G-renewal process as a model for statistical warranty claimprediction

摘要

A brief overview of the statistical aspects of warranty predictionis given as an introduction. The main discussion then focuses onwarranty claim prediction for repairable products. Introduced by Kijimaand Sumita (1986), a g-renewal process (GRP) can be considered as amodel for major repair assumptions encountered in repairable productreliability analysis. These assumptions include“good-as-new”, “same-as-old”, the intermediate“better-than-old-but-worse-than-new”, and“worse-than-old”. A statistical procedure is developed forestimation of the GRP parameter, which is suggested to have engineeringmeaning of the effectiveness of the repair actions. A practical exampleof the GRP application in statistical warranty prediction is given as anillustration of the proposed estimation method. The paper arrives to thefollowing conclusions: The GRP provides high flexibility in modelingreal life failure occurrence processes by covering major repairassumptions encountered in practice. A Monte Carlo simulation can beconsidered as a method for statistical estimation of the GRP. Warrantyclaim prediction based on GRP provides a higher accuracy compared to theORP or the NHPP