Optimal degradation-based hybrid double-stage acceptance sampling plan for a heterogeneous product

Yao Cheng, Haitao Liao, Zhiyi Huang

Research output: Contribution to journalArticlepeer-review

Abstract

Units of a product, even though produced in the same batch, may perform differently due to a variety of randomness in raw materials, manufacturing process and repair actions (e.g., refurbishment). If these units are also stored for a long period of time, their key characteristics will degrade randomly during storage. As a result, the product will exhibit heterogeneous performance upon delivery. Such a heterogeneity significantly affects the product's acceptability and is one of the utmost concerns to both the producer and consumer. In this study, we propose a degradation-based hybrid double-stage acceptance sampling plan (DHDASP) to determine the acceptability for a product exhibiting heterogeneous performance. Compared to ordinary acceptance sampling plans, the new plan provides the producer with a higher likelihood of product acceptability and offers the consumer an opportunity to explore the product's operational reliability metrics without unnecessary discard of units. In practice, it will balance the benefits of the producer and consumer and meanwhile, accurately assess the product's reliability metrics under flexible testing scenarios. A numerical study verifies the advantages and higher efficiency of the DHDASP over commonly used alternatives in handling heterogeneous products.

Original languageEnglish (US)
Article number107544
JournalReliability Engineering and System Safety
Volume210
DOIs
StatePublished - Jun 2021
Externally publishedYes

Keywords

  • Degradation, double-stage acceptance sampling plan
  • Heterogeneous product

ASJC Scopus subject areas

  • Safety, Risk, Reliability and Quality
  • Industrial and Manufacturing Engineering

Fingerprint

Dive into the research topics of 'Optimal degradation-based hybrid double-stage acceptance sampling plan for a heterogeneous product'. Together they form a unique fingerprint.

Cite this