Process monitoring and deviation analysis – the role of the Shewhart, CuSum and EWMA control charts

Tuesday, April 10, 2018

Time: 15.00-16.00 (CEST)

Sartorius Stedim Data Analytics (SSDA) invites to a webinar on the Umetrics® Suite and using SIMCA® 15 and SIMCA®-online 15 for multivariate statistical process control (MSPC)

Statistical process control (SPC) is the classical approach for investigating stability and reliability of industrial processes over time. The Shewhart, CuSum and EWMA control charts are the basic tools used for process monitoring. The Shewhart chart is the simplest and most straightforward. This chart allows the mean level and the variability of the process to be charted over time. It can be used for individual measurements and also for data arising from subgrouping. Subgrouping is used to make the charted data approach normality, which has the advantage that the resulting control charts often become smoother.

The Shewhart chart is useful for detecting large and sudden shifts in a process. However, the main drawback of the Shewhart chart in relation to other control charts is its poor ability to detect subtle process fluctuations. The CuSum and EWMA charting approaches are more efficient for tracking small process changes. This feature arises because in a Shewhart chart, the last plotted point depends only on the last measurement, while in a CuSum chart it depends on all previous observations, and in an EWMA chart, the last plotted point depends on the value of the adjustable memory parameter lambda.

When applying EWMA, the memory parameter lambda must be specified. In the SIMCA software it is possible to estimate lambda from the data, but it may also be specified by the user. Usually, we find that lambda is set at 0.2 ± 0.1. Note that with a low value of lambda, say below 0.15, the CuSum charting approach becomes applicable.

Attend this webinar and get familiar with the basic control charting possibilities inside the SIMCA software family.

Click here to register