Process Capability Calculator

What Is Process Capability?

Process capability measures how well a process can produce output that meets specifications. It compares the voice of the process (what the process naturally produces) to the voice of the customer (specification limits). A capable process consistently produces output within specifications with minimal defects.

Process capability analysis is fundamental to quality engineering, helping organizations understand whether their processes can meet customer requirements and where improvement efforts should focus.

Prerequisites

Before calculating process capability:

1. The process must be in statistical control (use control charts to verify)
2. Data should be approximately normally distributed
3. Specification limits must be defined by customer requirements

Capability indices calculated from an unstable process are meaningless—they'll change unpredictably.

Key Capability Indices

Cp (Process Capability)

$C_p = \frac{USL - LSL}{6\sigma}$

Cp compares the specification width to the process spread (±3σ). It measures potential capability if the process were perfectly centered.

• $C_p = 1.0$: Process spread equals specification spread
• $C_p = 1.33$: Typical minimum requirement
• $C_p = 2.0$: Six Sigma capable

Limitation: Cp ignores process centering.

Cpk (Process Capability Index)

$C_{pk} = \min\left(\frac{USL - \mu}{3\sigma}, \frac{\mu - LSL}{3\sigma}\right)$

Cpk accounts for both spread and centering by taking the minimum of upper and lower capability. It measures actual capability.

• $C_{pk} = C_p$: Process is perfectly centered
• $C_{pk} < C_p$: Process is off-center
• $C_{pk}$ can be negative if the mean is outside specifications

Cpm (Taguchi Index)

$C_{pm} = \frac{USL - LSL}{6\sqrt{\sigma^2 + (\mu - T)^2}}$

Where $T$ is the target value. Cpm penalizes deviation from target, not just from specification limits—aligned with Taguchi's quality philosophy that any deviation from target causes loss.

Cpu and Cpl (One-Sided Indices)

$C_{pu} = \frac{USL - \mu}{3\sigma} \quad \text{(upper)}$ $C_{pl} = \frac{\mu - LSL}{3\sigma} \quad \text{(lower)}$

Used when only one specification limit exists (e.g., "strength must exceed X").

Interpreting Capability Indices

Industry standards typically require:

• $C_{pk} \geq 1.33$ for existing processes
• $C_{pk} \geq 1.67$ for new processes or safety-critical applications

Sigma Level and PPM

The sigma level relates to defects per million opportunities (PPM):

$\text{Sigma Level} = 3 \times C_{pk}$

$\text{PPM} = 1,000,000 \times [1 - \Phi(3 \times C_{pk}) + \Phi(-3 \times C_{pk})]$

Where $\Phi$ is the standard normal CDF.

Note: "Six Sigma" methodology assumes a 1.5σ shift, so a "6 sigma process" actually produces 3.4 PPM defects, not the 0.002 PPM of a truly centered 6σ process.

Practical Applications

Manufacturing

Ensure machined parts meet dimensional tolerances, chemical processes maintain concentration specs, and assembly processes achieve proper fits.

Pharmaceuticals

Verify drug potency, tablet weight, and dissolution rates meet strict regulatory specifications.

Service Industries

Measure delivery times against service level agreements, call handling times, and transaction accuracy.

Improving Process Capability

When Cpk is inadequate:

1. Reduce variation (σ): Increases both Cp and Cpk

   • Better equipment
   • Standardized procedures
   • Operator training
   • Material consistency

2. Center the process (μ → target): Increases Cpk toward Cp

   • Process adjustment
   • Better targeting

3. Widen specifications: If genuinely possible without compromising function

   • This is often not an option

Pp and Ppk: Performance Indices

$P_p$ and $P_{pk}$ use overall standard deviation (including between-subgroup variation) rather than within-subgroup variation:

$P_p = \frac{USL - LSL}{6s} \quad P_{pk} = \min\left(\frac{USL - \mu}{3s}, \frac{\mu - LSL}{3s}\right)$

• Cp/Cpk: Short-term or potential capability (within-subgroup σ)
• Pp/Ppk: Long-term or actual performance (overall s)

If $P_{pk} < C_{pk}$, special causes are adding variation over time.

Non-Normal Data

For non-normal data, options include:

• Transform data (e.g., Box-Cox) to achieve normality
• Use percentile-based methods (Cp based on 0.135 and 99.865 percentiles)
• Use distribution-specific capability indices

Related Calculations

• Control Chart Calculator — Verify stability first
• Standard Deviation Calculator — Process spread
• Normal Probability Calculator — PPM calculations
• Box-Cox Transformation — Handle non-normality

Summary

Process capability indices quantify how well a process meets specifications. Cp measures potential (if centered), Cpk measures actual capability, and Cpm penalizes deviation from target. Before calculating capability, ensure the process is stable using control charts. A capable process ($C_{pk} \geq 1.33$) consistently produces output within specifications, while incapable processes require improvement through variation reduction or centering. These indices provide a common language for quality improvement across industries.