Thesis Type: Postgraduate
Institution Of The Thesis: Middle East Technical University, Faculty of Engineering, Department of Industrial Engineering, Turkey
Approval Date: 2005
Thesis Language: English
Student: Yeşim Özalp
Supervisor: GÜLSER KÖKSALAbstract:
This thesis attempts to integrate the three supplier processes- supplier segmentation, supplier selection/evaluation and order allocation- based on a case study of a white goods manufacturer. These processes are dealt with in a hierarchical manner, where the decisions at an earlier stage (supplier segmentation) may affect on the results of later stages (supplier selection/evaluation and order allocation). Based on a wide set of variables gathered from the supplier segmentation literature and from the framework developed by the case company, a factor analysis is performed. The resulting eight factors (complexity and criticality, supply market dynamism, supplier̕s economic dependence, buyer̕s supplier dependence, uncertainty, supplier capabilities, supplier̕s specialization to white goods manufacturer and possibility of harm to user) are used for clustering analysis with the K-means method. Three different clustering schemes (4-clusters, 8-clusters, and 18-clusters) are analyzed to see the effect of number of clusters on the cluster means. A supplier selection model is constructed by the PROMETHEE method as was done in a former study at the same company. Two different scenarios are considered in the supplier selection phase: First, data from the former study are kept as they are. The second scenario considers a change of criteria weights. As a result, the qualified suppliers were found to be the same in both scenarios, but the rankings and net flows, which were inputs to the order allocation model, changed. The effects of these changes were discussed at a conceptual level, due to a lack of case data for the order allocation model.