An efficient motion planning algorithm for robot multi-goal tasks

Abstract

Motion planning is a fundamental aspect of every robotic system. The objective is to generate an efficient and collision-free path for the robot to reach a certain number of reference points. This paper presents a method to define a near-optimal sequence of target positions for a robot executing a multiple goal task. In the described context, optimal means that the robot reaches its target positions in a sequence that minimizes the total execution time of the task, visiting each goal once. The proposed technique takes into account also collisions with possible obstacles and objects in the robot work area, searching for the minimum execution time collision-free path for the robot. The result of the proposed work is a motion planner that has been successfully tested and verified both in simulation and in a real industrial application. The results achieved show its effectiveness and usefulness. Moreover, the proposed planner can be easily extended to many other different areas and sectors, also beyond the industrial context.