Statistical process control of an industrial process in real time

Abstract

Most of the real-time industrial control methods suffer from the fact that the corrective action is often taken after the instability has been detected, In other words the control strategy has always to catch-up with the problems by bringing the process back to optimum condition. Most of the time correct assignable cause of variation is not known. The conventional methods of quality control are always to catch-up with the product quality. In addition to this, conventional methods may cause financial losses due to machinery downtime and in many cases may not be able to predict the problem to avoid machinery downtime. The new computer-aided statistical process control (SPC) provides several windows in to the industrial process. These windows help collect statistical information which in turn can be used to find assignable causes of variation in product quality, quantitative measure of process capability, presence of extraneous influences and deterioration of certain equipment. We have introduced various types of problems into the industrial grade process control system and by using the SPC software tool we have collected data samples of the variables over time. These data sample were used to generate control charts of various process variables during runtime. Later, by analyzing the trend of the control charts we have been able to relate back to the problems which were originally introduced in to the process.