Time series techniques for dynamic, real-time control of wood-drying processes

Continuous improvement in any process necessitates reduction in the variability across the target value of the process. Traditionally, two different technical schemes have focused on realizing this goal: Statistical Process Control (SPC) used by statisticians and quality engineers and Engineering Process Control (EPC) used by control engineers. In the wood products industry, EPC in the form of Distributed Control Systems (DCS), Programmable Logic Controllers (PLCs), and single-loop controllers have prevailed for physical control of processes (Young and Winistorfer 1999). In the last decade, many wood products companies have used SPC techniques to monitor and analyze the resulting performance of these EPC processes. However, the full dynamic value of these techniques, that is, to combine the reactive control properties of EPC with the predictive control capability of SPC, has not been commonly adapted. This paper discusses the concept of combined EPC/SPC and demonstrates the use of this technique in improving a common wood-processing operation: hardwood lumber drying. In this application, the adoption of techniques similar to those described in this work will automate the kiln control problem through dynamic, real-time kiln schedule adjustments.