Executive Snapshot
Without stability, a process cannot be predicted or improved. When SPC and control charts are missing, teams
chase noise and miss signals. Problems are discovered late, usually through scrap, rework, or customer complaints.
SPC is the early warning system of the invisible factory. It separates common cause from special cause, and it gives
operators the confidence to act before defects pile up. Without it, firefighting replaces foresight.

Why it happens
SPC often gets sidelined for three reasons. First, it is seen as ‘statistical jargon’ that only quality engineers
understand. Second, managers view it as overhead compared to lagging metrics like yield or PPM. Third,
implementation is done poorly — wrong subgrouping, noisy data, or charts without reaction plans — so operators
lose trust. The result: SPC is abandoned, and teams run blind.

How it shows up
Unstable processes reveal themselves through constant firefighting. SPC charts, if they exist, show frequent rule
violations. Capability studies give contradictory results between runs. Operators don’t know whether to adjust or
wait. Meetings are filled with debate over whether a shift was real or just ‘noise.’ Meanwhile, bad parts slip through
undetected until customers raise complaints.

Consequences
When SPC is weak or absent, process capability is misunderstood. Engineers waste time chasing false signals.
Operators lose confidence in the data. Quality costs rise as problems are detected late in inspection or, worse, by
the customer. Management ends up relying on lagging indicators — a dangerous illusion of control.

The fix
The fix is to anchor SPC on critical-to-quality (CTQ) characteristics. Use proper subgrouping that reflects how the
process produces. Train operators in the basic rules for detection — trends, runs, and points beyond limits. Most
importantly, tie every chart to a reaction plan that defines what to do within 1–2 hours of a signal. SPC is not
paperwork; it is the nervous system of a stable factory.

Root causes (6M+E)
Dimension Typical issues
Measurement No SPC on CTQs; wrong subgrouping; noisy or unstable gauges.
Methods No control plan; unclear reaction plans; over-adjusting instead of letting process run.
Machines Unstable settings; drift; poor maintenance hides real variation.
Materials Inconsistent supply quality; uncontrolled lot-to-lot variation.
Manpower Operators not trained in SPC interpretation; charts ignored or filled in after the fact.
Environment Temperature, humidity, or vibration introduce uncontrolled variation.

Diagnostics & quick checks

• Confirm SPC charts exist for all CTQs in the control plan.
• Check if charts show stable periods with only common cause variation.
• Look for documented reaction plans tied to specific rule violations.
• Verify operators know what to do when a signal occurs.
• Audit whether reaction happens within 1–2 hours of signal.
  Acceptance criteria
  Stable charts show no rule violations and only common cause variation. Out-of-control signals should trigger
  documented reaction within 1–2 hours. Capability studies should be run only on stable processes. Operators and
  engineers should interpret signals consistently with SPC rules.
  30/60/90-day playbook
  0–30 days
• Identify CTQs and set up basic control charts.
• Train engineers and operators in SPC rules of detection.
• Pilot SPC on one high-risk process; link to a simple reaction plan.
  31–60 days
• Expand SPC to other CTQs; standardize subgrouping methods.
• Develop clear reaction plans and embed them in work instructions.
• Audit first responses to signals; track reaction time.
  61–90 days
• Integrate SPC metrics into management reviews.
• Tie SPC signals into continuous improvement projects.
• Reduce reliance on end-of-line inspection by using SPC as early warning.
  Sustain & scale
  Sustain SPC by embedding it into layered audits and control plans. Link recurring signals to structured problem
  solving. Review reaction effectiveness regularly and adjust plans as needed. When SPC is treated as a living
  system, not a chart-filling exercise, the invisible factory gains foresight — and firefighting declines.
  References & further reading
  [1] NIST/SEMATECH e-Handbook of Statistical Methods — Control Charts
  https://www.itl.nist.gov/div898/handbook/pmc/section3/pmc3.htm
  [2] AIAG SPC Manual (2nd Edition)
  https://www.aiag.org/quality/automotive-core-tools/spc
  [3] ASQ Quality Resources — Statistical Process Control
  https://asq.org/quality-resources/statistical-process-control