The value of quality grading in remanufacturing under quality level uncertainty

In remanufacturing, variability in quality levels of available cores (end-of-life products) has an impact on both the process cost and the process time. While previous research suggests that quality grading adds value, there are also concerns raised regarding how reliably the grades can be identified. We argue that uncertainty is inherent to the grading process and investigate the value of grading by taking into account the underlying uncertainty. We develop a robust optimisation model for remanufacturing planning, where both the per-unit cost and resource requirement to remanufacture a core are uncertain parameters that are assumed to reside in two different uncertainty sets; box and ellipsoidal. We analyse both uncapacitated and capacitated cases, and based on extensive numerical analysis, conclude that while on average, there is still value in grading, it becomes significantly smaller when the inherent uncertainty is accounted for. For the capacitated case, we also consider a cost for grading and find that it may cause a significant deterioration in the value of grading, if not rendering the grading totally useless. We show the validity of our approach through extensive numerical analyses.