AGENT SKILL

spc-control-charts

Statistical Process Control (SPC) — select the correct control chart, interpret out-of-control signals using Western Electric rules, calculate and interpret Cp, Cpk, Pp, Ppk. Use when setting up SPC for a new characteristic, interpreting control chart signals, responding to special cause variation, or auditing SPC implementation. Covers AIAG SPC 2nd edition and IATF 16949 §8.3.3.

适用职业数据分析师
使用场景分析与可视化
License 许可
MIT
最近更新
2026-11-14
Stars
★ 1.6k
Forks
491

01 前言 Front Matter

name
spc-control-charts
license
MIT
metadata
{ "author": "RBraga01", "domain": "quality-engineering", "status": "approved", "created": "2026-06-06", "version": "1.1", "iso-9001": "9.1", "subdomain": "measurement", "iatf-16949": "8.3.3, 9.1.1", "industries": "automotive,electronics,aerospace,medical,general", "updated_by": "migmcc", "reviewed_by": "RBraga01", "last_updated": "2026-06-06", "aiag-reference": "AIAG SPC 2nd Edition", "standard_edition": "AIAG SPC 2nd Edition (2005)" }
description
Statistical Process Control (SPC) — select the correct control chart, interpret out-of-control signals using Western Electric rules, calculate and interpret Cp, Cpk, Pp, Ppk. Use when setting up SPC for a new characteristic, interpreting control chart signals, responding to special cause variation, or auditing SPC implementation. Covers AIAG SPC 2nd edition and IATF 16949 §8.3.3.

02 资源文件 1

03 SKILL.md 正文

Statistical Process Control (SPC)

When to use

Use this skill when:

  • Selecting the correct control chart type for a process characteristic
  • Interpreting control chart signals — is this special cause or common cause?
  • Calculating Cp, Cpk, Pp, Ppk and determining if the process is capable
  • Responding to an out-of-control signal on a production line
  • Setting up SPC for a new special characteristic (PPAP/APQP requirement)
  • Auditing an SPC system for correctness and adequacy
  • Explaining SPC results to a customer or during an audit

Prerequisites

  • The characteristic to monitor (variable or attribute data)
  • Production data (minimum 25 subgroups for control limits, 100+ pieces for capability)
  • MSA study completed and %GRR < 30% for variable charts
  • Process specification (nominal + tolerance) for capability calculations
  • Subgroup size decided based on rational subgrouping principle

Workflow

Step 1 — Select the correct control chart

Variable data (measured values — length, weight, pressure, temperature)

Chart When to use
X̄-R (X-bar/Range) Subgroup size 2–9; most common in manufacturing
X̄-S (X-bar/Sigma) Subgroup size ≥ 10; better sensitivity to spread
I-MR (Individuals/Moving Range) Subgroup size = 1; one measurement per inspection (slow processes, destructive tests)

Attribute data (counts, pass/fail, defect rates)

Chart When to use
p-chart Proportion defective; variable subgroup size
np-chart Number defective; constant subgroup size
c-chart Count of defects per unit; constant inspection area
u-chart Defects per unit; variable inspection area

Decision rule: If you can measure it with a number, use a variable chart. Variable charts are more sensitive and require smaller samples to detect process shifts.


Step 2 — Calculate control limits

X̄-R chart

From at least 25 subgroups of size n:

  • Centre line (X̄̄): Grand average of all subgroup means
  • UCL_X̄ = X̄̄ + A₂ × R̄
  • LCL_X̄ = X̄̄ − A₂ × R̄
  • Centre line (R̄): Average of all subgroup ranges
  • UCL_R = D₄ × R̄
  • LCL_R = D₃ × R̄ (= 0 for n ≤ 6)

Constants for common subgroup sizes:

n A₂ D₃ D₄
2 1.880 0 3.267
3 1.023 0 2.574
4 0.729 0 2.282
5 0.577 0 2.114

Important: Control limits are calculated FROM THE DATA — never set them to match the specification limits. Specification limits and control limits are completely separate concepts.


Step 3 — Apply Western Electric Rules (out-of-control signals)

Divide the control chart into zones: Zone A (2–3σ from centre), Zone B (1–2σ), Zone C (0–1σ).

Rule Signal Interpretation
Rule 1 1 point beyond 3σ (outside control limits) Large, immediate shift
Rule 2 9 consecutive points on same side of centre line Process mean has shifted
Rule 3 6 consecutive points steadily increasing or decreasing Trend — tool wear, drift
Rule 4 14 consecutive points alternating up and down Systematic variation — two alternating distributions
Rule 5 2 of 3 consecutive points in Zone A or beyond (same side) Large shift signal
Rule 6 4 of 5 consecutive points in Zone B or beyond (same side) Moderate shift
Rule 7 15 consecutive points in Zone C (either side of centre line) Stratification — data from two separate distributions
Rule 8 8 consecutive points beyond Zone C (either side) Mixture — sampling from two processes

Action required for ANY rule violation: Stop and investigate immediately. Do not reset control limits. Do not restart until root cause is identified.

Most commonly applied in automotive: Rules 1, 2, 3 minimum. Rules 1–8 for safety-critical characteristics.


Step 4 — Calculate and interpret process capability

Capability data requirements: Process capability is only valid when calculated on a process that is in statistical control (no out-of-control signals) and with a minimum of 100 consecutive parts from that stable process. Fewer parts produce unreliable Cpk estimates — a Cpk calculated on 30 parts can vary by ±0.3 from the true value. Do not report Cpk based on fewer than 100 parts as a production capability figure; label it "preliminary" and state the sample size.

Short-term capability (within-subgroup variation):

  • Cp = (USL − LSL) / (6σ̂) — capability — process spread vs. tolerance
  • Cpk = min[(USL − X̄̄) / (3σ̂), (X̄̄ − LSL) / (3σ̂)] — centred capability — accounts for process mean location

σ̂ = R̄ / d₂ (for X̄-R chart)

Long-term performance (total variation including between-subgroup):

  • Pp = (USL − LSL) / (6s) — same formula but uses overall standard deviation s
  • Ppk = min[(USL − X̄̄) / (3s), (X̄̄ − LSL) / (3s)]

Acceptance criteria

Index Minimum Target
Cpk 1.33 1.67
Ppk 1.33 1.67
Cpk Interpretation PPAP action
≥ 1.67 Excellent ✅ Accepted
1.33 – 1.67 Acceptable ✅ Accepted — monitor
1.00 – 1.33 Marginal ⚠️ Customer approval required; add control measures
< 1.00 Not capable ❌ 100% inspection required; corrective action mandatory

Cp vs. Cpk relationship

  • Cp = Cpk: process is perfectly centred
  • Cp > Cpk: process is off-centre — improve centering before widening control limits
  • Never report only Cp without Cpk — a process can be off-centre and still show a good Cp

Step 5 — Respond to an out-of-control condition

  1. Stop the process (or place affected output on hold) — do not continue producing to an out-of-control process
  2. Contain — identify affected output since last in-control point
  3. Investigate — ask: what changed? (material lot, operator, shift, tooling, environment)
  4. Identify root cause — use 5-Why or Fishbone (is-is-not to scope the problem first)
  5. Correct — implement correction and verify the process returns to control
  6. Document — note the signal, investigation, and action taken on the chart (or in the log)
  7. Update PFMEA and Control Plan if the root cause reveals a new failure mode

Do NOT simply recalculate control limits after a shift to make the chart "look in control."


Step 6 — Audit an SPC implementation

When reviewing SPC in production or at a supplier:

  • [ ] Correct chart type selected for data type (variable vs. attribute)
  • [ ] Minimum 25 subgroups used to calculate initial control limits
  • [ ] Control limits calculated from process data — NOT set to specification limits
  • [ ] Western Electric rules applied (minimum Rule 1, 2, 3)
  • [ ] Out-of-control signals are annotated on the chart with the action taken
  • [ ] Process capability calculated on stable (in-control) process only
  • [ ] Cpk ≥ 1.33 minimum; 1.67 for special characteristics
  • [ ] MSA study complete and %GRR < 30% for the gauge being used
  • [ ] Control chart is used for decision-making — not filled in retroactively

Validation criteria

An SPC implementation is adequate when:

  • Chart type matches data type and subgroup size
  • Control limits calculated from minimum 25 subgroups of production data
  • Process in statistical control before capability is calculated
  • Cpk ≥ 1.33 (minimum) for all monitored characteristics
  • Out-of-control signals trigger documented investigation and corrective action

Common mistakes

  • Setting control limits equal to specification limits (a fundamental SPC error — these are different concepts)
  • Calculating capability on an out-of-control process (meaningless — must be stable first)
  • Reporting Cp but not Cpk — hides off-centre processes
  • Using only n=1 individual charts when subgrouping would reveal more
  • Filling in control charts retroactively at end of shift — defeats the purpose of real-time monitoring
  • Recalculating control limits to eliminate out-of-control points without identifying root cause
  • Reporting Cpk = 1.45 based on 30 parts — too few; minimum 100 pieces for reliable capability

Output Format

At the start of each use, ask the user:

"How would you like to receive the output? A — Structured Markdown (formatted tables and sections, ready to copy) B — Plain tables (simplified structure for Excel or Word) C — Narrative report (flowing text for a formal document or email)

Default: A."

Adapt all output sections to the chosen format. If the platform or session context already defines a format preference, skip this question.

Changelog

Version Date Author Change
1.0 2026-06-06 @RBraga01 Initial release
1.1 2026-06-06 @migmcc Added 100-part minimum requirement for valid capability study in Step 4; clarified in-control prerequisite before capability calculation

04 同仓库更多 Skills

hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-review-with-annotations

Create, inspect, reply to, update, or resolve situated Worktable annotations. Use when feedback belongs to a Doc section, block, text range, or HTML Doc, or when the user asks to address existing review feedback and coordinate a requested artifact edit.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-find-and-synthesize-context

Find, assess, and synthesize context from Worktable Docs and Records. Use when the user asks what a workspace contains, wants workspace-backed research, needs current facts reconciled across artifacts, or requests a cited synthesis without changing workspace state.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-create-or-update-docs

Create, edit, rename, or maintain narrative Worktable Docs. Use for plans, decisions, briefs, notes, research syntheses, and other durable prose when the user requests a new artifact or a safe update to an existing one.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-create-or-manage-records

Organize repeated structured information as durable Worktable Records. Use for research datasets, inventories, directories, pipelines, trackers, and other collections whose items need identity, fields, filtering, links, validation, or individual updates.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-create-or-manage-interactive-html

Create or improve durable visual and interactive artifacts in Worktable. Use for dashboards, trackers, calculators, mini apps, slide presentations, and other requests that benefit from a polished HTML Doc.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-collaborate-threads

Read, answer, and hand off messages through durable Worktable threads with users or other authenticated participants. Use when the user asks to check addressed messages, reply to a participant, continue a thread, request another agent's help, wait briefly for a response, or watch for comments and messages sent to the agent.

★ 1.6k