Collision-Free Algorithm for Industrial Robots Based GJK and OBB Theories

This paper proposes a better way for industrial robots to avoid collisions while manipulating objects. The method uses the Oriented Bounding Boxes (OBB) theory. OBB is effective because it can work with objects turned in any direction, which makes it very exact. To detect if two OBBs are colliding, the algorithm adopted double-check collision detection to achieve a collision-free workspace. First, Threshold distance collision detection by calculating the nearest line between an object’s corners. Then by the Gilbert-Johnson-Keerthi (GJK) algorithm. The algorithm uses Z-Y-Z Euler's angles to handle orientation between robot grippers and objects during paths. This updated algorithm will make industrial robots safer and more efficient when handling objects. We have adopted Coppellia Simulation software to verify the accuracy of the algorithm with different scenarios.