On the optimization of a single-stage generalized kanban control system in manufacturing

We consider a manufacturing system that uses a pull control mechanism to determine when to release raw parts at its input based on actual customer demands for finished parts at its output. This pull control mechanism depends on two parameters: the number of kanbans, which are cards authorizing the release of raw parts into the system, and the base stock of finished parts. We call this system the single-stage generalized kanban control system (GKCS). The conventional single-stage KCS is a special case of the single-stage GKCS in which the base stock of finished parts is equal to the number of kanbans. The single-stage KCS therefore depends on only one parameter. In this paper it is shown that the computational complexity of optimizing the single-stage GKCS is the same as that of optimizing the single-stage KCS even though the former system depends on two parameters, whereas the latter system depends on only one parameter.