A Bayesian Approach to Determine Test Sample Size Requirements for Reliability Demonstration Retesting after Product Design Change

作者：Kleyner Andre^*; Elmore David; Boukai Benzion

来源：Quality Engineering, 2015, 27(3): 289-295.

DOI：10.1080/08982112.2014.990035

• sample size reduction
• mixture priors
• retesting
• attribute test
• reliability
• redesign
• beta priors
• reliability demonstration
• Bayesian analysis

Last minute design changes after already completed product validation testing is a common occurrence in the engineering world. It is also common for a redesigned product to be retested to demonstrate compliance with the original reliability requirements. Therefore, there is often a question of how to demonstrate the required reliability metrics without repeating already completed test flow on a redesigned product. This article discusses the application of Bayesian techniques to reduce the sample sizes required for retesting of engineering products/systems after design modification. This method utilizes attribute test data collected on previous version(s) of the product via success-based testing prior to the design change occurrence. The proposed method helps to reduce the test sample size for attribute testing while demonstrating the required reliability and helping to reduce the overall cost of a product development cycle.

• 出版日期2015-7-3