Designing Vibrotactile Scales Based on Human Perception for Encoding Abstract Data

This study presents a perceptually grounded method for designing vibrotactile scales. These scales consist of structured sets of calibrated vibrotactile patterns to facilitate reliable differentiation between the values and help users perceive and interpret variations in stimulus through touch. Systematic methods for designing vibrotactile patterns and scales that are perceptually valid and effective for data encoding are lacking. To address this, we investigated vibrotactile properties, frequency, duration, and rhythm to design perceptually distinct scales that support user identification and ordering tasks. We conducted an iterative experimental process involving the estimation of perceptual limits for frequency and duration, the construction of perceptible scales, and user evaluation of the proposed scales. We created a custom vibrotactile device with multiple actuators for experimental support, and a set of three scales using parameters: frequency (four, five, and six values) and duration/rhythm (four, six, and eight values). Participants identified all vibrotactile patterns with 100% accuracy. Ordering task accuracy reached 100% for four value scales, 91.6% for six value scales, and 88.1% for eight value scales for some properties. The findings indicate that lower-granularity scales yielded smaller perceptual errors and greater consistency, whereas higher-granularity scales increased perceptual load and error dispersion. These results validate the use of perceptual limits for designing vibrotactile scales and support their integration into vibrotactile interfaces for encoding data.