A Low-Cost Super-Resolution Tactile Sensor: Design, Fabrication, and Validation

Abstract

Tactile sensing, which serves as a modality parallel to vision and auditory, provides rich contact information that is irreplaceable by other modalities. Although tactile sensing technology has made great progress over past decades, existing sensors still lag far behind human skin in infinite-resolution sensing, large-area sensing, and thinness. Inspired by the bionic mechanism of human skin that has various receptors embedded into soft tissues, here we design a low-cost and super-resolution tactile sensor by embedding flexible pressure sensors into a soft silicone layer. Different to the traditional functional imitation of human receptors (e.g., Pacinian corpuscle), our focus is on the selection of the soft silicone materials to better mimic the soft tissues of human skin. When an external force is applied on the soft silicone layer, the deformation of the soft silicone can excite the response of multiple pressure sensors, which mimic the Pacinian corpuscles in human skin. Based on experimental data and practical applicability, the optimal parameters for the soft silicone layer are determined to enable more responsed Bionic human receptors for the same contact force. Then the position and magnitude of the normal force are estimated based on a reconstruction algorithm to achieve super-resolution and larger-area sensing. In addition, a human-computer interaction interface for signal collection and tactile real-time display is designed to vividly show the contact status. Experiments show that the tactile sensor can achieve normal force estimation with an average error of 0.61 N and millimeter-level super-resolution localization within a range of $22.8\times 22.8$ mm. Moreover, our sensor is more compact (only 6 mm in thickness) than visuo-tactile sensors (15 mm), which are the current state-of-the-art.