A manufacturing process often consists of many procedures, with a certain feature being added to the product in each procedure. Normally, these procedures need to be completed one after another according to a predetermined sequence. It is widely recognized that the sequence would affect not only the variabilities of production volumes but also the long run average cost of the manufacturing system. This paper aims to determine the circumstances under which we can reduce the long run average cost of the manufacturing system by reversing two consecutive procedures of the manufacturing process. We formulate an inventory model for products with two or three features. We provide several sufficient conditions by using the technique of stochastic order for determining whether operations reversa...[ Read more ]

A manufacturing process often consists of many procedures, with a certain feature being added to the product in each procedure. Normally, these procedures need to be completed one after another according to a predetermined sequence. It is widely recognized that the sequence would affect not only the variabilities of production volumes but also the long run average cost of the manufacturing system. This paper aims to determine the circumstances under which we can reduce the long run average cost of the manufacturing system by reversing two consecutive procedures of the manufacturing process. We formulate an inventory model for products with two or three features. We provide several sufficient conditions by using the technique of stochastic order for determining whether operations reversal could really optimize the production process in terms of reducing costs. These sufficient conditions imply that the less variable feature should be processed first. We also provide counter examples proving that the criteria of a small total variability itself can not ensure a low operating cost.