A new on-line singularity avoidance approach for manual guidance of industrial robots using variable impedance control

This paper studies path generation using manual guidance procedure for industrial robots by considering real-time singularity avoidance. Main feature of the proposed approach is singularity avoidance by variating impedance control parameters in preset distance from singularity in order to warn operator. Robot end-effector is equipped with a force sensor which operator grasps it and produces desired path. The desired end-effector path is generated by operator’s manual guidance for applications such as welding and spray painting and is recorded by robot controller. Robot singular configuration is possible during the manual guidance. So real-time detection of singularity position and orientation have to be considered during path generation because it can lead to unexpected high robot joints velocity. This problem is not safe due to physical human-robot interaction. Manipulability ellipsoid method is utilized so as to singularity identification. The method can be utilized in on-line due to its simple and low calculation process. On the other hand, the end-effector velocity is saturated in a specific value in the approach considering safety issues. Two main advantages of the proposed approach are real-time application and high safety because of the singularity avoidance. Experiments are applied on a SCARA robot to study the effectiveness of the proposed approach. Experimental results show the ability of proposed approach in dealing with singularity problem during the manual guidance.