On assumptions in optimisation of warranty policies

Optimisation of warranty policy has been a hot research topic in both operations research and statistics communities since warranty providers hope to balance cost-benefit analysis in the nowadays competitive market. Some assumptions are inevitably needed for such research. Most of the existing publications, however, make assumptions that may not be true in practice, based on which biased decision may be made. This paper discusses pitfalls in the assumptions, which include causes of warranty claims, pattern of warranty claims, warranty claim models, field reliability vs product reliability, the relationship between usage and age in 2-dimensional warranty. A real-world example is used to elaborate the arguments.