Supplier Selection for Mixed-Model Production: A Case Study from the Apparel Industry

The mass-customization trend as flexible production philosophy has been adopted by many industries as a reaction to the fast changing customer demand environment. In relation to that mixed-model production (MMP) is a possibility for increasing flexibility, especially in the assembly department. However, many apparel companies have failed in their transition from traditional manufacturing to new, flexible manufacturing, which makes it difficult for them to be selected as suppliers who are capable of MMP. In most of the studies that have been done before, the problem of supplier selection was based on productivity and cost as the main selection criteria. As an alternative solution this research presents a new group of criteria for supplier selection with respect to the goal of MMP. The methodology of combining the analytical hierarchy process (AHP) with the support of the commercial software package Expert Choice and cluster analysis (CA) is proposed in this study. The results obtained present a new group of criteria, ranked by their importance, that the supplier needs to fulfil in order to be selected for MMP. Furthermore the large number of available suppliers was grouped into four clusters based on their similar characteristics by using CA. Finally the most acceptable cluster for MMP was selected using AHP. The methodology presented in this study can support the apparel industry in better decision making in the process of selecting the best group of suppliers for mixed-model production.

W wielu przemysłach wprowadzono system zindywidualizowanej produkcji masowej jako reakcję na szybko zmieniające się wymagania konsumentów. Dotychczas selekcja dostawców oparta była na produktywności i analizie kosztów. Ten artykuł prezentuje alternatywne rozwiązanie z wyborem nowej grupy kryteriów. Przedstawiono metodologię połączenia metody hierarchicznej analizy z pakietem Expert Choice i analizą klastrów. W proponowanej metodzie dostawcy zostali podzieleni na cztery klastry na podstawie podobieństwa charakterystyk. W efekcie najbardziej akceptowalny klaster zostaje wybrany przy zastosowaniu metody hierarchicznej analizy. Metoda ta może wspomóc prawidłowe funkcjonowanie przemysłu odzieżowego.