Synthetic Data for Manufacturing SPC / Cpk
A measurement column of independently random floats can't sit at a declared Cp/Cpk, can't split into the right within/between variance an Xbar-R chart actually estimates from, and can't trigger a Western Electric Rule violation at a believable rate. Misata generates subgroup data where sigma is estimated the real way (Rbar/d2), control limits come from a Phase I baseline, and one process carries a genuine tool-wear drift that trips the rules exactly where it should.
The tables Misata generates
processesFive processes with real spec limits and a target Cpk spanning excellent to failingsubgroupsXbar, range, and three Western Electric rule flags per subgroupmeasurementsThe 5 raw readings behind every subgroupcapability_summaryMeasured Cp/Cpk and control limits from each process's Phase I baselineWhat holds true, every time
- measured Cpk (from a real Xbar-R baseline) reconciles to the declared target on every process
- Cp >= Cpk always, with one process built off-center specifically so Cp measurably exceeds Cpk
- one process is a genuine failing example, measuring below the 1.33 capability minimum
- an engineered tool-wear drift trips Western Electric rules on 100% of its final subgroups and 0% before it starts
Frequently asked
Do I need real manufacturing SPC data to generate this?
No. Misata builds the dataset from a specification, not a sample. There is no real manufacturing SPC data to source, anonymize, or leak. You describe the tables you need and the engine constructs them with referential integrity and realistic distributions.
Is the generated manufacturing SPC data privacy safe?
Yes, by construction. Nothing is learned from real records, so there is no membership to infer and nothing to leak. It runs entirely on your machine with no API key for the core engine.
Can I control the outcomes, like rates and totals?
Yes. Declare a target such as a monthly volume curve or an event rate and Misata produces rows that hit it exactly, while foreign keys stay intact and roll-up columns reconcile after a JOIN.
How is Cpk actually calculated here?
The real way an Xbar-R chart estimates it: sigma_within = Rbar/d2 (d2=2.326 for a subgroup of 5, the standard ASQ/Montgomery control-chart-constants table), not the pooled standard deviation of every raw measurement. Cp = (USL-LSL)/(6 x sigma); Cpk = Cp x (1-k), where k is how far off-center the mean sits.
What are the Western Electric Rules?
Standard out-of-control detection rules from the 1956 Western Electric Statistical Quality Control Handbook. This dataset checks three: Rule 1 (a point beyond the 3-sigma control limit), Rule 3 (2 of 3 consecutive points beyond 2-sigma, same side), and Rule 4 (8 consecutive points on the same side of the centerline).
Is the out-of-control signal real, or just randomly flagged?
Real. One process is given a genuine, engineered tool-wear drift over its last third of subgroups -- the most common real out-of-control pattern on a shop floor -- and the rules are checked against it: they fire on 100% of the final subgroups and stay silent (0%) before the drift starts, a signature independently random columns can't produce.
Is there a free sample manufacturing SPC dataset?
Yes, at https://misata.studio/datasets/manufacturing-spc: 5 processes, 450 subgroups, and 2,250 measurements, public domain (CC0), no signup. The zip includes an INTEGRITY.txt with every capability figure re-measured against the exact files.
Choosing a tool? How Misata compares