A multimodal interaction-based approach to situational driving anger emotion regulation
Ting Wei, Yuanwu Shi, Jiachen Sun · Expert Systems with Applications · 2025
Driving anger can impair a driver’s perception of the environment, leading to hazardous driving behaviors that threaten driving performance and safety. The regulation of driving anger through multimodal interaction allows intelligent cockpit assistance systems to provide adaptive feedback to the driver’s anger, mitigating potential dangerous actions. To identify reliable and effective anger regulation modalities for drivers, a 2 (with and without anger regulation) x 3 (auditory-only interaction + auditory-visual bimodal interaction + auditory-olfactory bimodal interaction) mixed-factor experimental design was employed, utilizing three distinct modal conditions in a simulated driving experiment. After inducing driver anger, different interactions were provided: no feedback, auditory-only materials, auditory-visual bimodal materials, and auditory-olfactory bimodal materials for emotion regulation. The Self-Assessment Manikin (SAM) scale was used to evaluate anger levels before and after the interaction, and the Likert scale was used to assess the effectiveness of the anger regulation. The results indicate that multimodal interaction significantly positively modulates driving anger. The average pleasure level increased from 2 to 6, reflecting a transition from anger to pleasure. Our findings suggest that different interaction modalities can provide positive information for driver emotional regulation during episodes of driving anger. The interaction modality combinations identified here can serve as outputs for future adaptive driving assistance systems, which can provide adaptive interaction feedback after detecting driver anger to alleviate it.