Robot Path Planning

Planning for a robot to execute a task from abstraction to concretion can be roughly divided into three stages: task planning, path planning, and motion planning. The task planning divides a complex task into several simple subtasks. For instance, for a forage robot to execute a task of grasping food back to home, this task may be divided into subtasks of searching food, moving toward the food if found, grasping the food, moving the food back to home, and putting the food at home. For executing each subtask, path planning needs to produce a path for the robot to execute. For example, for executing the subtask of moving toward the food, the robot needs to calculate executable paths from its current position to the food position and determine which route to take if multiple paths exist (generally the shortest path has the highest priority). Similarly, for executing the subtask of grasping the food, the robot needs to produce a path for its arm to hold the food. Finally, the motion planning based on the planned path to plan its kinematic and dynamic motions for the robot to execute. The main focus of this article is path planning, and part of task planning and motion planning is also mentioned as they are related. Keywords: hierarchical robotic architecture; behavior-based robots; hybrid robots; multiple robot systems