving state, visual nudges, voice interaction, ambient lights, and music have been exploited to design an HMI prototype of that kind. This study presents the results of a focus group performed with daily drivers to assess the proposed HMI approach in terms of its perceived usefulness, engagement and explainability in partial autonomous driving scenarios. The feedback gathered shows that the proposed solution matches participants' expectations in supporting emotion regulation and attention to keep a safe driving state. Furthermore, it highlights the need for the design of different types of explanations of the driver complex state as well as engagement techniques to adapt to personality traits and mood.