Robust Attribute Control Chart for Median‐Truncated Life Tests Under the Odd Exponential Half‐Logistic Distribution

Abstract

In this research, an attribute control chart is developed in order to monitor the product lifetimes having odd exponential half‐logistic distribution (OEHLD) with truncated life tests. The novelty of the suggested scheme is in truncation time that is assumed to be a multiple of the median of lifetime distribution instead of using any other quantile including mean. Such a feature eliminates the lack of a closed‐form expression of mean of OEHLD and makes the suggested approach more robust for skewed data. This control chart is constructed to detect changes in the scale parameter while considering the shape parameter as a known quantity. Based on the binomial approximation, the control limits are derived, and then its performances are evaluated by computing average run lengths in case of both in‐control and out‐of‐control scenarios. Simulation studies show the efficiency of the proposed chart in terms of rapid process identification. Moreover, the applicability of the suggested approach is verified by means of real‐life example of carbon fiber failures together with the generated artificial data. It can be stated that the use of median as a multiple for truncation time makes it possible to achieve more reliable process identification and control, which contributes to better lifetime control in industries. The proposed chart offers an easy‐to‐implement and theoretically sound solution for industries where lifetimes are positively skewed and non‐normal, providing a valuable tool for quality engineers and reliability analysts.