Workflow Modules
The SPC Engine scores each sample as you log it: outliers past three sigma, or eight points on one side of the mean. You make the call.
How it works
Each sample keeps its time, its value and who captured it, tied to the lot and the characteristic. Its statistic is stored with it, so the lot view and the history read the same points.
Capture
Log the reading right at the line, tied to the lot that’s running. Every point is stored, not just the ones that trip a flag, so if a value starts drifting on second shift, it’s on record before anyone asks.
Outliers
Every point you log is scored against the recent mean and spread. Anything past three sigma is flagged, and the flag carries its z score so you can see how far out the reading landed.
Trend rule
Eight points on one side of the mean read as drift. You get the flag while every reading is still inside three sigma, with the run length shown, so you can see the trend build before it becomes an outlier.
Lots
Open the chart for the lot that’s running, or go back through the full history behind one characteristic. Any commentary the shift left on the run stays beside the points it explains.
Checks and chart
You get two tests and one chart. Xbar-R charts, Cp and Cpk aren’t included, so keep the package you use for those.
Flag review
A flag is a starting point, recorded with the point that raised it. What it means for the lot is your quality process’s call. That call might open a case in the Investigation Engine, or line up with what the Auditor sees on a lot check.
You log the reading at the line, against the lot that's running. The point is captured whether or not it trips a flag, so second-shift drift is already on record.
Eight readings landing on one side of the mean count as drift. The flag comes while every point still sits inside three sigma, so the shift sees the trend building early.
Two. An outlier test at three sigma on every point, and a run test that flags eight points on one side of the mean.
No. There are no capability indices and no chart families, so keep the package you use for those.
Three sigma by default, and you set it per characteristic. On thin data the flag is held back until there are enough points to mean anything.
Readings captured at the line against the running lot. Each one is stored with its statistic, so the lot view and the history are built from the same points.
30 minutes on your samples and how the SPC Engine scores them against the recent mean.
We use cookies to understand how the site is used and to connect form submissions to earlier visits. No analytics or marketing cookies are set until you choose. Privacy Policy