Control Chart

QW5 Help Files

The Control Chart screen is a very important investigative tool for determining what may be causing a process  upset or control rule violation.  Viewing this multi purpose screen can give you a clear picture about the current state of a particular variable.  In QW any numeric, calculated, average or range type variable, (field type N,C,A,R), can be charted on this screen.

The screen is divided into four sections.

Each on these four areas have a distinct function and used together provide a very powerful, statistically based, visual tool that will help you track down the cause of process variation.

The Control Chart

A Control Chart provides you with a graphic comparison of a measured variable against set control limits.  It plots variation over time and helps you identify special versus normal causes of variation.  

Using the Control Chart

The Control Chart is used to understand the nature of variation for  a variable.  Data points which deviate greatly from targets, or trends over a number of data points can be visually identified from the Control Chart. 

The following is an outline of the features of the Control Chart which will help in the task of tracking down the cause of variation in a variable.

Chart Features

Self Scaling - the Control Chart will display all historical points currently in memory for a variable.  The number of historical points is set in QWSETUP for each template or defined in a View.  The number of points currently displayed on the Control Chart is displayed in the Statistics Box - No. of Points.  The horizontal scale of the chart is proportional to the number of records displayed in the bottom right hand corner of the screen, Record 100 of 100.  The minimum scale size is 25 points and the scale increases by 5 until current number of records has been reached.

Color Zones - The Control Chart uses the same color interpretation of the limits set in QWSETUP for a variable  that is used consistently throughout QW.

on target and in control

in control but closer to a control limit than target

out of control but within specification limit

out of specification limit

Legend - at the bottom of the Control Chart a five character legend is displayed.  Which variable to be used as a legend is defined by you in QWSETUP.  It can be the Date or Time variable or a different variable such as Batch Number or Order Id.  Only the first five characters of the variable selected will be displayed on this area.  The scaling of the chart dictates the number of legend entries that are displayed.  For 100 points a legend value for every 5 points is displayed. 

Positional Cursor Bar - This is a very useful feature which allows you to scroll or position the cursor bar on individual data points on the Control Chart.  The cursor bar is always positioned on the current record number on the Log Sheet.  What is displayed in the two Annotated Fields and Data field is controlled by this feature.

Annotated Field 1 [Value: 180, Axis 20 B-TEAM 2005-06-19] is located in the yellow highlighted box below the Control Chart and is tied to the position of the Positional Cursor Bar

Annotated Field 2  [OVER ADJUSTED OVEN TEMP.DOWN] is used in conjunction with the Positional Cursor Bar feature. As the Cursor Bar is moved right and left, this window, at the top right corner of the Control Chart screen, displays the contents of the variable field defined in QWSETUP as Annotated Field 2 for the template.  The variable defined in QWSETUP  can be any variable within the template.  Normally a text field is used so a comment can be entered to indicate an action taken or the cause of a process upset.

You should set the two annotated fields to provide the best information for the user of the Control Chart screen with most useful information based on the application.  For instance, if you have defined the legend of the Control Chart to be the Time field then setting the Annotated Field 1 value to the Date variable would provide additional useful and related information.

Rule Field [Rule UCL: The last point is greater than the Upper Conreol Limit] to the right of the Data Field, displays the three character RuleID and rule description if a control rule, as defined and activated in QWSETUP, has been triggered.  The contents of this field changes as the Cursor Bar moves left and right across the Control Chart.

The Capability Bar

Understanding the Capability Bar

To the left of the histogram is a vertical bar chart representing the Capability Bar and answerd the question "How has the capability changed"?

The Histogram

This is a second graphical presentation for the distribution of points.  A histogram counts points within equal range groupings or intervals and displays the counts within these intervals in a bar chart format.

The histogram in QW differs slightly from the traditional histogram.  We have turned it on its side and displayed the sub-groupings or intervals over the standard intuitive color zones to make relating the control chart and histogram data simpler.   This gives you a clear picture as to the range of variation and where the process is currently centered. 

Understanding the Histogram

An important use of the histogram is to determine if the distribution of the data is normal.  A normal distribution is one that resembles a bell curve and is equal on both sides of the center or average.

Histogram Features

Color Zones - The Histogram uses the same color interpretation of the limits set in QWSETUP for a variable that is used consistently throughout QW. 

on target and in control

in control but closer to a control limit than target

out of control but within specification limit

out of specification limit

Intervals - QW divides the area between the control limits into nine equal intervals and divides the specification zones into as many equal intervals as will fit.  QW then accumulates the number of points in each interval and displays the information as a bar chart.

Limits - The actual target, control and specification limits are labeled on the histogram for reference for both the histogram and the control chart.

Highest Frequency - QW places a single number on the bottom of the histogram to indicate the number of points in the longest bar on the histogram.

The Statistics Box

The Statistics Box contains a grouping of key statistics that are very important in determining the status of the process for a particular variable.  For details on how the statistics are calculated see Statistics in QW 5.

QW uses the highlighting feature to identify significant values.  This allows you to zero in on a problem and provides you with a hint to its possible cause.   

Understanding the Statistics Box

It is important to view all the statistics within this area to obtain a clear understanding of the cause of an alarm. One statistic alone does not give the whole or complete story.  For instance, Variation describes how well a process is centered but does not mean that some points are not outside the specification limits.

@POP Notes

If a Calculated variable is defined with the @POP function the following differences should be noted for the Control Chart.

The @POP function combines all of the data currently in memory for the range of variables defined.

Statistics performed on @POP variables will treat all individual points in the range as its population

Example: @POP[V18..V23] will include all of the data for variables 18, 19, 20, 21, 22 and 23 on the Control Chart.

@POP Changes to the Control Chart

For each record (Date and Time entry) a vertical line will be displayed showing a horizontal line at the top of the vertical line for the maximum value, another horizontal line at the bottom of the vertical line for the minimum value and a blue dot for the calculated average of the data.

Example:        @POP[V18..V23] would be displayed as

Maximum value found in the V18 to V23 range

Average value in the V18 to V23 range

Minimum value found in the V18 to V23 range

Only the average value is stored in the data file and displayed on the Log Sheet.