عنوان مقاله [English]
In this paper, we consider a dyadic supply chain with an uncertain supplier. Most existing studies in supply chain and inventory literature assume that the supplier is continuously available to get the order from the retailer. In this study, this assumption is relaxed and we assume that the supply process is unreliable. An unreliable supplier alternates between available (ON) and unavailable (OFF) states. The ON and OFF durations are considered to be independent exponential variables. Both the retailer and the supplier use a (R, Q)-type continuous review policy. The retailer faces Poisson demands. The transportation time between the outside supplier and the supplier, as well as the transportation time between suppliers and the retailer, are assumed to be constant. The lead time that the retailer experiences is a non-zero random variable, which is composed of constant transportation time and a random delay that occurs due to lack of inventory at the supplier. The shortage at the retailer is backordered, and delayed retail orders are satisfied on a first-come, first-served base. The main contribution of this paper is considering both the integrity and uncertainty in the above supply chain. Using the idea of the one-for-one ordering policy cost, we derive a reliable approximation for the total cost function of the described system, as a weighted mean of costs of a one-for-one ordering policy. Finally, using simulation studies, we show that absolute errors are ignorable.