A3 Workflow · DMAIC
Overall progress 45% · Stage 5 of 11
Problem Overview — charter snapshot
High Priority
A3 Number
A3-2026-0418
Site
Sheffield Precision
Department
CNC Machining
Process Area
CNC Cell 3
Priority
High
Owner
MH M. Holloway
Sponsor
BG B. Griffiths
Created
12 May 2026
Due Date
30 Jun 2026
RAG Status
Amber
Background & Context
CNC Cell 3 machines safety-critical titanium brackets for the Tier-1 aerospace customer. Over Q1–Q2 2026 the first-pass reject rate climbed from 6.2% to 11.4%, well above the 4% contractual quality gate, coinciding with a 22% volume increase and a new alloy batch. Rework and scrap now total an annualised £129k and the cell is at risk of missing the customer's OTIF commitment.
First-Pass Reject Rate
11.4%
+5.2 pts vs target
OEE — Cell 3
62%
23 pts below target
Scrap + Rework
£129k
annualised run-rate
Lead Time
14 d
+4 d vs standard
Reject Rate Trend
Pareto of Defect Types
Defect Heat Map — by shift & machine
0–3%
4–8%
9%+
Pulled live from Data Point KPI · Cell 3 quality stream
Problem Statement
The first-pass reject rate on CNC Cell 3 has risen to 11.4% against a 4% target, driven primarily by dimensional out-of-tolerance and surface-finish defects. This is generating £129k of annualised scrap and rework and placing the aerospace customer's OTIF commitment at risk.
Safety Impact
Medium
Customer Impact
High
Business Impact
High
Cost Impact
£129k / yr
Drop photos here or browse
bracket-OOT.jpg
surface.jpg
CMM_inspection_report.pdf
Quality · 1.2 MB
scrap_log_Q2.xlsx
Data Point export · 340 KB
Target Condition — WDGLL
| Measure | UoM | Baseline | Target | By | Status |
|---|---|---|---|---|---|
| First-pass reject rate | % | 11.4 | 4.0 | 30 Jun 26 | Tracking |
| OEE — Cell 3 | % | 62 | 85 | 31 Jul 26 | On plan |
| Scrap & rework cost | £/yr | 129,000 | <30,000 | 30 Jun 26 | Tracking |
| Lead time | days | 14 | 9 | 30 Jun 26 | On plan |
5 Whys
1
Why are parts out of tolerance?
Tool wear drifts the cut dimension before it is detected.
2
Why isn't the drift detected?
No in-process measurement — parts are gauged only at batch end.
3
Why only at batch end?
SPC has never been implemented on the cell.
4
Why no SPC?
Operators aren't trained on SPC and there is no gauge at the machine.
5
Why no training or gauging?
Standard work and a tool-change schedule were never defined for the cell.
Verified root cause
Fishbone (6M)
Evidence Capture
| Cause | Category | Evidence | Confidence |
|---|---|---|---|
| No tool-change schedule | Method | Tool-life logs show 0 scheduled changes vs 6 expected | Confirmed |
| No in-process measurement | Measurement | CMM data only post-batch; drift undetected for 40+ parts | Confirmed |
| New alloy batch variation | Material | Defect rate +1.8% on new lot — minor contributor | Partial |
Countermeasure Plan
Implement in-process SPC + at-machine gauging
In ProgressExpected benefit
−5.0%
Actual to date
−3.2%
Define tool-change schedule by tool life
CompleteExpected benefit
−3.0%
Actual
−2.8%
Operator standard work + SPC training
In ProgressExpected benefit
−2.0%
Actual
—
Coolant concentration control loop
To DoExpected benefit
−1.0%
Actual
—
Action Plan — drag cards between lanes
Results Validation — before vs after
Verified by QA
First-pass reject rate
11.4%→5.1%
−6.3 pts
OEE — Cell 3
62%→74%
+12 pts
Scrap & rework
£129k→£58k
−£71k / yr
Lead time
14 d→9 d
−5 days
Quality (FPY)
88.6%→94.9%
+6.3 pts
Reject rate — before vs after
Cumulative savings realised
Standardisation
Embed the improvement into standard work, control plans and leader standard work. Confirm sustainment before closure.
Standard work updated & signed off
Control plan / SPC limits set
Tool-change schedule added to PM system
Roll out to CNC Cells 1, 2 & 4
Lessons Learned
Daily SPC review at the cell gave operators ownership of quality and caught drift in minutes rather than at batch end. Pairing tool-life scheduling with at-machine gauging delivered most of the gain.
Standardise the in-process SPC + tool-life pattern as default for all precision machining cells. Add to the Data Point cell-launch checklist and the Skills Matrix training path.
Closure & Approval
Final sign-off by owner, sponsor and QA. On approval the A3 is archived and surfaced to Brett Bot for similar future problems.
Owner
M. Holloway
SignedQA
Subhalekshmi
SignedSponsor
B. Griffiths
PendingExecutive Summary
Open A3s
38
Overdue A3s
6
Completed (YTD)
124
Avg closure time
47 days
Savings generated (YTD)
£2.46M
Highest impact A3
A3-2026-0391 · £312k
Site Performance
1Birmingham94
2Sheffield81
3Bristol76
4Glasgow68
Brett Bot AI
Suggest root causes
Generate 5 Whys
Recommend countermeasures
Find similar historical A3s
Predict success probability
Auto-generate exec summary
