Supplier selection and involvement in one-of-a-kind production (OKP).

摘要

Global competition is forcing manufacturing companies to develop the capability of efficiently producing customized products with high quality and competitive prices, which is now widely referred to as mass customization. The mass customization capability is particularly critical for manufacturing companies in developed countries to compete with the ones in those low labour cost countries. One-of-a-kind production (OKP) is a manufacturing paradigm for achieving mass customization. It is also observed that the OKP companies are often small and medium sized enterprises (SMEs). Under the pressure of cutting cost and improving efficiency in OKP, a relatively new concept of supplier involvement in product development and production process has attracted more and more attention from OKP companies. Thus, the supplier evaluation and selection, along with the mechanism and methodologies of supplier involvement in OKP, are worth of being researched. It has been found in practice that although relatively small in the batch size, the 20% procurement makes nearly 100% varieties of the products. The purchase of these customized items is critically influencing OKP efficiency and cost. This thesis work aims at development of the methodology for OKP companies to improve their cooperation with the suppliers. A virtual supplier involvement (VSI) concept is proposed. Under this concept, the supplier selection and involvement are considered to be two interrelated stages, viz. early generic stage and late specific stage. In the early stage, the research focuses on a generic evaluation and selection model to support formulating and maintaining a qualified supplier base. In the late stage, the research places its emphasis on developing computational algorithms for OKP companies to quickly select suppliers from the qualified supplier base. In line with the stage classification, a DEA (data envelopment analysis) based piecewise fuzzy AHP (analytical hierarchy process) approach and a multi-objective fuzzy mathematic programming model are proposed. Finally, the feasibility and effectiveness of the proposed models and supplier selection methods are validated and tested through industrial case studies.