Insights on statistical process control, quality engineering, and data-driven manufacturing
Six Sigma’s core insight — reduce variation to reduce defects — is timeless. But the normality default, manual data collection, and belt-certification gatekeeping need updating. Here’s what modern Six Sigma looks like with distribution-free methods and Quality 4.0.
First pass yield says 98.2%. Cpk says 0.94. One measures what happened. The other predicts what will happen next. When they disagree, something important is hiding — and knowing which to trust prevents costly mistakes.
Your PPAP got rejected — not for bad parts, but for bad statistics. OEM auditors now scrutinize whether your Cpk method matches your data. Build a PPAP capability evidence chain that withstands the toughest audits.
Distribution fitting replaces the normality assumption with a different guess. With typical sample sizes, Weibull, lognormal, and gamma all pass goodness-of-fit tests — giving different Cpk values. The distribution fitting step that should fix your analysis becomes its own error source.
Process drift hides under false alarms. Shewhart charts catch sudden shifts but miss gradual process drift — while Nelson rules fire on stable data. Entropy-based homogeneity testing separates real drift from noise without chart configuration.
Hidden clusters from multi-cavity molds, shift changes, and material lots produce aggregate Cpk that looks capable — while one subpopulation ships defects. ELDF detects what Cpk can’t see.
Minitab offers non-normal options. EntropyStat is distribution-free. Those aren’t the same thing. Offering a menu of distributions to choose from is distribution-flexible — not distribution-free. Here’s why that distinction determines whether your Cpk is correct.
With a small sample of 10 parts, traditional Cpk has a confidence interval 0.6 units wide — your 1.38 could be anywhere from 1.05 to 1.71. Entropy-based methods extract more from limited data without the normality assumption.
Your SPC software computes Cpk assuming data follows a bell curve — but 60–80% of manufacturing data doesn’t. That silent assumption produces capability numbers that are confidently wrong, costing real money in both directions.
Alarm fatigue isn’t a people problem — it’s a chart problem. These five control chart mistakes drive false alarm rates through the roof and teach operators to ignore real signals.
IATF 16949 requires “appropriate statistical methods” — not normal-based Cpk. For the 60–80% of datasets that fail normality tests, the standard actually supports distribution-free methods like entropy-based analysis.
What is EGDF? The Entropic Global Distribution Function skips distribution fitting entirely — learning your process shape from the data. Here’s how EGDF and ELDF work, when they outperform traditional methods, and why they produce reliable Cpk from small samples.
The mathematical proof that your Cpk is wrong: Kolmogorov–Smirnov convergence, Chebyshev’s inequality, and why σ-based intervals fail when the empirical CDF diverges from the Gaussian hypothesis. A master’s-level walkthrough of what the Cpk normality assumption actually assumes — and where it breaks
When a KS test rejects the Cpk normality assumption, the ±3σ interval stops meaning 99.7%. Here are five real process shapes where sigma-based reasoning falls apart — and what your data is actually telling you.
The Cpk normality assumption fails for 60–80% of industrial datasets. Here’s what that costs you — and how distribution-free methods fix it.