A staged supplier pre- evaluation model to determine risky, potential and preferred suppliers

Abstract

In the late 1980s, the proportion of outsourced materials in the cost of high-tech products was around 80%. In this respect, with increasing globalization and ever-expanding supply chains, interdependencies between organizations have increased and the selection of suppliers has become more important than ever. This exploratory research study intends to develop a novel approach for a specific type of supplier selection problem which is supplier pre-evaluation. A two-staged multi-layered feed forward neural networks (NN) algorithm for pattern recognition was used to pre-evaluate suppliers under strategybased organizational and technical criteria. Data for training, validation and testing the network were collected from a global Tier-1 manufacturing company in the automotive industry. The results show that the proposed approach is able to classify candidate suppliers into three separate groups of risky, potential or preferred. With this classification, it becomes feasible to eliminate risky suppliers before doing business with them.