Cumulative degradation methodology to predict reliability of electronic systems

Abstract

Purpose The purpose of this paper is to describe a novel methodology for predicting reliability for consumer electronics or any other hardware systems that experience a complex lifecycle environmental profile. Design/methodology/approach This Physics-of-Failure–based three-step methodology can be used to predict the degradation rate of a population using a Monte Carlo approach. The three steps include: using an empirical equation describing the degradation of a performance metric, a degradation consistency condition and a technique to account for cumulative degradation across multiple life-cycle stress conditions (e.g. temperature, voltage, mechanical load, etc.). Findings Two case studies are provided to illustrate the methodology including one related to repeated touch-load induced artifacts for displays. Originality/value This novel methodology can be applied to a wide range of applications from mechanical systems to electrical circuits. The results can be fed into the several stages of engineering validation to speed up product qualification.

Investigation on reliability of classical industrial motor drives using GoldSim Monte Carlo reliability workbench

“…On the other hand, reliability plays an important role from product improvement to cost-saving during the manufacturing process (Rao and Chitra, 2023). The term “reliability” refers to designers’ efforts to decrease the likelihood of failure as much as feasible (Thiel et al , 2022). The establishment of the reliability indices is not a trial problem; many factors like technology, quality and complexity of the components get involved with the reliability of the system.…”

The Digital Twin Landscape at the Crossroads of Predictive Maintenance, Machine Learning and Physics Based Modeling

“…A class of methods that reflect this concept is known as degradation models. They can be implemented as a linear stochastic process or an exponential stochastic process if the equipment experiences cumulative degradation effects [69]. Degradation models typically work with a single condition indicator, although data fusion techniques can be used to combine multiple indicators of degradation.…”