The Workaround is the Spec
Before you capture what a retiring operator knows, ask why they know it.
It is 6:40 on a Tuesday and second shift left a note. Don’t trust the counts on the screen. Check them against the whiteboard by the press. Nobody wrote down why. The person who would know retires in November.
The advice that follows is old and not wrong: get the knowledge out of people's heads and into systems. Manufacturers have heard it for 30 years. Plants still run on memory. That gap is worth examining.
Start with the workaround itself. It is not a discipline problem. It is information. It marks the exact point where what the system says stopped matching what transpires on the floor. Someone noticed and built a bridge out of a whiteboard and their own attention. What nobody recorded is the gap that made the bridge necessary.
So, when a plant sets out to capture what a soon-to-be-retired operator knows, it transcribes the bridge. The workaround becomes official. The mismatch remains.
That also explains the decay. Knowledge about a mismatch has a short shelf life because mismatches move. Product mix changes, a machine gets added, a customer adds a requirement. The document goes stale and everyone goes back to asking the person.
It helps to sort tribal knowledge into two kinds because they need opposite responses.
The first is the workaround — how people get around the system. This has no long-term value. It should be designed out, not written down.
The second is judgment — why the process behaves the way it does. Why that resin runs differently in August. What the line sounds like a week before a bearing goes. That is real expertise, and the expert often cannot fully explain it, which is why binders never captured it.
Technology has finally made both responses practical.
Workarounds used to be rational. Changing the ERP was slow and expensive, so people compensated in spreadsheets and memory. Integration, machine connectivity and low-code tools have changed that math. Making the system fit the process is now a decision, not a capital project.
Judgment no longer has to be interviewed out of anyone. It can be measured. Put a sensor on the vibration the senior operator hears. Match it against the failures they called. The pattern becomes a signal the system holds, and the expert validates it rather than authoring it.
The plant's downtime and defect history is already written down, just not anywhere searchable: maintenance work orders, shift logs, quality holds. AI reads that kind of record well and shows where the same problem keeps coming back. That is a workaround inventory, and most plants have never run one.
None of this replaces mentoring or leadership. It changes the question. Instead of asking what the retiring operator knows, ask what they have been compensating for and how long the plant has been paying for it.
Workarounds are the cheapest requirements document in the building. Every one of them has an expiration date attached to a person.
About the Author
Mike Schipper is Founder & CEO of InsITe. He may be reached at 616-421-8324 or mike.schipper@trustedinsite.com .
InsITe was a sponsor of the 2026 MMA Workforce Solutions Summit. Learn more about MMA events.