A multi‐attribute preventive replacement model

Abstract

We formulate a multi‐attribute decision model for preventive replacement of a “magnetic sealing head” in a soft‐drink producing factory in the Kingdom of Saudi Arabia. In case of failure of this part, the opportunity cost (for production losses) is very important, as the entire production line will be idle. We determine in a first case the replacement policy that minimizes the total expected unit cost of replacement (preventive and corrective). Next, we determine the optimal policy that maximizes the expected multi‐attribute utility of the decision‐maker in the factory. Four attributes are considered in the replacement problem, namely cost, quality, labor productivity, and cash flow availability. The optimal policy in each case outperforms by far the one applied in the plant, which turns out to be costly and inefficient with respect to the utility of the decision‐maker.