A Comparative Overview of Electronic Devices Reliability Prediction Methods-Applications and Trends

Reliability prediction is vital in the conception, definition, design, development, operation and maintenance phase of electronic devices. It is needed at various system levels and degrees of detail, in order to evaluate, determine and improve the dependability measures of an item when designing electronic devices in view of the high level competition among device manufacturers. Different reliability prediction methods or models are available for electronic devices. This paper comparatively examined the commonly used methods such as empirically based failure rate modeling methodologies used in reliability prediction handbooks, and physics of failure (PoF) based models. Three empirical approaches such as MIL-HDBK-217F – a conservative standard applicable principally to military equipment, and Bellcore TR-332/Telcordia SR-332, which are applicable to commercial devices are reviewed in closer details. Also reviewed is Recueil de Donnes de Fiabilite (RDF) 2000, used in Telecom industry. Some PoF based methods such as Arrhenius law; Eyring model, Black Model for Electromigration, and Coffin Manson Model for fatigue are also examined. Additionally, the respective merits and demerits of the prediction methods which provide the basis for use are noted. The paper also attempts to highlight future trends and challenges in RP of electronic devices.