Supplier Portfolio of Key Outsourcing Parts Selection Using a Two-Stage Decision Making Framework for Chinese Domestic Auto-Maker
Fuli Zhou,Xu Wang,Mark Goh,Lin Zhou,Yandong He
DOI: https://doi.org/10.1016/j.cie.2018.12.014
IF: 7.18
2018-01-01
Computers & Industrial Engineering
Abstract:The Chinese domestic auto-industry has been urged to perform quality improvement practice due to the soaring warranty cost and quality loss under the “two-dimensional warranty policy” regulation. The quality of key vehicle parts provided by its outsourcing suppliers plays a significant role not only on the quality and reliability of the vehicle products, but also on the reputation, customer loyalty, warranty cost and other quality loss of the assemblers. To improve the economics of the vehicle quality, preventive quality improvement strategy through the strategic procurement studies on the key parts is addressed. This study develops a two-stage decision making framework to deal with the supplier portfolio of the key outsourcing parts (SPKOP) selection by highlighting the economics of quality within the stipulated warranty period. In stage I, on the basis of customer feedback, the extended VIKOR-based multi-criteria decision making (MCDM) technique is developed to select the key outsourcing parts, which influence the economics of the vehicles with higher quality improvement priorities. According to the sequences of the key outsourcing parts, the supplier candidates of the key parts are investigated and determined. In stage II, a nonlinear mixed 0–1 integer programming (NLMIP) model is formulated to select the optimal supplier portfolio based on the total quality-related cost, system reliability, delivery time and customer complaints. The RPN-based method considering the nonlinear characteristics of failure severity is modified to determine the relative importance of each key part subject to its reliability. To resolve the SPKOP problem, the utility function and combined weighting technique are employed to cope with multiple conflicting objectives by the weighting technique. Next, a hybrid genetic algorithm (HGA) with a local search strategy is designed to improve the search efficiency. A case study is conducted to verify the effectiveness of the proposed model and the decision-making framework.