An adaptive maintenance policy for degrading units with an imperfect and a perfect inspection

This paper presents an adaptive condition-based maintenance policy for degrading units where at the inspection time an imperfect inspection is firstly carried out to decide whether to immediately replace the unit, to define its future replacement time, or to immediately carry out an additional, more costly, perfect inspection. When it is performed, the perfect inspection is used again to decide whether to immediately replace the unit or to define its future replacement time. The imperfect inspection is assumed to provide a degradation measurement affected by error. The perfect inspection is supposed to give an exact degradation measurement. The distinguishing feature of the proposed policy is that the future replacement time is defined adaptively, unit-by-unit, based on the outcome of the inspections. It is assumed that the inspections can be performed only once, at most, over the lifespan of the unit. The policy accounts for the joint presence of temporal variability, measurement error, and random effect. The temporal variability is modeled by using a gamma process. Two comparative studies demonstrate the superiority of the proposed adaptive decision rule with respect to the classical control limit rule and show how its use allows to decide whether and which inspection is convenient to perform, depending on inspection costs and features of the degradation process.