SPC Charts
The SPC (Statistical Process Control) Charts processor analyzes a window of process data and calculates statistical measures that assess a process's performance and capability against its specification limits, using control charts and predefined rules.
Why use SPC Charts
Use this processor to monitor whether a process stays within its specification limits and to catch early warning patterns using Nelson Rules, rather than only watching a raw value against fixed thresholds.
How it works
The processor calculates a current mean and standard deviation (stdDev) from the input data within the window, then derives control limits and capability indices from them:
- Upper Specification Limit (USL) and Lower Specification Limit (LSL) are calculated from the current mean and standard deviation.
- Process capability (Cp) is the ratio of the difference between USL and LSL to six times the standard deviation (6σ).
- CPu (Upper Capability Index) and CPl (Lower Capability Index) measure how centered the process is within the specification limits: CPu = (USL - average) / 3σ, and CPl = (average - LSL) / 3σ.
- Process performance (Cpk) is the smaller of CPu and CPl.
- If valid data is present, the processor also calculates Cpm and Cpkm, which incorporate Target Mean.
If your input tag doesn't publish at a steady rate, use Time Interval to publish a value on a fixed schedule instead of waiting for the next incoming event. Enter 0 to disable the timer if your input already publishes at the expected rate.
Nelson Rule violations
The processor flags a true value for each Nelson Rule your data violates:
- One point is more than three standard deviations from the mean.
- Nine or more points in a row fall on the same side of the mean.
- Six or more points in a row continually increase or decrease.
- 14 or more points in a row alternate direction, increasing then decreasing.
- Two out of three, or three out of three, points in a row fall more than two standard deviations from the mean in the same direction.
- Four out of five, or five out of five, points in a row fall more than one standard deviation from the mean in the same direction.
- 15 points in a row all fall within one standard deviation of the mean, on either side.
- Eight points in a row occur, with none within one standard deviation of the mean, and the points fall in both directions from the mean.
Parameters
Parameter | Details |
|---|---|
Window Size | The number of data points considered for calculations. Affects the current mean, standard deviation, and specification limits. Defaults to 50. |
Target Mean | The desired average value for the process, used as a reference point for evaluating performance. Required, along with valid data, to calculate Cpm and Cpkm. Defaults to 100. |
Pass Through Value | Determines whether the processor's output is the calculated control chart fields or the original unchanged input value. |
Time Interval | Sets a timer that publishes output on a fixed schedule if the input tag isn't polling regularly. Enter 0 to disable the timer. |