IDB-RMP-021
Production · ramp · pilot · MP
Production ramp-up plan
Reference for transitioning from pilot to mass production — yield ramp curves, gate criteria, operator training, capacity planning, and the milestones that signal readiness for full-volume production.
Abstract
Production ramp-up is the planned transition from low-volume pilot (50–500 units) to mass production (10 000+ units). Done well, yield ramps from 70 % to 95 % over 4–8 weeks, defect rates drop, and unit cost stabilises. Done poorly, yield stays low, schedule slips, and the first 3–6 months of production lose money.
Section 1 covers pilot-to-MP gate criteria. Section 2 covers yield ramp targets. Section 3 covers operator training and line balancing. Section 4 covers capacity planning. Section 5 covers production process documentation. Section 6 covers risk management during ramp.
1.Pilot-to-MP gate criteria
Before scaling production from 100 to 10 000 units/week, the project must meet specific gate criteria. Below these thresholds, the line is not ready for MP.
1.1Gate criteria checklist
- [ ] First-pass yield ≥ 90 % in pilot run.
- [ ] All critical dimensions at Cp/Cpk ≥ 1.33 on production samples.
- [ ] All known defects catalogued in QC plan with severity classification.
- [ ] Standard Operating Procedures (SOPs) written for every station.
- [ ] Operators trained and qualified for their assigned stations.
- [ ] Production test fixtures validatedsame fixture catches same defects across batches.
- [ ] BoM and Spec sheet revisions locked at production version.
- [ ] All compliance certificates issued (CE, FCC, etc.).
- [ ] Material allocation securedCritical components have safety stock + lead-time visibility.
- [ ] Logistics + customs plan confirmed.
- [ ] Rework path defined for each defect type.
- [ ] End-of-line audit process in place (FQC).
- [ ] First-pass-yield (FPY) target agreed with supplier as contract metric.
1.2Pilot production batch sizes
| Stage | Quantity | Purpose |
|---|---|---|
| Engineering Sample (ES) | 5–10 | Process feasibility check |
| Pre-Production Sample (PPS) | 20–50 | Full process verification |
| Pre-Production Run (PPR) | 100–500 | Yield + cycle time baseline |
| Pilot production | 500–2 000 | Operator training, fixture validation |
| Ramp 1 (initial MP) | 2 000–5 000 | First commercial batch |
| Ramp 2 (steady-state MP) | 5 000+ per week | Full production volume |
1.3Risk-based gate exceptions
Some products may pass MP gate with first-pass yield <90 %, if:
- Defects are recoverable at rework station with known fix time.
- Yield improvement plan in writingSpecific actions + timeline to reach 90 %.
- Buyer accepts the risk financiallyHigher unit cost during ramp.
This must be a deliberate choice, not a default.
2.Yield ramp curve
First-pass yield improves predictably as the line stabilises. Track and forecast it explicitly.
2.1Typical yield ramp (consumer hardware)
| Production batch | Cumulative units | First-pass yield | Defect rate |
|---|---|---|---|
| 1 (pilot) | 0–500 | 65–75 % | 25–35 % |
| 2–3 (early ramp) | 500–5k | 75–85 % | 15–25 % |
| 4–8 (mid ramp) | 5k–20k | 85–92 % | 8–15 % |
| 9–15 (late ramp) | 20k–50k | 92–96 % | 4–8 % |
| 16+ (steady state) | 50k+ | 95–98 % | 2–5 % |
2.2Yield improvement levers
| Lever | Typical gain | Time to apply |
|---|---|---|
| Operator skill (per shift) | +2–5 % | 1 week |
| Fixture refinement | +3–7 % | 2–3 weeks |
| Process parameter tuning | +2–4 % | 1–2 weeks |
| Component variability reduction | +1–3 % | 2–4 weeks |
| Tooling refinement | +2–8 % | 4–8 weeks |
| Design change (ECN) | +5–15 % | 6–12 weeks |
| Material substitution | +1–5 % | 2–6 weeks |
2.3Common yield killers
| Defect | Cause | Fix |
|---|---|---|
| Solder bridges (PCBA) | Stencil aperture too large | Reduce paste volume; adjust aperture |
| Tombstoning | Reflow profile asymmetry | Adjust profile; balance pad design |
| Component misalignment | Pick-and-place feeder issues | Adjust feeder positioning; clean nozzle |
| Surface defects (plastic) | Mold contamination; wear | Clean mold; refurbish |
| Mis-aligned screws | Torque setting; tool wear | Re-calibrate; replace driver |
| Connector damage | Operator force; tool fit | Adjust insertion path; train operator |
| Foreign material in pack | Loose particles in line | Improve cleanliness protocol |
3.Operator training + line balance
The production line is human + machine. Training and balance determine throughput.
3.1Operator skill curve
First-week operators perform at 40–60 % of experienced operator productivity. Productivity reaches 90 %+ at 2–3 weeks. Don't ramp on first-week operators.
3.2Training stages
| Stage | Duration | Activities |
|---|---|---|
| Familiarisation | 1 day | Walk through line; understand the product |
| Single-station training | 2–3 days | Master one station |
| Quality awareness | 1 day | Defect identification, SOP compliance |
| Cross-station rotation | 3–5 days | Learn 2–3 adjacent stations |
| Certification | 1 day | Demonstrate proficiency to QA lead |
| Continuous improvement | Ongoing | Kaizen suggestions, defect feedback |
3.3Line balancing
Each station's cycle time should be close to the line's bottleneck. If station A takes 60 s and station B takes 30 s, B is starved 50 % of the time.
| Cycle time issue | Symptom | Fix |
|---|---|---|
| Bottleneck at one station | Other stations idle | Add operator, automate, simplify |
| Cycle time imbalance | Throughput < capacity | Rebalance, redistribute tasks |
| WIP buildup | Inventory between stations | Pull system, takt-time discipline |
| Operator fatigue | Late-shift quality drop | Job rotation, breaks, shorter shifts |
3.4Cycle time + takt time
- Cycle timeTime for one operator to complete their station task.
- Takt timeCustomer demand interval (working hours / units required).
- GoalEach station's cycle time ≤ takt time.
Example: 50 units/hour demand → takt time = 72 s. Each station must complete in ≤ 72 s.
3.5Standard Operating Procedure (SOP)
One SOP per station. Includes:
- Task descriptionWhat to do, in plain language.
- Tools usedList of tools per task.
- Visual aidsPhotos at each step.
- Quality checkWhat to verify before moving to next station.
- Quality issuesCommon defects, how to recognise.
- Cycle time targetExpected duration per task.
- Sign-off blockOperator and QA lead initial when complete.
4.Capacity planning
How fast can the line scale? What's the volume ramp from week 1 to week 12?
4.1Line capacity formula
``` Daily capacity = Working hours × shifts × stations × units/station-hour ÷ (1 + reject rate) × OEE (Overall Equipment Effectiveness, typically 0.75-0.90)
Weekly capacity = Daily capacity × working days (typically 5-6) ```
Example: 1 line, 1 shift × 8 hours, 10 stations × 30 units/hour, 5 % reject, 0.85 OEE:
- Hourly: 300 units / 1.05 × 0.85 = 243 units
- Daily: 1 944 units
- Weekly: 9 720 units
4.2Capacity ramp
| Week | Hours/day | Shifts | Lines | Estimated weekly capacity |
|---|---|---|---|---|
| 1 (pilot) | 4 | 1 | 1 | 2 000 |
| 2 | 6 | 1 | 1 | 4 000 |
| 3-4 | 8 | 1 | 1 | 8 000 |
| 5-8 | 8 | 1 | 2 | 16 000 |
| 9-12 (steady) | 8 | 2 | 2 | 30 000 |
Plan ramp on the supplier side. Capacity is often the bottleneck during early production.
4.3Component supply during ramp
Material supply must lead production by safety stock buffer + lead time:
- Week 1 productionMaterials ordered 8–12 weeks earlier.
- Re-order pointMaterials triggered when stock falls below (weekly demand × supplier lead time × 1.5 safety factor).
- Critical componentsHigher safety stock (4–8 weeks) for single-source items.
5.Production process documentation
The line generates documentation; the documentation supports the line.
5.1Documents on the production floor
- SOPs per stationVisible at workstation; updated when process changes.
- Quality planAQL, inspection points, defect catalogue with photos.
- Bill of materialsPosted at each station for verification.
- Defect logPer-shift, per-station defect counts.
- Visual standardsBoards showing good/bad examples at cosmetic stations.
- First-article inspection reportOne unit per shift start, verified against spec.
5.2Trend monitoring
- Daily yield reportsFirst-pass yield + reject rate + defect breakdown.
- Weekly trendYield trajectory, recurring defects, process drift signals.
- Cp/Cpk trackingCritical dimensions trended over time.
- Operator performancePer-operator productivity + defect attribution.
5.3Process change control
| Change type | Owner | Approval |
|---|---|---|
| Tooling adjustment (within spec) | Production supervisor | Floor approval |
| Process parameter change | Process engineer | QA + Engineering |
| Material substitution | Engineering | ECN + sign-off |
| Design change | Engineering | Full ECN process |
| SOP revision | Production supervisor | QA review |
6.Risk management during ramp
The first 4–8 weeks of production carry the highest risk. Plan for surprises.
6.1Common ramp risks
Supply-side
- Component shortage (allocation)
- Quality drift in materials
- Tooling failure or wear
- Supplier scheduling conflicts
- Currency / cost fluctuation
Production-side
- Operator skill gap
- Fixture wear or failure
- Process parameter drift
- Defect identification gaps
- Capacity bottleneck unforeseen
6.2Risk mitigation tactics
| Risk | Mitigation |
|---|---|
| Component shortage | Safety stock 4–8 weeks, dual-source critical parts |
| Tooling failure | Spare tooling on standby; refurbishment plan |
| Operator turnover | Cross-training; documented SOPs |
| Quality drift | Daily Cp/Cpk monitoring; intervention triggers |
| Capacity shortfall | Second shift availability; outsourcing backup |
| Customs / shipping delay | Buffer stock at destination; alternative ports |
6.3Daily standup during ramp
A short morning meeting catches problems before they propagate:
- Yesterday's actual production vs. target
- Defect breakdown by station and type
- Open issues from production floor
- Risk indicators (material levels, fixture status, operator gaps)
- Today's plan + adjustments
6.4Escalation criteria
Issues that warrant immediate escalation:
- First-pass yield drops below 75 %
- Single defect type accounts for >50 % of defects
- Critical component shortage (less than 2 weeks of supply)
- Tooling damage requiring repair
- Compliance-relevant defect (mis-marked, mis-labelled)