Abstract Reducing variation increases both quality and productivity. To exploit a system to its maximum capacity we need to control variation. To identify the sources of variation, the information transmitted from an operating system (Box, G. E. P. (1990–1991). Good quality costs less? How come? Quality Engineering 3(1): 85–90.) is used. With suitable receivers (quality control techniques), the manufacturing process behavior of a polymer plant is studied in order to exert better control over it. A separate study of this plant's historical data showed that long-term process improvement is feasible if the focus of quality control system changes is upstream after the polymerization reaction process. In order to identify the real process variation and/or its sources, an observational study was planned to collect data containing (quantitative/qualitative) quality characteristics of the reactor's input and output material and to plot the statistical control chart of the polymer quality characteristic measured right after polymerization. The article shows that the observational study and the use of different statistical tools (such as correlation analysis presented as quality function deployment [QFD] diagrams, covariation chart, and sliced inverse box [SIB] plots) are powerful enough for the identification of a list of variables correlated to the product quality characteristics in such process industry and can lead to the identification of the reactor's input/output relationships.
[1]
David Banks,et al.
Pre-analysis of superlarge industrial data sets
,
1992
.
[2]
Louis Cohen,et al.
Quality Function Deployment: How to Make QFD Work for You
,
1995
.
[3]
W. T. Tucker,et al.
Algorithmic statistical process control: concepts and an application
,
1992
.
[4]
Jan Degrève,et al.
Quality Control in a Semi-continuous Polymer Production Process
,
2004
.
[5]
J. Macgregor,et al.
Monitoring batch processes using multiway principal component analysis
,
1994
.
[6]
Yoji Akao,et al.
Quality Function Deployment : Integrating Customer Requirements into Product Design
,
1990
.
[7]
William Y. Fowlkes,et al.
Engineering Methods for Robust Product Design: Using Taguchi Methods in Technology and Product Development
,
1995
.