Bayesian Analysis Enhances Sales and Warranty Strategies for Repairable Industrial Products by Considering Hybrid Deterioration Modes

Abstract

An effective warranty policy not only fulfills a manufacturer’s or vendor’s obligations but also plays a crucial role in enhancing customer confidence and encouraging future purchases. To attract more customers and drive sales, companies may extend the service life of their products. However, they cannot offer unlimited warranties to dominate the market, as the associated warranty costs will ultimately outweigh the profits. Therefore, manufacturers must strike a balance between the advantages of providing longer warranties to foster customer trust and the potential financial implications involved. Despite the significance of this issue, there has been limited research on hybrid deteriorating systems that encompass both maintainable and non-maintainable failure modes. Furthermore, conducting preventive maintenance analyses is challenging when historical failure data is insufficient. To address these gaps, this study introduces a Bayesian statistical approach to handle preventive maintenance challenges. The system’s deterioration is modeled using Non-Homogeneous Poisson processes (NHPP) with power-law failure intensity functions. A mathematical model, along with a solution algorithm, has been developed to assist manufacturers in making informed decisions regarding pricing, production, and warranty strategies. Furthermore, to facilitate the practical application of these models, the study offers solution algorithms and a computerized framework that enables decision-makers to implement automated decision-making processes.