Effects of Parameter Estimation on the Modified and Acceptance Control Charts

• Activity Number: 615 - Statistical Process Control
• Type: Contributed
• Date/Time: Thursday, August 1, 2019 : 8:30 AM to 10:20 AM
• Sponsor: Quality and Productivity Section
• Author(s): Felipe Jardim* and Bruna Ker and Subhabrata Chakraborti and Pedro Oprime
• Companies: Pontifical Catholic University of Rio De Janeiro and Federal University of São Carlos (UFSCar) and University of Alabama and Federal University of São Carlos (UFSCar)
• Keywords: Type I and Type II Errors; False Alarm Rate; Control Limit Adjustments; Guaranteed In-Control Performance; Average Run Length

Abstract

Control charts are powerful tools used by many industries to monitor the quality of processes and detect special cause of variations. The Shewhart Xbar chart is one of the most popular control charts to monitor the process mean, i.e., to detect a change from an in-control (IC) to an out-of-control (OOC) situation. In its original formulation, if the actual process mean is different from the specified IC level, the process is considered to be in an OOC state. However, in many practical situations, the process might still be capable from a practical point of view even if the process mean may slightly differ from the IC level. In such situation, the Modified and the Acceptance Control Charts are the appropriate tools to monitor the process mean. To design these charts, usually, the IC process standard deviation must be estimated from reference samples. In this paper, we show that this estimation has a negative impact on the performance of the Modified and the Acceptance charts. Solutions to this impact are investigated.