Remaining Useful Lifetime (RUL): Probabilistic Predictive Model

Reliability evaluations and assurances cannot be delayed until the device (system) is fabricated and put into operation. Reliability of an electronic product should be conceived at the early stages of its design; implemented during manufacturing; evaluated (considering customer requirements and the existing specifications), by electrical, optical and mechanical measurements and testing; checked (screened) during fabrication; and, if necessary and appropriate, maintained in the field during the product’s operation. Prognostics and health monitoring (PHM) effort can be of significant help, especially at the last, operational stage, of the product use. Accordingly, a simple and physically meaningful probabilistic predictive model is suggested for the evaluation of the remaining useful lifetime (RUL) of an electronic device (system) after an appreciable deviation from its normal operation conditions has been detected, and the corresponding increase in the failure rate and the change in the configuration of the wear-out portion of the bathtub curve has been assessed. The general concepts are illustrated by a numerical example. The model can be employed, along with other PHM forecasting and interfering tools and means, to evaluate and to maintain the high level of the reliability (probability of non-failure) of a device (system) at the operation stage of its lifetime.