Realization of Controllers for Manipulation of Deformable Objects Based on Hybrid Automaton and Human Skill

Abstract

Requirement of handling deformable objects such as leather, paper, and rubber is growing. Since it is very hard to make physical models of them, design of controller to manipulate them becomes one of the significant problems in the robotics field. When we look at the operation of human workers, the deformable objects seem to be handled naturally and smoothly because of human skill. This paper presents a new method to design a controller for assembly tasks which involve the manipulation of deformable objects. In the proposed method, human skill can be expressed as a hybrid automaton because human workers change the dynamics according to the state of the task. The human skill is extracted as the series of dynamics that human workers used in executing the demonstrated task, and the dynamics are embedded into an event driven controller. The proposed method is applied to a hose insertion task and is verified through experiments.