It is common that food, drugs, oil, chemicals, cosmetics, and many other products are manufactured by continuous production processes, that is, the production takes place in a continuous product flow. Continuous production plants are typically inflexible, expensive, and large. Almost all production personnel is involved in the experiments due to the around-the-clock production, making the information flow and the coordination essential. The complex production is highly automated to maintain product quality and process control. Despite extensive automation, these processes require improvements to remain competitive.
The research area of experimental design includes a set of powerful tools for process improvement and optimization. Automation, however, interferes with the possibility of conducting experiments. The traditional approaches of experimental design and analysis implicitly assume that the studied process is operating in open loop, i.e., not under automatic control.
In the study, the authors have studied closed-loop control experimentation, not for optimizing the controllers, which is often occurring, but for gaining knowledge on how to improve the processes and the products. The study explores important issues in designing and analyzing experiments in such environments and illustrates how to optimize continuous processes using realistic simulations. Results show how designed experiments are useful for studying the process phenomena but also for studying the effectiveness and efficiency of the automatic control.