Knowledge continuity in manufacturing: keeping the plant's memory
Manufacturing SMEs run on two operating manuals: the documented one, and the real one in the heads of the plant manager and a few senior operators. The real one holds the process tuning, the exception recovery moves and the machine quirks — knowledge built from years on the same floor. It is invisible while those people are present and irreplaceable the day they leave.
What knowledge is at risk in manufacturing
The knowledge at risk clusters around the physical plant and its history:
- Process tuning knowledge — why the settings deviate from the manual's defaults, which drift is seasonal and which announces a failure — held by the plant manager and senior operators, written nowhere.
- The exception playbook — the recovery moves for every recurring disturbance, learned incident by incident over decades — the knowledge behind every quick restart.
- Machine-specific quirks — what unit 7 needs that its twin doesn't; the vintage problems per production line — keyed to serial numbers in one or two memories.
- Quality judgment — which deviations matter and which don't; what the customer will actually accept versus what the spec says — the difference between scrap costs and pragmatism.
- Supplier and maintenance calibrations — whose 'in stock' is real, which maintenance can slip — experience-priced promises.
The typical trigger moment
The classic manufacturing trigger is the retirement wave: the generation that installed and tuned today's lines is reaching pension age together, and each plant tends to lose its process memory in clusters rather than one person at a time. The second trigger is the night shift incident after a departure — the first disturbance that nobody present has seen before, resolved at premium cost, revealing how much playbook just left.
What a Knowledge Space looks like here
MentX is a personality-aware knowledge base: it captures the knowledge of a company's key person — from meetings and a guided intake track — into a living, temporal knowledge graph with full source attribution. A manufacturing Knowledge Space connects the daily production meetings and shift handovers — where the exception knowledge is exercised — and runs the intake track through the incident history: the last ten disturbances per line, the alarm noise list, the quirk catalog per unit, the quality boundaries per customer. The graph keys knowledge to machines and lines, so answers surface where faults do. From there, DigiME answers the 3 a.m. questions the departed expert used to take by phone.
Questions we hear from manufacturing leaders
Our senior operators say everything important is 'in the system'. Is it?
The MES and ERP hold parameters and history — the what. The why (why this setting, why this deviation is tolerated, why this alarm is ignored) lives in people. A simple test: pick one non-default setting and ask the system why it is what it is. That gap is what a knowledge base fills.
We run on SOPs and ISO documentation. What would a knowledge base add?
SOPs cover the procedures for foreseen situations. The expensive knowledge is what your best people do when the SOP doesn't apply — the exceptions, the diagnostics, the judgment. A knowledge base captures that layer with sources, and keeps it current as the plant evolves.
Can this work for a plant with three shifts and high time pressure?
That is the environment it is designed for: capture rides on meetings that already happen (production meetings, shift handovers), and the intake sessions are scheduled around the expert. Consulting the base takes less time than finding the one person who knows.
