Chapter 11: Control Chart Signals


In 1924, Dr. Walter Shewhart developed the first control chart. He designed control charts as graphical tools to identify and separate common and special variation. It is important, however, to recognize that to identify the type of variation, one must be cognizant of the intent of the particular chart and the information that it provides to the individual who is doing the analysis. In this chapter, we will attempt to give an explanation of the charts and the meaning of out-of-control conditions.

OVERVIEW

First, in the variable charts, one encounters two charts: Xbar, Median, and X, in which we measure variation between subgroups over time; and R and s, in which we measure variation within subgroups over time.

Second, in the attribute chart, the analysis is around one chart that measures a fraction or number defective, number of defects, or number of defects per unit. This means that in essence, the measurement of variation is between samples.

Regardless of which chart we evaluate, the idea is to be able to separate the common and special variation in relationship to the control limits. We want to be able to determine that the process under study is in statistical control. A process is in statistical control if it is not affected by special causes. That means that all points on a given chart are randomly spread about the mean.

Another point of understanding is that of the issue of control . Control , as used in SPC, does not mean that the product or service will meet the needs of the customer. Rather, it means that the process is consistent and predictable. This is very important, and one must keep the control limits and specifications separate because they communicate different things. The intent of control is to progressively proactively react to the specific signal that the process is generating. Specifically, there are three general options that one must take:

  1. React to out-of-control points:

    Verify readings .

    Record the time, if required; the reason for the out-of-control points; and the corrective action on the chart itself.

    Stop the process if necessary.

    Notify supervisor.

    If needed, contact quality professional.

  2. React to out-of-control signals:

    Notify supervisor and customer if necessary.

    Identify and record the reason for the out-of-control condition.

    Correct process if necessary.

    If needed, contact quality professional.

  3. React to out-of-specification limits:

    Stop process.

    If necessary, shut down particular machine.

    Notify supervisor or process office.

    Start containment action and notify customer.

    Hold parts ” an approved alert must be issued.

Finally when we say as a general comment that a process is out of control, we have to be very careful. In reality, at that moment, we are not sure that the process is in fact out of control. We suspect that the process is behaving very unusually, and we need further investigation to make the final disposition. In essence then, when we talk about the process being out of control, we mean that the process is giving a signal that needs to be investigated.

To verify the special cause, we must do some additional probability calculations on the process that are not very pleasant for the average person to do. That is why we recommend that the process be stopped for further investigation. It is possible to have the traditional out-of-control conditions in the process and not do anything.




Six Sigma and Beyond. Statistical Process Control (Vol. 4)
Six Sigma and Beyond: Statistical Process Control, Volume IV
ISBN: 1574443135
EAN: 2147483647
Year: 2003
Pages: 181
Authors: D.H. Stamatis

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net