Integrated Degradation Models in R Using iDEMO

Abstract

Degradation models are widely used to assess the lifetime information for highly reliable products with quality characteristics whose degradation over time can be related to reliability. The performance of a degradation model largely depends on an appropriate model description of the product's degradation path. The cross-platform package iDEMO (integrated degradation models) is developed in R and the interface is built using the Tcl/Tk bindings provided by the tcltk and tcltk2 packages included with R. It is a tool to build a linear degradation model which can simultaneously consider the unit-to-unit variation, time-dependent structure and measurement error in the degradation paths. The package iDEMO provides the maximum likelihood estimates of the unknown parameters, mean-time-to-failure and q-th quantile, and their corresponding confidence intervals based on the different information matrices. In addition, degradation model selection and goodness-of-fit tests are provided to determine and diagnose the degradation model for the user's current data by the commonly used criteria. By only enabling user interface elements when necessary, input errors are minimized.