Maintenance and CMMS glossary.

Corrective maintenance

Reactive maintenance, breakdown maintenance · FR: maintenance corrective

Corrective maintenance is work done to restore equipment after it has failed or been found faulty. It can be immediate, when the failure stops production, or deferred to a convenient moment when it does not.

Run to failure is a legitimate strategy for cheap, non-critical equipment with spares on the shelf. The problem is corrective work on critical equipment that nobody chose: its share of total maintenance hours is one of the clearest signs of how planned a maintenance function really is.

Preventive maintenance

PM, planned maintenance · FR: maintenance préventive

Preventive maintenance is work done on a schedule to reduce the chance of failure: inspections, lubrication, adjustments and part replacements. The trigger is a calendar interval or a usage count, not the condition of the equipment.

Its risk is doing too much or the wrong thing: a part replaced on a fixed interval may have had plenty of life left, and intervention itself can introduce faults. Plans work best when their intervals are reviewed against the failures that still happen, rather than set once from the manufacturer’s manual.

Condition-based maintenance

CBM · FR: maintenance conditionnelle

Condition-based maintenance triggers work when a measured indicator crosses a threshold: vibration, temperature, oil analysis, a wear measurement. Work is done when the equipment shows it needs it, not on a fixed date.

It needs a measurement that changes before the failure, with enough warning to plan the work. Thresholds are usually set from experience and the manufacturer’s guidance, and the hard part is less the sensor than the routine that makes sure somebody acts when it alarms.

Predictive maintenance

PdM · FR: maintenance prévisionnelle

Predictive maintenance forecasts when equipment is likely to fail, from trends in condition data and failure history, and schedules work just before. It goes a step beyond condition-based maintenance by estimating remaining life rather than reacting to a threshold.

A forecast needs failure history to learn from, and well-maintained equipment fails rarely, so the data is often thinner than the project assumed. Most plants that succeed start with condition-based thresholds on a few critical assets and extend from there.

TPM

Total Productive Maintenance · FR: maintenance productive totale

Total Productive Maintenance is a method that involves operators in the routine care of their own equipment, alongside the maintenance team, with the aim of eliminating breakdowns, short stops and defects. OEE was introduced as its measure.

Its foundation is autonomous maintenance: cleaning, inspection and lubrication done by the people who run the machine, who are also the first to notice a change. It is a change in who does what, not a software feature, and it stalls when operators are given tasks without the time to do them.

Root cause analysis

RCA · FR: analyse des causes racines

Root cause analysis is a structured search for the underlying reason a failure occurred, so that the fix prevents it from recurring rather than just repairing the symptom. Common tools are the five whys, the fishbone diagram and fault tree analysis.

It is costly, so it is kept for failures that matter: repeated stops, safety events, major breakdowns. Its quality depends on the record it starts from, and an analysis built on a reason typed hours later from memory tends to find the cause somebody already believed.

CMMS

Computerised Maintenance Management System · FR: GMAO

A Computerised Maintenance Management System is the application a maintenance team plans and records its work in: requests, work orders, preventive plans, the equipment register and spare parts. Where an MES records what production did, a CMMS records what maintenance did to the equipment.

Some MES include maintenance on the same record as production; a dedicated CMMS is built around the maintenance organisation itself, including equipment no MES reads. The MES vs CMMS comparison sets out how to tell which a plant needs.

EAM

Enterprise Asset Management

Enterprise Asset Management covers the whole life of physical assets across an organisation: acquisition, maintenance, performance, cost and disposal, often across many sites. It includes what a CMMS does and extends into finance and capital planning.

The difference from a CMMS is scope rather than kind: a CMMS is centred on the maintenance work at a site, an EAM on the asset as an investment across its life. Many plants run a CMMS locally and report into an EAM or ERP centrally.

Work request

Maintenance request · FR: demande d’intervention

A work request is a report that something needs maintenance attention, raised by whoever noticed it: an operator, a supervisor, a technician. It describes the problem; it is not yet a plan to fix it.

Separating the request from the work order lets the maintenance team triage: approve, merge duplicates, reject or schedule. The easier a request is to raise, the more small problems are reported before they become breakdowns, which is the point.

Maintenance work order

FR: ordre de travail (OT)

A maintenance work order is the authorised, planned job to carry out maintenance on a piece of equipment: what is to be done, by whom, with which parts and by when. When it is closed, it records what was actually done and how long it took.

Its closing record is the equipment’s history, and that history is only as good as what technicians write when they close it. A work order closed with a cause, the parts used and the time spent is what MTBF, MTTR and cost per asset are later computed from.

Maintenance backlog

FR: carnet de travaux

The maintenance backlog is the work that has been approved but not yet done, usually expressed in labour hours or in weeks of the team’s capacity. It is the queue between a request and its completion.

An empty backlog suggests the team waits for work or that requests are not being raised; a growing one means more work arrives than the team completes. Its age matters as much as its size: a backlog of old preventive jobs is deferred risk on the equipment they cover.

PM compliance

Preventive maintenance compliance · FR: taux de réalisation du préventif

PM compliance is the share of scheduled preventive work orders completed within their allowed window over a period. It measures whether the preventive plan is being carried out as written.

The window has to be stated: completed at all, completed within a few days of the due date, or completed on the day give very different figures. A high figure on a plan that is too light is no protection, so it is read together with the failures that still occur.

Lockout/tagout

LOTO · FR: consignation

Lockout/tagout is the safety procedure that isolates equipment from its energy sources and keeps it isolated while it is being worked on. Each person working on the equipment places their own lock, and the equipment cannot be restarted until every lock is removed.

It covers every form of energy, not just electrical: pneumatic, hydraulic, stored mechanical energy and gravity. Isolation points are recorded per asset, and the procedure is part of the job plan, so the time it takes belongs in the planned duration of the work order.

Asset register

Equipment hierarchy · FR: arborescence des équipements

The asset register is the list of every piece of equipment the maintenance team looks after, organised as a hierarchy from site to line to machine to component. Each asset carries its identity, location, documentation and history.

The hierarchy decides what can be reported: costs and failures roll up only along the structure that was built. A register organised by location answers different questions from one organised by system, and most plants settle on location first with the system recorded as an attribute.

Criticality

Asset criticality · FR: criticité

Criticality ranks equipment by the consequence of its failure: on safety, on the environment, on production and on cost. It decides where maintenance effort, spare parts and monitoring are spent first.

It is usually scored on a simple matrix of consequence and likelihood, agreed by maintenance and production together. A machine with no redundancy in front of the bottleneck scores higher than an identical one with a standby beside it, which is why criticality belongs to the asset in its place, not to its model.

Spare parts

MRO inventory · FR: pièces de rechange

Spare parts are the components held in stock to repair or maintain equipment, from consumables to critical spares that are expensive and slow to source. MRO, maintenance, repair and operations, is the wider category of stock that supports running the plant.

Two failures pull in opposite directions: a part missing when a critical machine is down, and shelves of parts nobody will ever use. Linking each part to the assets it fits, and recording its use on work orders, is what lets stock levels be set on evidence rather than by habit.

Reorder point

Minimum stock · FR: point de commande

The reorder point is the stock level at which a part should be ordered again, set so that the new stock arrives before the old runs out. It covers the expected use during the supplier’s lead time plus a safety margin.

For parts used regularly it can be computed from consumption; for critical spares used once every few years it cannot, and the decision is a judgement on the cost of the downtime against the cost of holding the part. Lead times change, so reorder points set once drift out of date.

MTBF

Mean Time Between Failures · FR: MTBF

Mean time between failures is the operating time of a piece of equipment divided by the number of failures over the same period. It measures how often the equipment fails.

It means something only when the period and what counts as a failure are stated with it. Counting every micro-stop as a failure produces a figure of minutes; counting only stops that needed a technician produces one of days. Under AFNOR NF E60-182 micro-stops are a performance loss and are kept out of it.

MTTR

Mean Time To Repair · FR: MTTR

Mean time to repair is the total time spent restoring equipment after its failures, divided by the number of those failures. It measures how long the equipment takes to come back.

Whether the clock starts when the machine stops, when someone is told or when the technician arrives changes the figure, so the starting point has to be stated. Split into waiting, diagnosis, parts and repair, it shows where the time actually goes, which is often not the repair itself.