Bayesian zero-failure reliability demonstration tests: Sample size estimation and its limitation.

Saved in:

Bibliographic Details
Title:	Bayesian zero-failure reliability demonstration tests: Sample size estimation and its limitation.
Authors:	Yoon, Yonghan^1,2 (AUTHOR), Yang, Min-Sik³ (AUTHOR), Bai, Cheol-Ho¹ (AUTHOR), Shim, Jaesool¹ (AUTHOR) jshim@ynu.ac.kr
Source:	Journal of Mechanical Science & Technology. Jul2025, Vol. 39 Issue 7, p3651-3666. 16p.
Subjects:	Mathematical statistics, Frequentist statistics, Sample size (Statistics), Random variables, Weibull distribution
Abstract:	A zero-failure test minimizes both the number of required samples and the test duration needed to ensure component reliability. Currently, most companies use the classical zero-failure test method, based on frequentist statistics, to determine the necessary sample size during reliability demonstration test planning. This method assumes that the shape and scale parameters are unknown and treated as constants—a strict assumption that can be difficult to meet when life testing is limited. In contrast, the Bayesian approach treats these parameters as unknown random variables, assigning probability distributions to reflect uncertainty of parameter estimation, offering a more realistic assumption. In this study, we numerically estimated the required sample size for zero-failure tests using the Bayesian method across various shape and scale parameters and compared the results with the classical method. While the Bayesian method generally requires a larger sample size, under specific conditions, we observed an unexpected phenomenon where it required fewer samples than the classical method. We identified the limitation of the Bayesian method for zero-failure tests and proposed criteria for its application. This study also provides guidelines for selecting between the classical method and the Bayesian method when planning zero-failure tests. [ABSTRACT FROM AUTHOR]
	Copyright of Journal of Mechanical Science & Technology is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	A zero-failure test minimizes both the number of required samples and the test duration needed to ensure component reliability. Currently, most companies use the classical zero-failure test method, based on frequentist statistics, to determine the necessary sample size during reliability demonstration test planning. This method assumes that the shape and scale parameters are unknown and treated as constants—a strict assumption that can be difficult to meet when life testing is limited. In contrast, the Bayesian approach treats these parameters as unknown random variables, assigning probability distributions to reflect uncertainty of parameter estimation, offering a more realistic assumption. In this study, we numerically estimated the required sample size for zero-failure tests using the Bayesian method across various shape and scale parameters and compared the results with the classical method. While the Bayesian method generally requires a larger sample size, under specific conditions, we observed an unexpected phenomenon where it required fewer samples than the classical method. We identified the limitation of the Bayesian method for zero-failure tests and proposed criteria for its application. This study also provides guidelines for selecting between the classical method and the Bayesian method when planning zero-failure tests. [ABSTRACT FROM AUTHOR]
ISSN:	1738494X
DOI:	10.1007/s12206-025-2409-1