Warranty data analysis for a fast reliability prediction

In recent years, the product renewal rate has quickly increased and, consequently, the product life cycle has decreased. For this reason, it is necessary to check, as soon as possible, the product quality, in terms of its reliability performance. The use of warranty field data may be useful, in such a context. This paper presents an application of the warranty call rate estimation, by means of the field data of a very common product: a washing machine. Using a Weibull distribution to fit the data, it has been possible to achieve a field-reliability prediction, basing only on a few months of field data. The following step of the work, was a comparison of the above mentioned prediction with the results of an alternative approach. The latter methodology derives from censored data analysis and is based on the filed data of the entire product warranty period. A final comparison of the two approaches, allowed to validate the fast prediction method, since the differences of findings of the two approaches are negligible.