Order statistics based sampling design for reliability sampling

• Main Authors: Low, Chan Kee, Balasooriya, Uditha
• Other Authors: Nanyang Business School
• Format: Research Report
• Published: 2008
• Subjects: DRNTU::Business
• Online Access: http://hdl.handle.net/10356/7041
• Institution: Nanyang Technological University

We consider the problem of designing variables-sampling plans for life test-ing. Leiberman and Resnikoff, Schneider and many others have considered this problem for location-scale type lifetime distributions. However, in prac-tice often the actual lifetime distribution is unknown. In this article, we pro-pose a procedure based on order statistics for designing a reliability sampling plan which is applicable to any lifetime distribution. It is shown that, under the proposed procedure, subject to MIL-STD-105D specifications, only a few specimens need to be tested until failure, thereby reducing total test time con-siderably. A large-scale simulation study shows that the proposed procedure performs well. Two examples are given to illustrate the procedure. Keywords: Acceptance sampling; Order statistics; Life testing; Censoring; OC Curve