Abstract
The widespread adoption of automation to work alongside humans has opened opportunities to integrate soft robotic grippers with a sense of touch for executing a wide variety of tasks. In this work, we demonstrate a visuo-tactile sensor for soft grippers, which is based on cholesteric liquid crystal elastomers (CLCE) that uses an embedded camera-vision system to convert the force distribution on the contact surface into an RGB map, to detect and distinguish the applied force without complex transmission and processing of electrical signals. A new method was developed for rapid and cost-effective fabrication of wafer-scale CLCE sensing layers with high sensitivity and excellent mechanochromic performance. By integrating the CLCE sensing layer with a 3D-printed array of microtips, mili-Newton forces can be detected by color changes with a resolution of similar to 2 mN. Through an optimization algorithm, the computational load of color image processing is minimized, and real-time fast processing of color signals and high-precision tactile force prediction are performed using a low-cost Raspberry Pi 4 camera. The processed signal guides a soft robotic gripper to complete grasping actions through a closed-loop force control. The integration of such mechanochromic films with suitable mechanical actuators could lead to miniaturized and untethered tactile sensors.