Verifying a factory's Statistical Process Control (SPC) implementation requires a systematic approach that examines documentation, data, practices, and results. Here’s a step-by-step guide to ensure SPC is effectively implemented and driving quality improvements:

• SPC Plan: Check for a documented SPC plan outlining:
  • Critical processes/products to control.
  • Key characteristics to monitor (e.g., dimensions, defects).
  • Sampling plans (subgroup size, frequency).
  • Control charts used (e.g., X-bar/R, I-MR, p-charts).
  • Reaction plans for out-of-control signals.
• Control Chart Specifications: Verify:
  • Correct chart types for each process.
  • Rational subgrouping (e.g., samples taken close in time).
  • Clear control limits (calculated from process data, not specs).
• Training Records: Confirm operators/technicians are trained on SPC concepts and chart interpretation.

Audit Chart Integrity & Data Accuracy

• Data Collection:
  • Observe data collection in real-time (e.g., check if measurements are taken per plan).
  • Verify data is recorded accurately (no rounding errors, transcription mistakes).
  • Ensure data is timely (not back-filled).
• Control Chart Calculations:
  • Recalculate control limits for a subset of charts to verify accuracy.
  • Check if limits are based on initial process data (e.g., 25+ subgroups) and updated after significant process changes.
  • Ensure specification limits are not used as control limits.
• Chart Maintenance:
  • Verify charts are updated regularly.
  • Check for out-of-control points marked and investigated (e.g., notes on root causes).

Assess Process Control & Reaction

• Out-of-Control Response:
  • Ask operators/managers how they react to signals (e.g., "What happens if a point is outside control limits?").
  • Review reaction logs: Are investigations documented? Corrective actions implemented?
  • Ensure actions address root causes (e.g., tool wear, material changes), not just "adjusting the process."
• Process Stability:
  • Analyze charts for patterns (e.g., trends, cycles, shifts). Stable processes show random variation.
  • Use statistical tests (e.g., runs tests) if needed to confirm stability.
• Link to Corrective Actions: Verify SPC triggers corrective actions (e.g., preventive maintenance, material inspection).

Evaluate SPC Integration

• Operator Engagement:
  • Do operators understand their charts? Can they interpret signals?
  • Is SPC part of daily routines (e.g., morning checks)?
• Management Review:

  Is SPC data reviewed in management meetings? Is performance tracked?

• SPC in Quality Systems:
  • Confirm SPC aligns with ISO 9001, IATF 16949, or other standards.
  • Check if SPC results feed into PPAP submissions or customer reports.

Verify Impact on Quality & Cost

• Defect Reduction:
  • Compare defect rates before/after SPC implementation.
  • Check if out-of-control signals correlate with defect spikes.
• Process Capability:
  • Review Cp/Cpk values. Processes should be capable (Cpk ≥ 1.33).
  • Verify capability studies are performed for critical characteristics.
• Cost Savings:

  Quantify reductions in scrap, rework, or inspection costs linked to SPC.

Common Red Flags During Verification

• ❌ Charts are "wallpaper": No updates, no reactions to signals.
• ❌ Using spec limits for control: e.g., UCL = USL.
• ❌ Overly tight control limits: Artificially signals out-of-control conditions.
• ❌ Ignoring special causes: No action taken for out-of-control points.
• ❌ Data tampering: Erasures, inconsistent rounding, or "adjusted" values.
• ❌ Lack of training: Operators don’t understand chart interpretation.

Key Tools for Verification

• Checklists: Use audit checklists tailored to SPC standards (e.g., AIAG SPC manual).
• Data Sampling: Randomly pull 10–20 charts for recalculation and review.
• Interviews: Talk to operators, supervisors, and quality engineers.
• Spot Checks: Observe data collection live on the shop floor.

Example Verification Scenario

Process: Machining of a critical shaft diameter.
SPC Plan: X-bar/R chart, subgroup size 5, hourly sampling.
Verification Steps:

1. Recalculate control limits for 3 charts using raw data.
2. Check if operators investigate points beyond ±3σ.
3. Review reaction logs for the last 5 out-of-control events.
4. Compare defect rates pre/post-SPC implementation.
5. Ask operator: "What do you do if a point is above the UCL?"

Conclusion

Effective SPC verification goes beyond checking charts—it requires confirming that SPC is embedded in daily operations, driving decisions, and improving quality. If SPC is only a paperwork exercise, it fails its core purpose: reducing variation and preventing defects. Focus on evidence of real-time control, root cause analysis, and measurable impact to validate success.