System and method for trajectory planning for manipulators in robotic finishing applications

#This specification relates to systems and methods for automatically moving a tool using a robotic manipulator.

#Many manufacturing applications, such as robotic finishing, require robotic manipulator(s) to operate in a common workspace on same (or different) workpiece(s). During the operation, each manipulator considers the current state of the workspace and plans a collision-free trajectory in order to accomplish its desired task. Traditionally, the motions of the robotic manipulators were predefined and programmed by the expert human operator while deploying these manipulators on mass production lines. In case of any error in the operating environment, the whole process (or assembly line) was suspended into a standby mode before resolving the errors. This procedure causes large delays and leaves very small margin of error.

#The National Association of Manufacturers estimates that the US has close to 300,000 small and medium manufacturers (SMMs), representing a very important segment of the manufacturing sector. Small volume production, which represents a significant segment of SMM, uses highly automated processes to convert raw material into individual parts. Examples include 3D printing, CNC machining, laser cutting, injection molding, etc. The remaining operations are largely manual and often labor intensive. Examples include machine loading/unloading, inspection, cleaning, and assembly. In contrast, these manual operations are often performed by robotic manipulators on mass production lines. This clearly shows the potential of robotic manipulators in manufacturing. However, current generation of robotic manipulators are not considered useful in small production volume operations due to long programming time, and low reusability. Hence, SMMs in the US have largely stayed away from using them.

#Other systems, methods, features, and advantages of the present invention will be apparent to one skilled in the art upon examination of the following figures and detailed description. Component parts shown in the drawings are not necessarily to scale, and may be exaggerated to better illustrate the important features of the present invention.

#FIGS. 1A-1E illustrate a part, a robotic manipulator, and a tool, according to various embodiments of the invention.

#FIG. 2 illustrates the interaction between the collision detection engine, the sampling of configuration space, the kinematics solver, and the optimizer with the planning algorithm, according to various embodiments of the invention.

#The novelty of the systems and methods described herein is an anytime graph search-based trajectory planning algorithm that computes collision free trajectories for the high DOF manipulation system consisting of one or more robotic manipulators in real-time. Several enhancements have been developed for the planner that improves the computational efficiency of the planner and enables it to compute plans in real-time.

#The workspace analysis described herein enables the planning algorithm to efficiently compute the forward kinematic solution in constant time and is independent of the DOF of the manipulation system. The improved collision detection approach described herein reduces the time required to evaluating the collision between during the search and sampling of configuration space.

#The computational performance of the planning algorithm is improved by automating context dependent dynamic switching of heuristic functions to efficiently guide the search towards the goal state, automating switching and scaling of control action primitives by estimating the congestion around each state, and reducing the number of nodes expanded by automated elimination of nodes using the diversity criteria.