An empirical evaluation of attribute control charts for monitoring defects

The traditional C-chart by Shewhart has been widely applied for monitoring count data in industrial and nonindustrial processes. However, using C-chart always experiences an excessive amount of false alarms, since control limits of traditional C-chart are defined by impractical normal assumption. Thus, monitoring a process by traditional C-chart leads to the increase of unnecessary costs of inspection. There are many works that have attempted to improve C-charts. In this paper, 11 selected improved versions of C-charts are presented. The performances of improved C-charts are evaluated in term of numerical results to demonstrate the sensitivity of the charts and costs of inspection.