Manual guidance of manipulators: detection and control of intentional contacts without force sensors

The core idea of this study is developing a reliable method to manually guide robots in the physical cooperation scenarios, having human safety in priority. To this end, a safe human–robot cooperation algorithm in the presence of physical intentional interactions without employing force and vision sensors is developed. The physical interaction is recognized and controlled through available common sensors. Contact occurrence and its position are investigated by using online external torque vector obtained by residual based observer. Then, the force vector is estimated and employed in the suggested controller for manually guiding the robot. To completely handle the cooperation phase, four guiding scenarios based on actual conditions are recommended. Performance of the proposed controller-observer method during manipulation has been validated through multiple experiments on a KUKA LBR iiwa.