Six Sigma Kaizen logo

Calculator Library / Capability

Process Capability Analyzer

Analyze capability using raw data points or summary statistics. This app calculates Cp, Cpk, Pp, Ppk, estimated ppm defective, a histogram with normal curve overlay, a normality screening flag, and a control-chart preview with what-if adjustments for centering and tolerance changes.

Analyzer

Choose an input method

Capability core: Cp = tolerance width / (6 x within sigma), Ppk = min(Ppu, Ppl)

Input mode

Paste values separated by commas, spaces, or line breaks. CSV uploads can contain a single numeric column or mixed text; the app extracts numeric values automatically.

Mean shift: 0.00

Tolerance factor: 1.00x

Raw-data mode estimates overall sigma from the full data set and within sigma from average moving range. Summary mode uses the entered overall sigma and defaults within sigma to that same value unless you provide a separate within estimate.

Histogram

Capability histogram with normal overlay

Control Chart

I-chart preview

Instructions

How to use this app

  1. Select Raw data points if you have individual measurements, or Summary statistics if you only have mean and sigma values.
  2. Enter or upload the data, then provide LSL, USL, and a target value.
  3. Click Analyze to calculate Cp, Cpk, Pp, Ppk, Cpu, Cpl, and estimated ppm defective.
  4. Use the centering and tolerance sliders to explore how process mean shifts or tolerance tightening affect capability.
  5. Use Export Control Chart to download the current control-chart preview as a PNG image.

Cp and Cpk use within sigma, which is meant to reflect short-term process spread. Pp and Ppk use overall sigma, which reflects long-term observed spread. If Ppk trails Cpk significantly, the process may be drifting over time even if short-term performance looks acceptable.

The normality flag is a screening aid based on skewness and kurtosis. It is useful for quick review, but it is not a substitute for a full normality study when the decision is high stakes.

What This Process Capability Calculator Does

This process capability calculator helps manufacturing and quality teams answer a practical question: can the current process consistently produce inside the specification limits? The app accepts either raw data or summary statistics, then converts that information into Cp, Cpk, Pp, Ppk, ppm defective, and a visual distribution view.

Use it when you are reviewing launch readiness, investigating variable quality, checking whether a process is centered, or preparing a capability summary for leadership, customers, or APQP documentation.

Core Capability Formulas

Metric Formula Meaning
Cp (USL - LSL) / (6 x within sigma) Potential short-term capability if the process is centered.
Cpk min[(USL - mean) / (3 x within sigma), (mean - LSL) / (3 x within sigma)] Actual short-term capability after centering is considered.
Pp / Ppk Same logic using overall sigma Overall long-term capability including drift and broader variation.
PPM defective Probability outside spec x 1,000,000 Estimated nonconformance rate at the current process spread.

Worked Example

Imagine a bore diameter with LSL = 8.50, USL = 11.50, mean = 10.00, and within sigma = 0.25. Cp would be 3.00 / 1.50 = 2.00, which means the spread is tight relative to the tolerance. If the mean shifts upward to 10.90, Cpk drops because the process is now much closer to the upper limit even though the raw spread did not change.

That is why capability review cannot stop at Cp alone. The centered potential can look excellent while the actual process still creates risk because of drift or poor centering.

How to Interpret the Results

Process Capability Frequently Asked Questions

What is the difference between Cp and Cpk?

Cp measures the potential capability based on spread alone. Cpk adjusts that view for centering, so it shows whether the process is actually running close to one of the limits.

What is the difference between Cpk and Ppk?

Cpk uses within variation and reflects short-term capability. Ppk uses overall variation and reflects long-term performance including drift, instability, and broader process noise.

Can a process have a good Cp and a bad Cpk?

Yes. That usually means the process spread is small enough, but the process mean is not centered between the limits.

When should raw data be used instead of summary statistics?

Use raw data whenever possible. It allows better checking of distribution shape, moving-range behavior, and the visual relationship between the process and the spec window.

What is a common mistake in capability analysis?

Running capability on an unstable process. If the process is not statistically stable, the capability numbers can look precise while giving a false sense of control.

Related Templates and Guides

Read the Standard Work Guide

Use capability data to decide whether the current method is actually stable enough to standardize and sustain.