The bathtub model has often been cited as a hazard rate profile that predicts not only human mortality but also the hazard rate of manufactured products. This paper examines the inapplicability of the bathtub model to predict the hazard rate of electronic components, products, and systems. The paper reviews various literature and uses numerous experiments and field failure data to support its argument. The factors affecting the hazard rate of electronics are then discussed, common hazard rate trends for electronics are identified, and recommendations for the assessment of hazard rates are given. This paper recommends that a preconceived model of hazard rate over time should not be used.

No abstract