A Hybrid Quantitative Method for Evaluating HMI Layout Design in Service Robots

Evaluating the human–machine interface (HMI) of service robots remains challenging due to the complex integration of perceptual aesthetics and functional rationality. To address this, we propose a hybrid multidimensional HMI evaluation method that quantifies three key dimensions—layout aesthetics, color aesthetics, and functional layout rationality—by integrating visual cognition theory and axiomatic design (AD). The framework operationalizes five layout principles (balance, proportion, unity, regularity, density) and a four-component color model (color difference, distribution, harmony, and personality), complemented by a biologically grounded metric—visual perceptual intensity (VPI)—derived from cone cell response theory. Subjective weights from expert judgments (via analytic hierarchy process, AHP) and objective weights from the entropy weighting method (EWM) are fused within an AD-based information axiom framework to enable balanced, data-driven assessment. Applied to five candidate HMIs for a medical service robot (N = 15 participants), the method identified the design scheme x₃ as optimal when the balancing coefficient α ≥ 0.5 (reflecting greater emphasis on subjective judgment), whereas design scheme x₂ was preferred when α < 0.5 (prioritizing objective data). Given the modest sample size, correlation analysis revealed moderate-to-large—though not reaching conventional significance—between evaluation indicator scores and eye-tracking behavior: unity correlated with total fixation duration (Pearson_r = 0.682), and color harmony with first fixation duration (Pearson_r = 0.788), suggesting alignment between design attributes and visual attention patterns. These preliminary findings suggest that key design attributes may influence visual attention patterns, supporting the framework’s potential to link aesthetic and visual choices to measurable perceptual outcomes.