A new single‐sample failure model and its application to a special CNC system

Purpose – This paper is to present a new failure model that can be applied to single‐sample failure data of a single system under testing. Design/methodology/approach – The Bayesian method is used for the reliability evaluation. The weighted least squares method is used for determining the parameters of the reliability function. Findings – The authors have observed the operation of a special computer numerical control (CNC) system for a period of over two years, and maintained a reliability database will all the collected failure data, from which the main source of failures can be identified. Research limitations/implications – Preliminary research results are very encouraging. However, more work will be necessary to validate the new failure model. Practical implications – The determination of the parameters of the reliability function of a system under testing helps to identify its failure characteristics and potential quality problems. Originality/value – It is hoped that the paper can help understand some of the challenges in modeling the failure behavior of special CNC systems.