Collision avoidance and path planning for industrial manipulator using slice-based heuristic fast marching tree

Abstract

With the emergence of Industry 4.0 , high productivity is critically dependent on robot manipulators . However, building an efficient and safe work environment with robot manipulators remains a challenge of hardware capability. The optimal path planning of the robot manipulator usually encounters shortcomings in low computational speed and tedious training after changing assembly lines and increases the risk during human–robot collaboration (HRC). To solve such a problem, we propose a path planning, named slice-based heuristic fast marching tree, based on joint space to achieve real-time path planning speed without modeling or training the workspace in advance. Our experimental results indicate that the time consumed for path planning in static environments is only from 0.51 s to 1.63 s, tested by a 6-DOF general-purpose industrial manipulator and different cylindrical obstacle placements. The time for path replanning in dynamic environments is from 0.62 s to 0.88 s. • Path planning is in 1.63 s with static obstacles and 0.88 s with dynamic obstacles. • The proposed path planning is faster, smoother, and shorter than RRT, RRT*, and FMT*. • The joint angles of revolutions in total are less than RRT, RRT*, FMT*.