ABSTRACT
The quality control technique of monitoring a process in real time to characterize its status has been discussed in previous chapters. The objective is to observe those process variables which have been proven to be the best indicators of the quality of the manufacturing operation. Then, undesirable shifts in the status of these key operating parameters can be sensed before the production quality suffers. This early warning of increased parameter variation or other excursions provides information that permits the appropriate corrective actions to be initiated in a timely manner. However, this is often a task that is easier to discuss in general terms than it is to implement in a production environment. One of the more difficult objectives to accomplish can be the~isolationof the most appropriate process variables for use in the control strategy. To obtain a process model that is not prohibitively complex, it is necessary that only a few critical process parameters be selected for monitoring and/or control activities. Models that utilize a redundant multitude of process parameters that are relatively insensitive to the status of the operation, are neither more accurate nor cost effective. In addition, another consideration that should be evaluated in the design phase of the control/monitoring system is the use of in-process versus post/process measurements. Also, the tradeoff involved in the use of direct versus indirect measurements of the process
variables for estimation/control of the quality of the process must be fully understood and accepted.
