Chapter 9 Inference for one-shot devices with multiple failure modes ⍟ ⍟ This book has a companion website hosting complementary materials. Visit this URL to access it: https://data.mendeley.com/datasets/879xmdz3d8/1.

Many practical devices are usually composed of multiple components, which may lead to different failure modes. In this chapter, we present a comprehensive overview of the different inferential methods developed for one-shot devices with multiple failure modes under constant-stress accelerated life tests (CSALTs). These include the competing risks model, assuming independence between the different failure modes, and some copula models that relax this assumption by allowing the presence of correlation between times to failure modes. Finally, we define the family of weighted minimum density power divergence (DPD) estimators and Wald-type tests for this competing risks scenario, as a potential robust alternative to the classical maximum likelihood estimator (MLE).