Motion Planning In Industrial Design

Abstract

Automation of the design and manufacturing process of complex mechanical systems could save countless man-hours and reduce costs. The development of practical motion planning algorithms could aid this automation by increasing the power of computer-aided design tools so that they can be used to perform manufacturability and maintainability studies. In this paper we present a new motion planning method called single-shot obstacle-based probabilistic roadmap method, or ssOBPRM. This new method is based on previous motion planning methods but addresses observed weaknesses in these methods for the industrial applications we consider. This new method is aimed at efficiency, only performing enough processing to obtain a collision-free path from one robotic configuration to another or else to determine that a path cannot be computed in a reasonable amount of time. A description of this algorithm is provided along with insight into its performance and possible future improvements and directions upon this project. From our experimental results we hope to establish that this new planning method is indeed successful and more efficient than existing ones, and therefore a viable motion planning algorithm for use in computation intensive design and manufacturing studies necessary in industry.