Reliability prediction for electronic components: Difference between revisions

Revision as of 03:20, 17 March 2025

Latest revision as of 07:46, 27 August 2025

(2 intermediate revisions by 2 users not shown)

Line 6:

A [[prediction]] of reliability is an important element in the process of selecting equipment for use by [[telecommunications]] [[service providers]] and other buyers of [[electronic equipment]], and it is essential during the design stage of engineering systems life cycle.<ref>EPSMA, “Guidelines to Understanding Reliability Predictions”, EPSMA, 2005</ref> Reliability is a measure of the [[frequency]] of equipment failures as a function of time. [[wikt:reliability|Reliability]] has a major impact on maintenance and repair costs and on the continuity of service.<ref>Terry Donovan, Senior Systems Engineer Telcordia Technologies. Member of Optical Society of America, IEEE, "Automated Reliability Prediction, SR-332, Issue 3", January 2011; "Automated Reliability Prediction (ARPP), FD-ARPP-01, Issue 11", January 2011</ref>

Every product has a [[failure rate]], λ which is the number of units failing per unit time. This failure rate changes throughout the life of the product. It is the [[manufacturer]]’s's aim to ensure that product in the “infant mortality period” does not get to the [[customer]]. This leaves a product with a useful life period during which failures occur randomly i.e., λ is constant, and finally a wear-out period, usually beyond the productsproduct's useful life, where λ is increasing.

== Definition of reliability ==

Line 34:

== Data-driven reliability predictions ==

Data-driven models for reliability prediction utilise data aquiredacquired from tests to failure on electronic components by establishing relationships between the different variables presented in the data. As such relationships can be complex, data-driven models often require computations in high dimensions, which means that a large dataset is needed to optimize the output of the model.<ref>{{cite conference |last1=Ghrabli |first1=Mehdi|last2=Bouarroudj |first2=Mounira | author3=Chamoin, Ludovic|author4=Aldea, Emanuel |date=2024 |title=Hybrid modeling for remaining useful life prediction in power module prognosis |url=https://ieeexplore.ieee.org/document/10491493 |conference=2024 25th International Conference on Thermal Mechanical and Multi-Physics Simulation and Experiments in Microelectronics and Microsystems (EuroSimE)|___location=Catania, Italy |publisher=IEEE|doi=10.1109/EuroSimE60745.2024.10491493 }}</ref>

== Physics-based reliability predictions ==

Physics based reliability predictions use physical equations and formulae to determine failure. This approach requires precise knowledge of the degradation process and the physical properties to ensure accuracy. These models often utilise numerical simulations to infer the quantities needed by the model.<ref>{{ cite journal | title=Physics-informed Markov chains for remaining useful life prediction of wire bonds in power electronic modules | journal=Microelectronics Reliability | year=2025 | last1=Ghrabli | author2=Bouarroudj, Mounira | author3=Chamoin, Ludovic|author4=Aldea, Emanuel | volume=167 | pages=1–12 | first1=Mehdi | article-number=115644 | doi=10.1016/j.microrel.2025.115644| urlbibcode=https://www2025MiRe.sciencedirect.com/science/article/pii/S002627142500057516715644G | doi-access=free }}</ref>