The effect of non-constant process variance in determining the optimal process means and production run of a deteriorating process

Abstract

Many models reported in the current process target literature assume that the process variance is constant throughout the process. In the real-world industrial settings, however, the process variance changes as a result of the variations in materials, labour and manufacturing conditions. A better understanding of how variation propagates through a process will provide insight into optimising process operating conditions, reduce out of specification production and improve design and economic analysis. This article considers the process variance to be time dependent. This article provides a method of estimating time period at which a process may be stopped and reset periodically. A model is developed to determine the optimum process means and production run for such a process with multiple quality characteristics. To illustrate the application of the proposed model, a case study from the automotive industry is provided.

Keywords:

• deteriorating process
• drift
• loss function
• multiple quality characteristics
• process mean
• production run
• shift
• variance