Live Demonstration: Dynamic Grip-Force Control Using Real-Time Friction Estimation From Incipient Slip Events

Robotic/prosthetic gripping systems fall far short of achieving human-like dexterity. One major limitation of these systems is a lack of adequate tactile sensing in the gripper fingers - feedback required for grip-force control. The PapillArray is a sensor array of soft silicone pillars that can measure 3D deflection, 3D force and 3D vibration at each array element, and torque, incipient slip, and friction - important emergent tactile parameters for robotic dexterity. In this work we demonstrate real-time measurement of friction and 3D force from PapillArray sensors mounted on each finger of a two-finger gripper. A closed-loop grip-force control algorithm uses the sensor output to dynamically determine the target grip force. A simple lifting task using different objects (of varying weights and surface materials) demonstrates that knowing friction allows the grip force magnitude to be optimized to avoid dropping the grasped object; as opposed to applying a fixed, pre-determined grip force. Future work will incorporate real-time torque measurement into the grip force feedback control. This will significantly advance the state-of-the-art in artificial tactile sensing for robotic dexterity.