What Is the System Doing Now? Designing a Spatially Differentiated, State-Explicit HMI for Clear ADAS Boundaries

Abstract

Advanced Driver Assistance Systems (ADAS) increasingly combine longitudinal and lateral functions, complicating drivers’ understanding of which assistance is active, limited, or requesting intervention. This paper translates findings from a naturalistic on-road study of first-time interaction with a new ADAS interface into a spatially differentiated, state-explicit Human–Machine Interface (HMI). Think-aloud comments and post-drive interviews revealed recurring difficulty distinguishing control dimensions, interpreting state transitions, recognizing Operational Design Domain limits, and understanding system logic without experimentation. These findings were synthesized with Ecological Interface Design, Skills–Rules–Knowledge principles, and benchmarking of ADAS interfaces. Through participatory design with three human-centered designers, mid-fidelity prototypes were developed for Adaptive Cruise Control, Stop-and-Go, Steering Assist, and Lane Keep Assist. The concepts use distinct visual zones, progressive transition cues, persistent boundary signaling, and redundant non-color cues to support mode awareness, system understandability, and recognition of driver responsibility. Future work will evaluate interpretability, workload, accessibility, and transparency with ADAS experts.