A Study on the Effects of Different Interaction Modalities on Driving Trust in Automated Vehicles

The growing capabilities of Intelligent vehicle interaction systems contribute to the intricacy of interaction methods, subsequently influencing drivers’ trust. This study comprehensively examines prevailing interaction modes of Intelligent vehicles, employing simulated driving experiments and evaluating participants’ trust levels using an automated trust scale across diverse interaction modes. The experiment introduces three specific scenarios: traffic congestion, traffic accidents, and pedestrian crossing. During the formal experiment, participants navigate a simulated urban route, encountering random instances of the specified scenarios during autonomous driving. The interaction system issues prompt when these situations arise. Four distinct interaction modes were crafted: baseline, visual, auditory, and audio-visual. The experimental findings reveal a notably heightened level of human-machine trust in the combined audio-visual interaction mode in contrast to voice and visual interactions. Furthermore, visual interaction is more effective in enhancing driver trust than auditory interaction. In visual interaction, the HUD-based visual interaction significantly elevates driver trust compared with dash-board and central control screen interfaces. This research offers insights for refining the interaction design of autonomous driving vehicles.