Autonomous Maintenance / TPM
Autonomous maintenance makes operators the equipment’s first line of defence: daily cleaning-as-inspection, lubrication and tightening to standard — so wear and leaks are spotted weeks before they become breakdowns.
A 5S level is established in the area (fixed places, a cleaning standard exists)
Maintenance is on board as partners — they should train, not surrender territory under protest
Time for the daily routines is recognised in the plan (typically 10-15 min/shift) — not “when there’s time”
Operators and technicians clean the plant together — thoroughly. Everything abnormal is tagged: leaks, loose parts, worn cables, missing screws, defective gauges. A typical deep clean finds 50-200 deviations on a plant that “ran fine”.
The tag list is worked through: leaks are repaired at the source (not with drip trays underneath), guards that hinder inspection are rebuilt with windows/quick releases, lubrication points are made accessible. The purpose: keeping it clean should take minutes, not hours.
What, where, how, how often, how long — visual, with photos and marked points on the machine (lubrication points numbered and colour-coded). The standard must be followable within the agreed minutes per shift; if it can’t, the standard is wrong, not the operator.
The technicians teach the operators what to look, listen and feel for: bearing noise, vibration, temperature, play, hose fatigue. One component group at a time. The goal is that the operator spots deterioration long before the failure — that is the core of the whole method.
The operator’s observations must have a clear channel: small things are fixed by them (within an agreed limit), the rest is reported with a tag and handled visibly by maintenance within an agreed time. Findings that vanish into a black hole stop the stream of findings within a month.
A monthly audit of the standards and the tag flow on the pilot plant; breakdown data is tracked against the baseline. When the routine carries itself, expand to the next plant — with the pilot’s operators as trainers.
How it goes wrong in practice — and this is where most implementations part ways.
If the team finds 150 deviations, and 140 are untouched after two months, the message is received: this isn’t meant seriously. The tag list must be prioritised and worked off visibly — it IS the project in the first months.
10 minutes of inspection is dropped to hit output — and the price is paid as breakdowns with interest. Management decides this fight with its reaction the first five times it happens.
Without operator ownership it just becomes more preventive maintenance — good, but a fraction of the potential. The daily, close observation can only the operator deliver.
The method only works when it becomes daily routines and leadership behaviour. The operations check shows in 4 minutes where your operations stand — and what should come first.