Updating Potential Runaway Motion Volume in Task Space and Introducing Minimum Volume Enclosing Ellipsoid

Human-robot collaboration (HRC) has gained considerable interest among researchers, particularly in the manufacturing field. In such systems, new risks that have not previously been considered may now need to be assessed. One of these risks is that of a robot and worker colliding with one another. To avoid such collisions, a definition of the separation distance between a robot and a worker is needed. According to the ISO10218-1 standard, the safety of a worker needs to be maintained even in the case of a single safety system fault. Therefore, in defining the separation distance, a single fault needs to be considered, so that a robotic manipulator may safely run away when it occurs. As a method to quantify running away, a potential runaway motion in the task space (PRAM-t) was proposed. In the PRAM-t, a dynamic manipulability ellipsoid was used to calculate the acceleration as the robot runs away. However, when we consider running away originating from a single fault, it can be exaggerated. Therefore, this paper proposes a new method to compute acceleration during the running away of a robot using an algorithm to compute minimum volume enclosing ellipsoids. The proposed method and the dynamic manipulability ellipsoid (DME) method are compared using a case study.