What Quality Engineers Need to Know about Degradation Models

TL;DR

The paper addresses reliability analysis when degradation data are available by introducing two core modeling paradigms—General Path Models (GPMs) and Stochastic Process (SP) models—and detailing inference methods (ML and Bayesian) to estimate degradation paths, failure-time distributions, and remaining useful life. It emphasizes RMDT and ADDT data structures, discusses dynamic covariates and accelerated testing (e.g., Arrhenius for temperature), and demonstrates applications across coatings, polymers, metal fatigue, civil infrastructure, solar, aerospace, and pharmaceuticals. Key contributions include guidance on model choice (favoring GPMs for practical, physics-informed degradation paths), Bayesian frameworks for uncertainty quantification, and case-study-driven illustrations of reliability and RUL predictions with real data. The work has practical impact by providing practitioners with a structured toolkit for designing experiments, selecting models, performing inference, and deploying degradation-based maintenance and shelf-life decisions in industrial settings.

Abstract

Degradation models play a critical role in quality engineering by enabling the assessment and prediction of system reliability based on data. The objective of this paper is to provide an accessible introduction to degradation models. We explore commonly used degradation data types, including repeated measures degradation data and accelerated destructive degradation test data, and review modeling approaches such as general path models and stochastic process models. Key inference problems, including reliability estimation and prediction, are addressed. Applications across diverse fields, including material science, renewable energy, civil engineering, aerospace, and pharmaceuticals, illustrate the broad impact of degradation models in industry. We also discuss best practices for quality engineers, software implementations, and challenges in applying these models. This paper aims to provide quality engineers with a foundational understanding of degradation models, equipping them with the knowledge necessary to apply these techniques effectively in real-world scenarios.

Figures (16)

• Figure 1: Illustration of the degradation paths and the failure threshold, which together induce a failure-time distribution.
• Figure 2: Illustration of degradation data: (a) Laser degradation data, an example of linear degradation; (b) Outdoor weathering data, an example of nonlinear degradation. Figure reproduced from Meeker et al. (2011) with permission from Wiley.
• Figure 3: Illustration of degradation data with an accelerating variable using the Device B data and the fitted degradation paths. Figure reproduced from Meeker et al. (2011) with permission from Wiley.
• Figure 4: Plot of nine representative degradation paths (a) and corresponding dynamic independent variable information for daily UV dosage (b). Daily temperature and relative humidity data are not shown here. Figure reproduced from Hong et al. (2015) with permission from Taylor and Francis.
• Figure 5: Visualization of Adhesive Bond B data with scatter plot and fitted parametric degradation paths. Figure reproduced from Xie et al. (2018) with permission from Taylor and Francis.