Friction coefficient estimation before gross slip for slippage prediction during prosthetic hand grasping

Slippage feedback is critical for prosthetic or bionic robotic hand to realize steady grasp. An important way to judge the slippage is to detect the maximum coefficient of static friction (mu s) which is commonly measured when or after gross slip happened. Here, we proposed a new method to estimate the mu s before the gross slip in order to realize slippage prediction, leaving more time for the control system to adjust grasp force. Based on a finite element model, a good linear relationship (R2 = 0.99) between the global tangential force at the time when relative slip occurs on the monitoring point (GTFRS) and the mu s was obtained. An optical experiment was conducted and verified the accuracy of the simulation result. Finally, a slip sensitive soft fingertip was fabricated and tested, results show mu s estimation in samples of seven different roughness is well realized with average error of 6.68 %.