Data-Driven Decision-Making method for Functional Upgrade Remanufacturing of used products based on Multi-Life Customization Scenarios.

With rapid changes in technology and customer preferences, functional obsolescence of used products poses serious challenges to resumed remanufacturing. Upgrade remanufacturing is a potential solution for dealing with the problems of functional obsolescence. The multi-attribute remaining life and customized life of the functional unit are critical elements of decision-making for upgrade remanufacturing solution, yet there are many possible scenarios for the multi-attribute remaining life and the customized life. The optimal solution varies with the various scenarios, which makes the decision-making for choosing the optimal solution of Functional Upgrade Remanufacturing (FUR) very individual and complicated. To this end, the paper proposes a Data-Driven Decision-Making (DDDM) method for FUR of used products based on Multi-Life Customization Scenarios (MLCS). MLCS describes the relationship between the remaining physical, technical, economic life and customized life. Firstly, the used product is decomposed into several functional units that are taken as the objects for upgrade remanufacturing, and the mapping between MLCS and decision-making for FUR is established through data mining. Then the DDDM method of Bayesian network is employed to inference, which is constructed based on historical data, and the solution with the largest posteriori probability is taken as the optimal solution. Finally, a case study on decision-making for FUR of a used mechanical hydraulic power steering system is demonstrated to validate the proposed method. • Describe the value of Functional Upgrade Remanufacturing (FUR) to overcoming obsolescence. • Take the remaining life and customized life as elements of decision-making for FUR. • The concept of Multi-Life Customization Scenarios (MLCS) is proposed for FUR. • The mapping between MLCS and decision-making for FUR is established by data-mining. • The DDDM method of Bayesian network is employed to choose the optimal solution. [ABSTRACT FROM AUTHOR]