A scalable deadlock avoidance algorithm for flexible manufacturing systems with free choice in part routing

A new deadlock avoidance algorithm for flexible manufacturing systems that contain both multiple capacity resources and free choice in part routing is presented. The method determines whether moving a part to its next step is safe, unsafe, or undetermined. This classification is linear in complexity. Part movements that are classified as undetermined are further analyzed using another procedure, which attempts to empty the system to determine whether the move is safe or not. It is polynomial in complexity. By adjusting a simple tuning parameter, the size of the set of possible undetermined classifications returned can be made arbitrarily small at the cost of increasing the order of the algorithm. An example showing the application of the method is provided.