A fuzzy inference approach to control robot speed in human-robot shared workspaces

According to the current safety standard ISO 10218, the video shows a typical SSM (Speed and Separation Monitoring) collaborative scenario. Human-robot interaction is made possible by properly scaling robot velocity when indispensable and by using an advanced perception system to monitor collaborative operations and to compute the separation distance (S) between the robot and the operator. A fuzzy inference system, based on a risk analysis as allowed by the ISO/TS 15066, has been developed to compute the minimum protective separation distance and adjust the robot speed by considering different possible situations.

The graph on top compares the constant minimum protective distance computed according to the ISO/TS 15066 (red line) with the novel solution. When the separation distance between the robot and the operator, d (blue line), is less than the warning distance (yellow line) and more than the proposed minimum protective distance, S (green line), the robot slows down. When d is less than S, the robot completely stops. The novel approach avoids any collisions between operator and robot and minimizes cycle time as well.

The bottom graph shows the inputs and output of the fuzzy inference system: the time derivative of the distance between human and robot (blue line) and the scalar product between the robot and the operator velocity vectors (red line) are combined to compute S through the set of fuzzy rules. Then, d is compared with S to compute the robot speed scaling factor (green line).