How fast can capability be improved?

Centering issues often resolve in hours; variation reduction is typically a weeks-to-months project.

Process State

Process Capability Improvement

3 min read Last updated July 2026

Improving capability is a structured exercise. Skip the steps and the gains do not stick.

Open the Process Capability Calculator Quality Statistics Knowledge Center

The engineering question this page answers

Capability is short of target — what is the correct improvement sequence so that the gain is real, sustained, and defensible in PPAP?

Decision logic

Baseline: run capability, quantify Cp gap and Cpk gap separately ↓ MSA first — reject any improvement plan built on unqualified gauges ↓ Stability first — remove special causes before attacking common cause ↓ Split the problem: centering loss vs spread loss vs both ↓ If centering dominates → offset, wear compensation, warm-up rules (fast wins) ↓ If spread dominates → variance components + DoE (weeks-to-months) ↓ Implement via formal change control; do not touch multiple factors at once ↓ Confirm with a fresh capability study on a stable window ↓ Update Control Plan, PFMEA, reaction plan; monitor for regression

Engineering procedure

Baseline Cp, Cpk, Pp, Ppk with documented MSA and stability evidence.
Decompose the gap: k-based centering loss vs σ-based spread loss.
Prioritize centering first — it is faster and cheaper.
Address spread via variance components; do not skip to DoE without ranking sources.
Run a screening DoE on the vital few factors; hold others constant.
Confirm the effect size against a pre-registered target — not just "better".
Update tooling, program, fixture, or material spec via engineering change.
Re-run capability with the same sample size and estimator as the baseline for comparability.
Refresh Control Plan, PFMEA occurrence/detection, and reaction rules.
Set a re-verification checkpoint at the next tool life or PM boundary.

Typical failure modes

Changing multiple factors simultaneously; the true driver is lost.
"Improving" capability by widening the specification or removing a control point.
Reporting capability from a hand-selected window that excludes bad shifts.
Claiming improvement from Ppk when the sample plan actually measures Cpk (or vice versa).
Skipping MSA — the "gain" was a gauge fix all along.

Engineering insight

Improvement without a baseline is a story, not evidence. Always capture the pre-state under the same conditions as the post-state.
Solving centering first often changes which variation source dominates — do not commit to a DoE plan before re-baselining.
Improvements that require constant operator intervention are not improvements — they are hidden inspection.
Gains that appear on the first shift after a change often reflect novelty attention, not process change. Wait a week before claiming success.

When NOT to use this metric

Do not launch an improvement project before the process is stable — you will be chasing noise.
Do not use capability improvement as a substitute for a PFMEA update when the change alters risk.
Do not compare capability studies whose σ estimator or sampling design differs.

Relationship to other capability metrics

Cp gap vs Cpk gap: Cp gap = spread problem; Cpk-only gap = centering problem; both = combined.
Cpk vs Ppk trajectory: Both must improve together for a real gain. Cpk up, Ppk flat = new drift introduced.
Cpk vs sigma level: Convert to sigma to communicate improvement in customer-facing terms.

Recommended actions

Confirm stability and MSA.
Quantify the centering vs variation contribution.
Attack the dominant source (DoE if needed).
Update Control Plan and FMEA.
Re-run the capability study.
Use the free Process Capability Calculator to compute Cp, Cpk, Pp, Ppk, sigma level, and predicted PPM from your measurement data — no sign-up required.

Engineering notes

Never accept improvement claims without the pre and post datasets side by side.
Never close an improvement project before running one full production week under the new state.
Never remove a Control Plan characteristic to "improve" capability metrics.

Continue the investigation

cavity balance, cooling and process-specific improvement — Molding improvement depends on process-specific causes.

Sequence the work through Process Not Centered first, then Process Variation Too High, and validate the outcome via a fresh Process Capability Study. If the improvement was triggered by a customer complaint, run it under the discipline of an 8D Report and update the Process FMEA.

Verification checklist

Baseline Cp, Cpk, Pp, Ppk recorded with MSA and stability evidence
Gap decomposed: centering vs spread
One improvement lever changed at a time
Change routed through formal engineering change control
Post-state study run with same estimator and sampling as baseline
Sustained for at least one full production week
Control Plan, PFMEA and reaction plan updated

Assumptions and applicability

Process condition: statistical stability is required.
Distribution assumption: use a distribution model justified for the data.
Confirm process stability and measurement-system adequacy before interpreting a capability index.
Use a justified distribution model or non-normal method when the normal model is unsuitable.