Ramp Mistakes

Where Production Ramp and Launch Readiness Numbers Go Wrong

A troubleshooting guide to the ramp, yield, staffing, and readiness errors that quietly wreck a launch, each with its symptom, root cause, and a numbered fix.

The most common ramp error is a wrong denominator. A team reads a Production Ramp Rate of 40% and celebrates, but they divided passed units by units already run instead of the full scheduled window. Symptom: the rate looks healthy early yet output never reaches the committed date. Root cause: using a shrinking running total instead of the fixed 250-unit window. Fix: lock the denominator to the whole ramp plan. Eight passed units against 250 scheduled is 3.2%, not the 80% you get dividing by 10 units run. The honest 3.2% forces the escalation the inflated number hides.

Counting reworked units as good is the yield killer. A pilot shows 88% first-pass yield on the Yield Ramp Curve, then mass production drops to 71% and nobody can explain it. Root cause: the pilot counted 22 repaired units as good, masking a fixture defect. Fix: first-pass yield means passed with zero rework or repair. If 8 of 250 pass clean, that is 3.2%, even if 90 more limped through after touch-up. Rework hides the true process capability you are trying to prove, and it does not scale, so the defect resurfaces the moment volume climbs and headcount for repair runs out.

Ignoring the learning curve breaks staffing plans. Symptom: crews sized off Staffing Ramp Workload run 30% over the calculated hours in weeks one and two. Root cause: the calculator assumes a steady output rate, but operators at ramp start hit maybe 60% of spec rate. Fix: feed the rate you actually observe, not the ideal spec. At 120 units and a real 8 units per minute instead of the spec 12, base time is 15 hours, not 10. Or hold the rate and raise the allowance from 10% to 25% to cover the early disruption. Re-tune each build as repetitions accumulate.

Using the final steady-state yield target at an early gate wastes escalation on healthy builds. Symptom: an EVT build reading 55% first-pass yield triggers a crisis meeting against a 95% target. Root cause: the 95% belongs to the mass-production gate, not EVT. Fix: stage the targets. A realistic curve might be 50% at EVT, 75% at DVT, 88% at PVT, then 95% at full rate. Judge each checkpoint against its own stage number in the Yield Ramp Curve, and watch the slope. A build climbing on plan needs no fire drill; a build stuck flat two gates running does.

Scoring detection backwards inverts your whole risk list. On Launch Readiness Score or Ramp Capacity Gap, teams sometimes give an easy-to-catch defect a high detection number, so trivial risks float to the top while a hard-to-detect sealing failure sinks. Symptom: the mitigation plan attacks visible problems and misses the dangerous invisible one. Root cause: reversed scale. Fix: anchor detection so harder to detect scores higher. With severity 6, occurrence 4, detection 3 you get a composite around 4.55; flip detection to 8 for a truly hidden failure and it jumps to roughly 7.4, correctly ranking it above the obvious defects.

Treating a risk score as a capacity number strands the line. Symptom: a Ramp Capacity Gap score of 4.55 gets logged, but nobody knows the cell is 6 units per hour short at full rate. Root cause: the FMEA-style score ranks danger, it does not size the shortfall. Fix: pair every high-scored constraint with a real takt or line-balance calculation. If takt is 45 seconds and the bottleneck cycles at 58, you are 22% under rate regardless of the composite. The score tells you which bottleneck to model first; only the capacity math tells you how many machines, shifts, or people close the gap.

Skipping supplier readiness assumes parts keep pace with the line, and they rarely do. Symptom: internal yield hits target but the ramp still stalls on shortages. Root cause: Supplier Ramp Readiness was never scored, so a Tier 2 running 3-week lead times against a weekly step-up never surfaced. Fix: rate each supplier's rate capability, PPAP status, and buffer before the ramp, not after the miss. If a supplier can deliver 400 per week and week 6 needs 700, the 300-unit gap is visible months out. Also right-size the Launch Inventory Buffer so a two-week supplier slip does not become a line-down event.

The most expensive mistake is not costing the delay, so slips look free. Symptom: a program trades a 2-week schedule slip for convenience with no pushback. Root cause: nobody ran Launch Delay Cost, so the lost margin stayed invisible. Fix: quantify it. At 5,000 units per week and $60 margin each, a 2-week slip forfeits $600,000 before any expedite or overtime spend. Put that number next to the Line Qualification Workload hours you would need to hold the date, and the trade-off becomes an actual decision instead of a shrug. A slip is never free; it is only unmeasured.

Published 2026-07-01.