Analytical Modeling and Experimental Validation of a Gelatin-based Shape Sensor for Soft Robots

Abstract

Shape sensing of soft robots has been a challenge due to the large deformation of the soft robots and their low stiffness. In this study, a simple yet accurate soft sensor for soft robotic application with small force ranges was proposed, modeled, prototyped, and experimentally validated. The proposed soft sensor is based on a gelatin-graphite composite that exhibited piezoresistive properties. The sensing element was molded to a cylindrical shape and was embedded in a soft flexural structure as a common type of soft flexural robots. Afterward, a mechano-electrical model for predicting the changes in the resistance of the sensing element was proposed and its predictions were validated through an experimental study. To demonstrate usability for force sensing, the sensor was calibrated with a nonlinear model and exhibited a force measurement range of 0.035-0.82N with an average absolute error of 3.7% and a resolution of 4%. Also, the mechano-electrical model was fairly accurate in predicting the piezoresistivity phenomenon of the sensing element under large bending deformations.