Shadow robot for teaching motion

Human-friendly robots have begun to spread in society. In the future such robots and intelligent machines should be autonomous in open situations. To give dexterity to a robot, teaching motion is a good candidate. However, there are some problems from the operational point of view due to gravity and friction effects. In this paper, a shadow robot is proposed for teaching motion instead of force sensors. The shadow robot is a novel disturbance compensation method that consists of a twin robot system. Two of the same type of robot are required and they are controlled with the same position, velocity, and acceleration by bilateral acceleration control based on a disturbance observer. One robot is in the teaching motion controlled by a human and the other is unconstrained. Thus the purity of the human force is extracted by subtracting the disturbance torque in the unconstrained robot from the constrained one. As a result, the shadow robot observes the human force with gravity and friction compensation. Since it is possible to apply this concept to a multi-degree-of-freedom system, the human operationality in teaching motion are improved. The experimental results show the viability of the proposed method.