To enable the SPC Check function on your manufacturing process flow you must first select the Step Definition and then the Name Value(s) that you wish to monitor using the SPC test(s) of choice.

In this example we will activate the Equal SPC Check to test the latest 7 data points (SPC Default number) if they are the same value. The data collected from the Step Definition Name Value(s) is tested only when the traveler is submitted to move to the next step. If the SPC Check results in a failure then the traveler can be blocked and a SPC Alarm Event configured to trigger and perform a customized action for the Step Definition e.g. send email to line manager and supervisor, put traveler on hold etc.

SPC_Check

Now select the Name Value that you wish to configure in the Step Definition.

SPC_Check_2

In the Name Value window select the SPC Report and Analysis Settings Tab and configure the test as illustrated below. To see the current results for this NV, select the print-report button at the top of the window. This shows a standard report with the results. How the results can be changed is shown below where the different settings are explained. This standard report, and the report that you want to create to take actions are two different ones.

SPC_Check_4

Note: The number of points that are tested as equal can be customized by entering a value in the 'Equal Number (Default 7)' entry section.

Common Cause and Special Cause Variation

  1. Special Cause variation means something different has happened at a certain time or place.
  2. Common Cause variation is always present to some degree in a process. The goal is to minimize the variation. 

It's very important to distinguish between Special and Common cause variation, because each requires a different strategy.

SPC Strategy
The strategy employed after a SPC Check returns an error depends on the type of error flagged for the system. One must determine if the error is either Special or Common Cause. Use the following link to find out more about SPC Special and Common Cause Strategy.

Any SPC Check can be activated following the same procedure as detailed above. Every SPC Check available is detailed below in the SPC Run Chart and SPC Control Chart sections.

 

SPC Run Chart Check
A 'run' is a series of points on the same side of the median. (Note: A run can be any length from 1 to any number of points).

SPC_Check_3-
SPC Run Chart Configuration set to run all 5 SPC checks using default entries

1. Number of Runs (Dependent on the number of data points)
Check this box to run the Number of Runs SPC check on your data. Too few or too many runs are important signals of special causes - they indicate that something in the manufaturing process has changed.

Number_of_Runs

 

2. Run Length (8 Data Points)
Check this box to activate the Run Length check on your data. Run length is a signal of a special cause in your process. In SPC analysis 8 or more points in a row on the same side of the median is a signal of a special cause in the process.

Run_Length

 

3. Trend Up/Down (6 Data Points)
Check this box to run the Trend Up/Down check on your data. When 6 or more data points in a row are continuosly increasing or decreasing ("trend") it is a signal of a special cause in your process.

Trend_Up-Down

 

4. Alternating (14 Data Points)
Check this box to run the Alternating check on your data. If 14 or more data points in a row are alternating up and down then this is a signal of a special cause in your process.

14_Alternating

 

5. Equal (7 Data Points)
Check this box to run the Equal check on your data. If 7 or more data points in a row are all the same value then this is a signal of a special cause in your process.

Equal_7

 

SPC Control Chart Check
A control chart plots time-ordered data(just like a run chart). Statistically determined control limits are drawn on the plot. The centerline calculation uses the mean instead of the median.

 
SPC_Check_5
SPC Control Chart Configuration set to run all 8 checks using default entries

1. Outside Upper Control Limit (UCL) or Lower Control Limit (LCL)*
If a datapoint lies outside of either the upper or lower control limits then this is a signal of a special cause in the process.
Note: A control limit defines the bounds of common cause variation in the process.
 
 
 
 
Point_above_UCL
 
 

 

2. Two of Three in Zone A
If 2 out of 3 data points are inside the Zone A Control Limits then this is a signal of a common cause variation.

2_of_3_in_Zone_A

 

3. Four of Five in Zone B
If 4 out of 5 data points are inside the Zone B Control Limits then this is a signal of a common cause variation.

4_of_5_in_Zone_B

 

4. Within Zone C (15 Data Points)
If 15 data points in a row are within the Zone C Control Limits then this is a signal of a common cause variation.

Within_Zone_C

 

5. Outside Zone C (8 Data Points)*
If 8 data points in a row are outside the Zone C Control Limits then this is a signal of a common cause variation.

Outside_Zone_C

6. Above/Below Mean (9 Data Points)
If 9 data points in a row are either above or below the mean then this is a signal of a common cause variation.

 9_Points_above_mean

7. Trend Up/Down (6 Data Points)*
If 6 data points in a row are trending either up or down then this is a signal of a special cause variation.

Trend_6

 

8. Alternating (14 Data Points)*
If 14 data points in a row are continuously alternating then this is a signal of a special cause variation.
Same as run chart alternating as illustrated above.

Useful links:

SPC Check Example

SPC Report

SPC Strategy

Pareto Report