Why maintenance knowledge gets lost at the machine

A technician spends an hour on a fault that nobody has seen before. They try three things, and the third one works. They now know something useful: what the symptom looks like, what the cause was, and what fixed it. The next person who sees the same fault will need exactly that. In most teams, they will not get it.

Where the knowledge goes

Some of it goes nowhere. The machine runs again, the next job is waiting, and the work order is closed with "fixed".

Some of it goes to places nobody looks:

• a note in a personal notebook or phone,
• a photo in someone's camera roll,
• a message in a group chat, between a hundred other messages,
• a comment written in marker on the machine's housing,
• a free-text field in a system that is hard to search.

And much of it stays in people's heads. The experienced technician knows that the press on Line 3 jams when the material is cold, and that the manual has the wrong torque value for one bolt. That knowledge leaves when they change shifts, change jobs, or retire.

Why it is hard to capture

The moment of learning is the worst moment to write. The technician is at the machine, often with dirty or gloved hands, in a noisy place, under time pressure. The tool for recording is usually far away: a terminal in the workshop, a laptop in the van, a form that asks for twenty fields.

By the time they reach it, the details are gone. Which hose was it? Was the reading 4.2 or 2.4 bar? The note that gets written is shorter and vaguer than what they knew.

There is also little reward for writing. The technician who records a fix well rarely sees the benefit themselves. The benefit goes to someone else, months later.

Why it is hard to find

Even when knowledge is recorded, the next person has to find it. That means knowing it exists, knowing where to look, and searching with the right words. A technician standing at an unfamiliar machine may not know its asset number, which system its history lives in, or what the last person called the fault.

Manuals have the same problem. The right document exists, but it is a 400-page PDF in a shared drive, under a folder named after a project from years ago.

What it costs

Lost knowledge shows up in familiar places:

• longer repairs, because each technician diagnoses from scratch, which raises MTTR;
• repeat visits in field service, which lowers the first-time fix rate;
• repeat failures, because nobody connected this one to the last one, which makes root cause analysis impossible;
• new technicians who take a long time to become independent, because the knowledge they need is not written anywhere.

What helps

None of the fixes is a single tool. They are habits that make capture and retrieval cheaper.

1. Put the history at the machine. A label on the machine that opens its record, such as a QR code, removes the search. The technician does not need to know the asset number or the system. See Label machines with QR codes.
2. Make recording fast. Accept photos, short voice notes, and rough text at the machine. Structure can be added later, by the technician or a planner. A rough note recorded now beats a clean one never written.
3. Keep the original. When a note is cleaned up or summarized, keep the original and who wrote it. The next person can judge it better if they can see where it came from.
4. Read the notes. Supervisors who read closing notes, and ask about vague ones, get better notes. Planners who use the history to plan get better history.
5. Make the benefit visible. When a technician finds a fix because a colleague recorded it, say so. It is the best argument for recording.

The goal is simple to state: what one person learns at a machine should be in front of the next person who stands there. Most teams are far from it, not for lack of systems, but because recording and looking up do not fit the moment of work. That is the gap to close.