Artificial prehension and the detection of object slip

Abstract

An investigation of the fundamental operation of an artificial hand has been carried out with the aim to study the automatic control feedback for this system. The hand is modeled as a simple prehension system with three sensing elements; force, acceleration and slip. This prehension study focused on the object's gripping and slippage processes. An automatic closed loop feedback control algorithm is developed to re-grip the object when it starts to slip which is similar in form to Hooke's Law. The algorithm uses information from the distance of the object has slipped to re-grip the object and control the amount of force required. Also, a method called approximate entropy has been used to analyze and detect when the object begins to slip. This method can be used to prevent the object from slipping. © 2009 Springer-Verlag.