A delayed product differentiation model for cloud manufacturing

Cloud manufacturing (CMfg) enables in-depth customization but raises demand uncertainties. Delayed product differentiation (DPD) is one attempt to solve such problem. However, current DPD approaches are mostly focused on the production of a dominant company, which tend to hold the supply chains with fixed network structures. Nevertheless, CMfg requires more flexible structures (agile supply chain) to facilitate multiple manufacturers to access external resources with various manners. Therefore, linking DPD to CMfg becomes an important research topic, and the paper proposes an optimization model, namely DPDCM, for such purpose. The model is established on the basis of integrating the order-release, generic inventory and sourcing decisions, and is formulated as an integer programming problem, to meet diverse requirements of companies for carrying out DPD in CMfg environment. Case studies on bicycle and industrial-transformer manufacturing have been applied and analyzed, in which genetic algorithm is adopted to obtain near-optimal solutions. It validates the effectiveness, flexibility and universality of DPDCM.