End-of-Use Core Triage in Extreme Scenarios Based on a Threshold Approach

Remanufacturing is a significant factor in securing sustainability through a circular economy. Sorting plays a significant role in remanufacturing pre-processing inspections. Its significance can increase when remanufacturing facilities encounter extreme situations, such as abnormally huge core arrivals. Our main objective in this work is switching from less efficient to a more efficient model and to characterize extreme behavior of core arrival in remanufacturing and applying the developed model to triage cores. Central tendency core flow models are not sufficient to handle extreme situations, however, complementary Extreme Value (EV) approaches have shown to improve model efficiency. Extreme core flows to remanufacturing facilities are rare but still likely and can adversely affect remanufacturing business operations. In this investigation, extreme end-of-use core flow is modelled by a threshold approach using the Generalized Pareto Distribution (GPD). It is shown that GPD has better performance than its maxima-block GEV counterpart from practical and data efficiency perspectives. The model is validated by a synthesized big dataset, tested by sophisticated statistical Anderson Darling (AD) test, and is applied to a case of extreme flow to a valve shop in order to predict probability of over-capacity arrivals that is critical in remanufacturing business management. Finally, the GPD model combined with triage strategies is used to initiate investigations into the efficacy of different triage methods in remanufacturing operations.