User-centered multimodal interaction design for autonomous vehicles: a focus on cognitive load and accessibility for users with severe visual impairment

User-centered multimodal interaction design for autonomous vehicles: a focus on cognitive load and accessibility for users with severe visual impairment

연구 분야: Artificial Intelligence

학회: Universal Access in the Information Society

This study investigated the optimization of adaptive design in autonomous driving technology for Severe Low Vision Persons (SLVPs), focusing on enhancing environmental perception, cognitive load, and overall user experience through multimodal interaction design (voice prompts, tactile feedback, and spatial audio). Five information delivery modes were evaluated: (1) voice prompts, (2) tactile feedback, (3) spatial audio, (4) voice + tactile feedback combination, and (5) tactile + spatial audio combination. A total of 25 SLVP users participated in the study, assessing their accessibility, cognitive load, and user experience within a simulated environment. The results indicated that, compared to single-modality designs, multimodal interaction design improved environmental perception, reduced cognitive load, and enhanced user experience. Notably, the tactile + spatial audio combination demonstrated the best performance in information perception and substantially enhanced user satisfaction. This study confirmed the advantages of multimodal interaction in adaptive design for SLVP and provided empirical evidence supporting the optimization of autonomous driving interaction systems. Future research should further assess the effectiveness of multimodal designs in real-world driving conditions and expand the sample size to explore the adaptability differences across various visually impaired groups, thereby offering more inclusive design strategies for the barrier-free development of autonomous driving technologies.