The work comes home. The know-how doesn't.

Here is an uncomfortable observation from the reshoring debate: everybody prices the robots, the building and the freight, and nobody prices the twenty years of judgment left standing at the old supplier's machines. Quality Magazine's latest reshoring survey tells the familiar story – momentum building despite policy and workforce headwinds, with the skills gap listed as one challenge among several. It is not one challenge among several. It is the project. I learned that the expensive way when I was hired to build a quality function for a greenfield plant of 900-plus people: new hall, new presses, new workforce, no history. The absence of veterans has a price. It sits on no capex line, and it arrives late – disguised as defect costs, escaped parts and a customer phone call nobody wants to take.

What actually stays offshore

When production moves home, the paperwork moves with it. Drawings, PPAP files, control plans, work instructions – everything transfers within a week and looks complete on a shared drive. What stays behind is the operator who hears a press go half a tone flat before the first bad part, and the die-setter who knows the tool "isn't happy" from the finish on a first article. The quality engineer whose escalation instinct took two decades of near-misses to calibrate stays behind too. You inherit documents. You do not inherit senses. I saw the mirror image at Witte Automotive: a mature plant where half the real control plan lived in the heads of long-serving setters. It worked beautifully until somebody retired.

Tribal knowledge never appears on a shipping manifest. That is why it never arrives.

The quiet failure mode sits inside the control plan itself. A control plan written by experts silently assumes experts will execute it. Every reaction plan, every frequency, every line that says "check appearance" presumes someone who already knows what a good part feels like. Run that plan with a three-month workforce and you have not transferred a quality system. You have transferred a test the new team will fail politely, at rate.

Build the judgment into the system, don't rent it

The instinct at any new site is to close the gap with training hours. Training matters, but it is a depreciating asset that walks out with every resignation. We chose to engineer around the gap instead – and through my consulting work at FOREAST I have since sat in enough established plants to know how much of their "system" is veteran instinct wearing a documentation costume. Four mechanisms carried us:

• Poka-yoke before training hours. Error-proof the station so inexperience cannot create the defect. A few hundred euros of fixture beats a hundred classroom hours, and it works the same on shift three as on shift one.
• Layered process audits. New eyes cannot see drift because they have no baseline. Layered audits – operator, supervisor, plant management – catch the slow departure from standard that a novice already perceives as normal.
• QRQC with roles, not seniority. We ran quick-response quality control with ownership, containment and verification assigned by role, because we had no twenty-year veterans to defer to. The structure stood in for seniority, and the structure stayed.
• Attribute agreement analysis, month one. A fresh inspector is a measurement system too, and an uncalibrated one. Our first attribute MSA showed new inspectors disagreeing with themselves and with each other – worth more than any gauge study on the CMM.

Then there was the andon. If stopping the line takes twenty years of confidence, the line stops after the defect escapes, not before. So we made pulling cheap and praised the puller in public. Every single time.

The first eighteen months are the qualification

Treat your own ramp as an extended PPAP. Run capability studies at 30, 90 and 180 days exactly as you would demand from any new supplier, because that is what you have become. We turned VDA 6.3 on our own site before any customer thought to – self-imposed, supplier-grade scrutiny, every finding tracked in an A3 with a named owner and a date. It is uncomfortable. It is also the cheapest audit you will ever commission: one escaped part in automotive body work costs more in sorting, containment and premium freight than a year of those audits.

Then stop promising the old plant's PPM on day one. That number was built on the twenty years of judgment that did not ship. Quote customers an honest ramp curve instead, and budget year-one defect cost as a planned, declining line rather than a surprise. At SNOP that approach took a 900-person greenfield operation to a 70% cut in defect costs and a quarter with zero critical customer escalations. QRQC and A3 did the daily work. The plan bought them the time to do it.

Reshoring is a quality decision before it is a financial one. The survey reads the workforce gap as a headwind. A quality director should read it as the substance of the project. The work will come home – the drawings already travel at the speed of email – and the plants that survive the trip are the ones that engineered the knowledge transfer deliberately instead of assuming it rode home with the tooling.