Characterization of wheat-based biscuit cooking process by statistical process control techniques

Abstract

A commercial wheat-based biscuit cooking process was characterized using statistical process control (SPC) techniques. These techniques included histograms, analysis of means (ANOM), control charts, process capability studies and Gauge Repeatability and Reproducibility (Gauge R&R). The production variables studied included raw material variables (wheat protein and moisture content) and product variables (biscuit breaking strength, moisture content and thickness of biscuit). The plant studied has three independent production lines, so product variables were analysed separately for each line. ANOM indicated that several process means were significantly different (α=0.05) from the grand mean (breaking strength line 3, product moisture content line 3 and thickness of biscuit in all lines). Control charts indicated that some observed variables were not in statistical control (wheat protein and moisture content, breaking strength line 3 and thickness of biscuit for lines 1–2). Results from capability studies showed that all observed variables except wheat moisture content were not capable. Notably, there was a large variation in the thickness of biscuit. Gauge R&R suggested that the measurement system used was inadequate, contributing to the variation found in the process characterization. Initial investigation of the relationship between raw material variables and product variables in each line found no correlation.