Every rate increase is a process change nobody validated

Boeing is pushing 737 MAX production back up, and the industry is counting monthly deliveries like a scoreboard. Twenty years in aerospace and automotive quality teach you something the earnings calls skip: rate increases are where mature systems quietly come apart. Not because the system is bad — because it was validated for a world that no longer exists.

Every PFMEA scoring, every control plan limit, every operator standard work instruction, every capability study on file — they were established at a specific throughput. Change the rate and you are running an unvalidated process at industrial scale. The paperwork says you're compliant. The physics says you're experimenting.

Rate changes the physics of your process

This is not a metaphor. Throughput changes the physical variables your control plan depends on.

Thermal processes — curing, welding, heat treatment — have dwell times and ramp profiles calibrated for a specific part presentation rate. Double the throughput and your furnace load changes. The thermal mass per cycle shifts. A cure profile validated at 4 parts per hour behaves differently at 8, not because the oven is broken but because the system boundary moved.

Adhesive bonders, sealant application, composite layup — each has open-time windows defined in minutes. Increase the rate and you compress the gap between operations. A sealant that had 90 minutes of working time now has 45 because the downstream cell is cycling faster. Nobody updated the PFMEA because nobody flagged rate as a failure mode. Rate is the failure mode.

At Airbus, when we pushed throughput on the line I manage, we didn't just go faster. We rebuilt the routing verification framework from the ground up because the existing KPI structure was measuring a process that no longer existed in the same form. That rework delivered a 97% reduction in internal lead time — but the point isn't the number. The point is that the old measurement system was blind to the new dynamics. We had to re-establish what "in control" meant before we could claim it.

The quality debts that only surface at higher throughput

You can run a process at low rate with a surprising amount of hidden dysfunction. Operators have time to compensate. Inspection catch-up is possible because the queue isn't pressing. A marginal PFMEA severity rating doesn't bite because the occurrence frequency stays low at low volume.

Rate increase is when those debts come due.

The supplier quality problem that was 1-in-500 at 20 units a day becomes 1-in-250 at 40 — and suddenly it's on the weekly escalation list. The gauge that drifted every 200 parts now drifts within a single shift. The operator certification gap absorbed by a seasoned team lead becomes critical when you add a second shift of newer personnel.

When I built the QA/QC department at SNOP from scratch — 900+ employees, greenfield plant — I had the rare advantage of designing every process parameter for a defined production load. No legacy assumptions. No inherited control limits. That clarity cuts both ways: it showed me exactly how each parameter behaves under different loads, and how quickly a system validated at one throughput degrades at another. The same PFMEA that scored a risk as acceptable at design rate could justify a completely different severity and occurrence rating once you double the line speed.

Rate doesn't test your quality system. It invalidates it and waits to see how long before anyone notices.

What proper rate validation actually requires

A rate increase validated by a management decision and a project plan is not validation. It is a schedule. Real rate validation demands the same rigour as a new product introduction — because mechanically, it is one.

At minimum, it requires re-execution of the PFMEA. Not a review — a re-scoring. Occurrence ratings tied to throughput must be recalculated. Detection ratings for inspection steps facing higher volume must account for reduced time per unit. Severity may not change, but the risk priority numbers will, and that should drive action items before the ramp, not after the first escape.

Capability studies must be re-run at the new rate. Cp and Cpk values from the old throughput are historical documents. You need new short-run studies at production-intended cycle times before committing to customer deliveries. I've seen lines with Cpk above 1.67 at validation rate drop below 1.0 the moment tempo changed, because the variation sources were rate-dependent all along.

Control plans need to be walked at the new rate. Does the operator have time to execute the check? Is the sample frequency still statistically meaningful? Are reaction plans still actionable within the compressed window? Operator standard work must be re-time-studied, because standard work assumes a pace — change the pace and it becomes fiction until you re-time it. This is where most ramps fail quietly. People approximate, improvise, and variation enters through the gap between documented work and actual work.

The 50% reduction in EASA audit findings we achieved in one cycle wasn't because we discovered new problems. It was because we re-anchored our process controls to the actual production conditions and stopped auditing against an obsolete baseline. The quality system became honest about what it was controlling.

The next time a production ramp is announced — whether it's a 737 MAX rate increase, a defence supplier scaling for new programmes, or an automotive plant chasing rebound demand — don't ask how many units they'll deliver. Ask what was re-validated. If the quality system wasn't part of the ramp plan, the rate increase isn't growth. It's a loan against future recalls, and the interest rate is whatever the first field escape costs you.