A CUSUM chart with
two 1-sided CUSUMs consists of the following:
· Plotted
points (dots), which is the upper CUSUM, the cumulative sum of
deviations
above the target. The upper CUSUM detects upward trends in the process
level.
· Plotted
points (squares), which is the lower CUSUM, the cumulative sum of deviations
below the target. The lower CUSUM detects downward trends in the process
level.
· Center line,
which is located at zero. The plotted points should fluctuate randomly
around the center line.
· Control limits,
which are located (by default) 4 s
from the center line, provide a visual means for assessing whether the
process is in control.
Points located beyond the control limits signal an out-of-control
condition.
You should examine the CUSUM charts for
· Upward
or downward trends in the upper and lower CUSUMs. The plotted points should
fluctuate randomly around zero. If an upward or downward trend develops,
it is evidence that the process mean has shifted and you should look for
special causes.
· Plotted
points located beyond the control limits, which would indicate that the
process is out of control. Minitab displays out-of-control points in red.
You should initiate a search for potential causes for the special-cause variation.
Example Output |
|
Interpretation |
|
For the sample data, the lower CUSUM signals an out-of-control condition
from subgroup 24 on. These out-of-control points, which are indicated
by the red triangles, fall below the lower control limit. Thus, a search
for special causes should ensue.