Multiple sourcing decisions using integrated AHP and knapsack model: a case on carton sourcing

Abstract

Purchase allocation is a multi criteria decision making (MCDM) problem. Multitude of qualitative and quantitative factors is involved in the multiple sourcing decisions. Analytic hierarchy process (AHP) has widely been used to find out the relative rankings of suppliers. AHP can be combined with regular supplier quantitative audit process. In classical AHP, decision maker (DM) has to pair wise compare suppliers for each factor, whereas the proposed audit based simplified AHP will remove the complexity of comparison. Quarterly audit-based AHP rankings and supplier performance probability products can be used in place of revenues in the backward recursive resource allocation knapsack model. This combined model will decompose purchase allocation problem into different stages and combine one supplier at each stage and provide the optimum and feasible solution in the end. Solution at each stage is also a feasible option. This model is only applicable when the total order quantity and the capacity of all suppliers are integer multiples of economic or minimum order quantity. This integrated model thus provides number of orders/supplier.