We are developing a direct teaching system for a multifingered robot hand. This system teaches instinctively by manipulating an object with an attached sensor. The multifingered robot hand then realizes a manipulation similar to the input manipulation. In this paper, we propose a recognition method for object manipulation using contact state transition on an object's surface, where a pressure distribution sensor is directly attached to the manipulated object. The manipulation is recognized by the orbital motion of contact points generated by rolling contacts. To evaluate the proposed method, we conduct a recognition experiment, the results of which shows the effectiveness of the method and the importance of the contact information. Target tasks involve in-hand manipulation, which can change the position/orientation of the object using only the motion of the fingers
Published in:
Multisensor Fusion and Integration for Intelligent Systems, 2006 IEEE International Conference on
Date of Conference: Sept. 2006