Detecting special-cause variation ‘events’ from process data signatures

Abstract

The ability to detect the special-cause variation of incoming feedstocks from advanced sensor technology is invaluable to manufacturers. Many on-line sensors produce data signatures that require further off-line statistical processing for interpretation by operational personnel. However, early detection of changes in variation in incoming feedstocks may be imperative to promote early-stage preventive measures. A method is proposed in this applied study for developing control bands to quantify the variation of data signatures in the context of statistical process control (SPC). Control bands based on pointwise prediction intervals constructed from the Bonferroni Inequality and Bayesian smoothing splines are developed. Applications using the control band method for data signatures from near-infrared (NIR) spectroscopy scans of industrial fibers of Switchgrass (Panicum virgatum) used for biofuels production, Loblolly Pine (Pinus taeda) fibers for medium density fiberboard production, and formaldehyde (HCHO) emissions from particleboard were used. Simulations curves (k) of k = 100, k = 1000, and k = 10,000 indicate that the Bonferroni method for detecting special-cause variation is closely aligned with the Shewhart definition of control limits when the pdfs are Gaussian or lognormal.