Tool Compensation and Force Password Identification on Admittance-Controlled Robots in Walk-Through Programming

This paper presents an example of a walk-through programming technique that exploits the payload compensation to execute the teaching phase of the robot programming. During this phase the human operator grabs the tool attached to the end-effector and physically walks the robot through the desired positions without any effort, since the static and dynamics effects produced by the tool are compensated. In the meantime, the robot controller records all the poses and then it will be able to reproduce the trajectory back. The addition of a “force password” identification allows to start and stop the recording in a simple way, disregarding the use of additional external buttons or software commands. Experiments are used to validate the proposed control architecture and a pick and place example is presented, to show a realistic industrial application.