Applying the minimum relative entropy method for bimodal distribution in a remanufacturing system

This paper provides a further investigation of a remanufacturing system studied by Tang et al. [2007. Planned lead time determination in a make-to-order remanufacturing system. International Journal of Production Economics 108, 426-435]. Due to the uncertainty of the disassembly lead time, purchasing lead time and yield rate, the probability density function (PDF) of the lead time for obtaining a component takes a bimodality form. In this paper, the minimum relative entropy (MRE) method, in which high-order moments are utilized, has been applied to approximate the lead time distribution. Numerical result shows that the MRE approach captures well the bimodality property. After reinvestigating the remanufacturing system, we verify the statement in Tang et al. that improving the recovery yield in disassembly does not necessarily improve the system performance. However, more accurate optimal decisions and system performance can only be evaluated by the advanced density function approximation method, such as MRE. Using data in an engine remanufacturing case, we further obtain managerial insights for supporting decision making in labor allocation, supplier selection and process investment.