25 terms used on a plant floor, each defined in one sentence, with the caveat that makes the definition useful in practice. Written for maintenance teams comparing systems, not for a certification exam.
A CMMS (computerised maintenance management system) is software that records maintenance assets, work orders, preventive maintenance schedules, and spare-parts consumption in one connected system.
The test of whether a plant needs one is not headcount: it is whether questions like which machine failed most, what it cost, and whether PM was done can be answered without relying on someone's memory.
What is a CMMS? →An EAM (enterprise asset management) system covers the full asset lifecycle, including acquisition, depreciation, and disposal, where a CMMS covers maintenance operations only.
EAM is typically bought by asset-intensive enterprises that need the finance side of the asset, not just its repair history.
An asset registry is the authoritative list of maintainable equipment, each entry carrying its identifier, location, documents, linked spare parts, and maintenance history.
Every other maintenance number in the plant inherits its accuracy from this list, which is why an incomplete registry quietly invalidates the reporting built on it.
MRO (maintenance, repair, and operations) inventory is the stock of spare parts and consumables held to support maintenance work rather than to be sold.
The cost of MRO stock is visible on a balance sheet; the cost of a stockout that idles a line is not, which is why MRO is chronically under-held.
Multi-tenancy is an architecture where one application instance serves many customer organisations while keeping each organisation's data fully isolated from the others.
Isolation has to be enforced on every read and write on the server, not by hiding data in the interface.
Preventive maintenance is work scheduled by elapsed time or by meter reading to prevent a failure, rather than work triggered by a failure that already happened.
Calendar-based PM is simpler to run; meter-based PM tracks actual usage and avoids servicing an idle machine on schedule.
What is preventive maintenance? →Reactive maintenance, also called run-to-failure, is repair work that starts only after an asset has already failed.
It is not automatically wrong. It is wrong when it is the default for critical assets rather than a deliberate choice for cheap, non-critical ones.
Condition-based maintenance triggers work from an observed asset condition, such as vibration, temperature, or a health score crossing a threshold, rather than from a fixed schedule.
MTTR (mean time to repair) is total downtime across incidents divided by the number of incidents, measured over a defined period and asset set.
The clock should start at failure, not at technician arrival. Starting it late is the most common way plants flatter the number without improving anything.
What is MTTR? →MTBF (mean time between failures) is total operating time divided by the number of failures, and measures reliability rather than recovery speed.
MTBF and MTTR answer different questions: MTBF asks how often an asset fails, MTTR asks how long it takes to get back.
MTTA (mean time to acknowledge) is the average time between a fault being reported and someone accepting responsibility for it.
A high MTTA with a low repair time means the problem is dispatch and alerting, not the technicians.
Unplanned downtime is production time lost to a failure that was not scheduled, measured from the moment output stops to the moment it resumes.
The reported figure is almost always lower than the real one, because short stops, slow restarts, and scrap made during the ramp back up rarely get logged.
The real cost of unplanned downtime →OEE (overall equipment effectiveness) is availability multiplied by performance multiplied by quality, expressed as a single percentage of perfect production.
MOE (Machine Overall Effectiveness) is the Firmicore composite score that blends availability, maintenance compliance, reliability, and machine health into one number per machine.
Where OEE is a production metric, MOE is a maintenance metric: it is designed to flag which machine the maintenance team should look at next.
A machine health score is a 0-100 rating derived from an asset's failure history, maintenance compliance, and open issues, used to rank risk across a fleet.
PM compliance is the percentage of preventive maintenance tasks completed within their scheduled window over a period.
Compliance measured without a window is meaningless, because a PM done two months late still counts as done.
A work order is the record of a single maintenance job: what is to be done, on which asset, by whom, with which parts, and what was found when it was done.
A work order without a sign-off step is a to-do list. The sign-off is what turns it into an audit trail.
Work order software →Maintenance backlog is the total volume of identified but incomplete work, usually expressed in crew-weeks rather than job count.
A backlog of zero is a warning sign, not a triumph: it usually means work is not being identified.
Guided triage is a branching troubleshooting flow that walks an operator through safe diagnostic steps one question at a time, so a fault can be narrowed or made safe before a technician arrives.
On a plant floor it has to survive shared devices, gloves, poor light, and weak connectivity near the machines, so each step has to be a single screen that queues offline.
Guided operator safety triage →A shift handover is the structured transfer of open maintenance context from the outgoing crew to the incoming one: pending work orders, ongoing breakdowns, low stock, and machines to watch.
Handovers done verbally lose the items nobody thought to mention, which are reliably the items that cause the next incident.
A root cause is the underlying condition whose removal prevents a failure from recurring, as distinct from the visible symptom that stopped the machine.
Root-cause fields are only useful when the option list is short and specific to the plant; a free-text box produces a hundred spellings of the same cause.
QR-triggered reporting attaches a fault report to the correct asset by scanning a code fixed to the machine, removing the step where an operator has to identify the asset by number.
Misattributed reports are the main reason paper and chat-based logs cannot produce per-machine history, so removing the typing step is what makes the history usable.
Why QR reporting beats paper logs →A permit to work is a documented authorisation that a specific high-risk job may proceed, listing the precautions taken and the person who accepted them.
A near miss is an incident that could have caused injury or damage but did not, recorded so the underlying condition can be fixed before it repeats with a worse outcome.
Lockout/tagout is the practice of physically isolating and locking an energy source before maintenance work, with a tag identifying who applied the lock.
Isolation instructions only work when they name the actual isolation point for that asset; generic safety text is ignored because it is identical everywhere.
MTTR, PM compliance, machine health, and MOE are computed from the work your team already records, so the numbers come from the workflow rather than from a monthly spreadsheet exercise.
