There is no universally mandated formula in aerospace (including AS9100/AS9110/AS9120) for calculating corrective action effectiveness on NCRs. Instead, organizations define their own measures in procedures and QMS tools, then use a combination of lagging and leading indicators.
Start with a clear definition of “effective”
For aerospace NCRs, a corrective action is typically considered effective if it:
- Prevents or materially reduces recurrence of the same or similar nonconformance.
- Does not create new, related failure modes or escapes.
- Is implemented and sustained in the production or MRO environment (not just on paper).
Effectiveness should be evaluated at the cause level (root cause / contributing cause), not only at the single NCR number level.
Core metric: recurrence rate of like nonconformances
The most direct quantitative measure is recurrence of similar NCRs after the corrective action due date:
- Define the family: Use standardized codes (defect codes, operation codes, component families, supplier, etc.) to group “similar” NCRs.
- Set a baseline window: For example, the 6 or 12 months before implementation of the corrective action.
- Normalize by exposure: Use opportunities for defect as a denominator (e.g., number of parts, operations, flight hours, shop visits).
A typical working formula might look like:
Recurrence rate (post-CA) = (NCRs in family after CA) / (units or operations after CA)
Effectiveness is then evaluated by comparing the post-corrective-action rate to the baseline rate, using thresholds defined in your procedure (for example, >80% reduction sustained for 6–12 months).
Useful supporting metrics
To avoid relying on a single metric, many aerospace organizations track a small set of indicators per corrective action / RCCA:
- Defect rate trend: Defects per million opportunities (DPMO) or per 1,000 operations, baseline vs. 3, 6, 12 months after CA.
- Escape rate: Number of customer-found or field-found issues related to the same cause vs. factory-found issues.
- Repeat NCR indicator: Flag whether any NCR with the same root cause or same cause code occurred after the CA was closed.
- Containment robustness: Whether any similar defect escaped during interim actions (before permanent CA implementation).
- Implementation timeliness: % of CA actions completed by the planned due date and verified in the line / cell.
- Residual COPQ: Scrap, rework, MRB hours, or delay directly tied to the cause family before vs. after CA.
These can be combined into a simple internal effectiveness score or dashboard, but the score should not replace direct review of recurrence and risk.
Example: basic effectiveness scoring model
Many teams use a lightweight, procedure-defined scoring scheme for each closed corrective action, such as:
- Recurrence:
- 0 points: No similar NCRs in 12 months, normalized for volume.
- 1 point: 1–2 low-severity recurrences with decreasing trend.
- 2 points: >2 recurrences or any severe repeat event.
- Escape / customer impact:
- 0 points: No related customer or field issues.
- 1 point: 1 minor related customer issue.
- 2 points: Any safety, regulatory, or major customer impact.
- Implementation & sustainment:
- 0 points: All actions completed on time, verified in production/MRO, and controls integrated into WI/MES/QMS.
- 1 point: Minor delays or partial verification.
- 2 points: Significant slippage, control not fully embedded.
Total score 0–1 might be treated as “effective”, 2–3 as “needs monitoring”, and 4–6 as “ineffective / requires further action”. The thresholds and time windows must be defined in your QMS and applied consistently across programs, plants, and suppliers.
Consider severity and risk, not just counts
For aerospace, a corrective action that prevents a low-cost cosmetic defect and one that addresses a potential safety or airworthiness issue are not equivalent. An effectiveness assessment should consider:
- Risk classification: Safety, regulatory, certification, functional, cosmetic.
- Criticality of affected parts or systems: Flight safety parts, critical characteristics, key characteristics, life-limited parts.
- Detection location: In-process, final inspection, customer receiving, in-service.
Many organizations use a risk-prioritized approach where higher-risk causes require longer monitoring windows and tighter acceptance criteria for claiming effectiveness.
Brownfield and system integration realities
In most aerospace environments, NCRs, CAPAs, and RCCA actions are spread across QMS tools, MES, ERP, and sometimes spreadsheets. This directly affects how well you can calculate and trust effectiveness metrics.
Common constraints include:
- Inconsistent coding: Different plants or programs use different defect codes, making it hard to group “similar” NCRs.
- Fragmented traceability: Part, operation, and supplier identifiers are not harmonized across MES/ERP/QMS, undermining normalization by exposure.
- Data quality and late entries: Backdated or incomplete NCRs can distort trend analysis and time windows.
- Legacy systems: Older MES or paper travelers may not support structured capture of cause, action owner, or verification details.
Before relying on numerical effectiveness calculations, it is important to:
- Standardize defect and cause coding across sites as much as practical.
- Define explicit rules for what constitutes a “similar” NCR.
- Align NCR identifiers, part numbers, and operation IDs across systems where possible.
- Document the data sources and known gaps used for your metrics.
Why “full replacement” tools rarely solve effectiveness by themselves
Buying a new QMS or MES and trying to replace everything at once rarely fixes corrective action effectiveness in aerospace. Qualification and validation burdens, downtime risk, and complex integrations with existing ERP, PLM, and customer portals mean most organizations operate in a mixed environment for many years.
In practice, the most sustainable approach is usually:
- Incrementally improving NCR and RCCA workflows within current systems.
- Adding light integration or reporting layers that consolidate defect and action data.
- Improving governance around coding, risk ranking, and verification sign-off.
Effectiveness then becomes less about the specific tool and more about disciplined process, consistent data, and management review.
Governance and review expectations
To make any calculation credible for aerospace customers and auditors, your process for evaluating effectiveness should be:
- Documented: Criteria, formulas, and time windows defined in procedures or work instructions.
- Traceable: Each corrective action shows baseline data, post-implementation data, and who performed the review.
- Consistent: The same method applied across programs and suppliers unless a justified exception is recorded.
- Risk-based: More stringent criteria for high-risk, safety, or regulatory-related nonconformances.
This approach does not guarantee any regulatory outcome, but it does provide a defensible, transparent framework that aligns with typical AS9100 expectations around data-driven corrective action and continual improvement.
