Título da Apresentação:

Making Combining Life Data Analysis, Warranty Data and Quality Methods

Descrição:

One can prepare forecasts of expected warranty returns for a future period based on a life data analysis model. These forecasts can then, in turn, be compared with actual returns observed in that period. A set of tools from statistical process control can be used to determine if a period and lot combination indicates an isolated departure from the life data model or, more importantly, if there is evidence of a persistent model departure. A persistent model departure can indicate a bad lot or an incorrect life data model, usually due to emerging failure modes previously masked in the censored observations.

This tutorial will illustrate the concepts with an example, using the Weibull++ life data analysis software and some Visual Basic macros in Excel.

Autor(es)/ Empresa: