Some Suggestions for Successful Statistical Process Control Mini Paper Taken from the Spring 2001 Newsletter
Abstract: Three causes that result in the failure of statistical process control (SPC) systems are discussed. These are personnel issues, methodological errors and inadequate maintenance. Fortunately, there are preventive measures that can be taken in any of these situations. Personnel issues can be addressed by training operators to understand the SPC tools and providing an out of control action plan that the operators can execute. Further, incentives to line workers and managers should be designed to encourage a proper implementation of SPC. Since methodological errors tend to result from measuring the wrong parameters, designed experiments should be used to identify measurements that are proven influences on critical process outcomes. Furthermore, it is critical to achieve a balance between false alarm rates and successful detection of out of control conditions. Finally, since SPC is in itself a process, the SPC implementation should also be monitored and maintained. In this instance it is useful to have a written monitor and action plan with stated goals and responsibilities.
Keywords: Statistical process control - Out of control action plan - False alarms - Control charts - Measurement systems validation