Every shop has at least one process that only one person really knows how to run. Maybe it’s a legacy CNC program with quirks nobody documented, or a calibration routine on an older piece of gauge equipment, or simply the one inspector who has developed an eye for a defect that the written acceptance criteria don’t fully capture. It works fine for years, right up until that person takes vacation during a critical order, gets promoted, or leaves the company entirely, and the shop discovers how much of its process knowledge was never actually in the quality system at all.

Why This Risk Hides So Well

Single-point knowledge is dangerous precisely because it doesn’t show up as a problem until it’s suddenly a crisis. Day to day, everything runs smoothly because the one person who knows the process is present and available. There’s no nonconformance to trace, no scrap trend to flag, nothing in the metrics that would prompt a risk assessment to look twice. The gap is invisible in the record because the record only shows outputs, not who is capable of producing them. It typically surfaces during a resignation, an extended illness, or an unplanned absence right before a shipment deadline, which is the worst possible moment to discover a process gap.

A version of this shows up almost identically across shops: a senior programmer who’s run the same five-axis cell for a decade retires with two months’ notice after a family circumstance changes his plans. The handoff period looks fine on paper — he trains his replacement on the documented steps, and the new person can run the program by following the work instruction. What doesn’t transfer in two months is the accumulated judgment: which fixture clamps tend to need an extra check on humid days, which tool offset drifts faster than the schedule assumes, which visual cues on the finished part indicate a problem before it shows up on a CMM report. None of that was ever written down because it never had to be, right up until the person who carried it in his head was gone.

Where It Connects to Clause 6.1

Risk-based thinking under clause 6.1 is usually applied to process risks — tooling failure, supplier quality, equipment downtime — but workforce concentration is just as real a risk to product conformity and on-time delivery, and it deserves a line in the same risk register. The assessment doesn’t need to be complicated: for each critical process or inspection point, how many people are currently qualified to perform it, and what happens to output if the primary person is unavailable for two weeks. Processes with only one qualified person are, by definition, a single point of failure, and ranking them by how critical the process is to product conformity gives a shop a rational order in which to close the gaps rather than reacting to whichever gap happens to bite first.

Building Cross-Training Without Slowing the Floor Down

The instinct once a gap is identified is to schedule a broad cross-training initiative, which often collapses under its own ambition because it competes with production schedules for everyone’s time. A narrower approach tends to work better: identify the two or three highest-risk single-points, pair each with a secondary person who already has adjacent skills, and structure the cross-training around actual production runs rather than a separate training session. An operator learning a second process by working it under supervision during a normal shift produces both output and competence at the same time, instead of pulling capacity off the floor for a dedicated training block.

The competency record from that cross-training matters just as much as the training itself. A shop needs to be able to answer, on short notice, exactly who is currently qualified for each critical process — not who was qualified two years ago, and not who theoretically could pick it back up with a refresher. That answer needs to be retrievable in minutes, not reconstructed from memory when someone unexpectedly calls in sick the same week a key order ships.

The Cost When Cross-Training Gets Skipped Entirely

The consequence of leaving a single-point risk unaddressed rarely shows up as a dramatic failure on day one. It shows up as a slow erosion of margin: a rush order gets pushed to a less experienced operator because the qualified person is out, quality holds increase for a few weeks, a customer complaint traces back to a process someone was technically running for the first time under real production pressure rather than supervised training conditions. By the time it’s visible in the numbers, the shop is often already contractually committed to a delivery schedule that assumed the missing expertise would be there. Treating this as a risk worth budgeting time against before it happens is considerably cheaper than absorbing the cost of discovering it mid-order, when there’s no good option left except slowing everything down or accepting a quality risk nobody would choose deliberately.

Making the Record Actually Usable

This is where a lot of well-intentioned cross-training efforts quietly fail: the training happens, but the record of who is now qualified for what lives in a format nobody actually checks before assigning work. A qualification matrix in a binder or a spreadsheet that only the quality manager updates doesn’t help a shift supervisor scrambling to cover an absence at seven in the morning. The record needs to be something a supervisor can actually pull up and trust in the moment they need it, which is a much lower bar in principle than it turns out to be in practice for shops running training records on paper or in disconnected files. Shops that have moved qualification tracking into manufacturing QMS software that operators will actually use tend to close this gap because the qualification matrix stays current automatically as training events are logged, rather than depending on someone remembering to update a separate document.

Single-point knowledge will never be eliminated entirely — some expertise takes years to build and can’t be fully transferred through a training checklist. But the risk of it going unmanaged can be, and that starts with treating workforce concentration as a documented risk like any other, rather than something that only gets attention after it has already caused a problem.