Process Capability Analysis

Process Capability assesses whether a stable process can meet specification limits — quantifying how well natural process variation fits within the tolerance band.


Step 1 — Provide data and specifications

  • Process data — measured values
  • USL — upper specification limit
  • LSL — lower specification limit
  • Target (optional) — ideal value
  • Subgroup size (default 1)

Supports Excel Import, Sample Data Generator, and Manual Entry.


Step 2 — Results

Capability indices (within-subgroup variation):

  • Cp — potential capability (assumes centred)
  • Cpu / Cpl — one-sided
  • Cpk — actual capability (accounts for centring)

Performance indices (overall variation):

  • Pp, Ppu, Ppl, Ppk

Defect estimates:

  • PPM above USL, below LSL, total
  • Percentage out of specification

Visualisations: Histogram with normal curve and spec limits.

A UniversalChatBot is available for discussion.


Statistical methods used

Capability indices:

  • Cp = (USL - LSL) / (6 * withinSD)
  • Cpk = min((USL - mean)/(3*withinSD), (mean - LSL)/(3*withinSD))

Within SD: From mean moving range (MR-bar / 1.128) or mean subgroup range (R-bar / d2).

Interpretation:

Cpk Label
>= 2.00 Excellent (Six Sigma)
1.67 - 2.00 Very Good
1.33 - 1.67 Good
1.00 - 1.33 Marginal
< 1.00 Poor (not capable)

Defect rate: Based on normal distribution z-scores at spec limits.

Important: Capability indices only meaningful if process is in statistical control first.