What is a CMMS? Computerized maintenance management systems, and how maintenance and CI teams use one

What is a CMMS? The meaning in plain terms

CMMS stands for computerized maintenance management system. Before one, a plant's maintenance runs on a whiteboard, a job book, a spreadsheet of PM due dates and the memory of the senior technician. A CMMS puts the same things into one database: every machine with an asset number, every job raised against it, every PM with its interval, every part on the storeroom shelf, and who did what, when, and how long it took.

The work order is the unit of everything. A breakdown, a PM, a request from an operator, a small modification: each becomes a work order against one asset, with a priority, a craft, an estimate, the parts it needs and, once closed, the hours spent, the parts used and what was found. Reports, schedules and the asset history are all built from closed work orders, so the CMMS is only as good as the work orders the team writes.

The letters CMMS are also used for a maintenance management certification for people. This page is about the software.

How a work order moves through a CMMS

Most CMMS products move a work order through the same states, whatever they call them. The names below are common ones; the order matters more than the labels.

1. 1

   Requested

   Someone reports a problem or the CMMS raises a PM. A good request names the asset, what is wrong and what the person saw or heard, ideally with a photo.

2. 2

   Approved and prioritised

   A supervisor or planner checks it is real, not a duplicate, and sets the priority. A safety hazard or a line stopped goes straight to a technician; most jobs wait for planning.

3. 3

   Planned

   The planner writes the steps, picks the craft, estimates the hours, reserves the parts and notes any permit or lockout. The maintenance job plan template shows what a full plan holds.

4. 4

   Scheduled

   The job gets a date and a technician, inside a downtime window agreed with production. The weekly maintenance schedule loads the week against each craft's hours.

5. 5

   In progress

   The technician does the work and records time as it is spent, not from memory at the end of the shift.

6. 6

   Closed

   The technician records hours, parts, readings, what was found and the failure codes (what failed, why, what was done). Anything found that could not be fixed becomes a new work order. A supervisor checks the close-out before it counts.

Breakdowns skip planning: the job goes from request to a technician at once and is planned after the fact, with the codes added at close. That is fine for the emergency, but a plant where most work orders skip planning is running on breakdowns, and the planned maintenance percentage in its reports will show it. The maintenance work order form has the same fields on paper.

What the history gives you: MTBF, MTTR and the breakdown Pareto

Once a year of work orders is closed with honest hours and codes, the CMMS can answer questions nobody could answer from a whiteboard. This example is illustrative. Over 4,000 running hours, compressor 2 had three breakdowns, repaired in 1.5, 2.5 and 3.5 hours.

The same history, summed across a line, gives the breakdown Pareto. This example is illustrative: one quarter of breakdown work orders on a packaging line.

CMMS vs EAM vs ERP vs spreadsheets

The four get confused because they overlap. The difference is scope: how much of the asset's life and the business each one covers.

People often ask whether SAP is a CMMS. It is an ERP, and its plant maintenance module covers much of what a CMMS does (equipment records, maintenance plans and orders) inside the ERP, so some plants use it as their maintenance system and others run a separate CMMS that passes costs and part orders to the ERP.

When a spreadsheet stops being enough

• PMs are missed because nobody opened the file that week, and nobody finds out until the breakdown.
• Two people keep two copies of the PM list and they no longer agree.
• You cannot say how many hours of work are waiting, or which machine broke down most last quarter.
• Parts run out because nobody knew the last one was used.
• An auditor or customer asks for a machine's maintenance record and it takes a day to put together.

Moving from Excel to one system covers what to clean up before you move any list into software, and why to digitize equipment maintenance covers the case for moving off paper.

CMMS implementation: start with one line

Most CMMS projects that fail do so on data and habits, not software. Go live on one line, get the records honest there, then spread it.

1. 1

   Clean the asset list

   One asset number per machine, a naming rule, a location and a parent in the tree. Add the utilities that stop the line: compressors, chillers, dust extraction.

2. 2

   Rank criticality

   Score each asset on safety, quality, output and cost. Load PMs for the A class first.

3. 3

   Load PMs with limits

   Each task written as an action and a limit, with its trigger and interval. Do not load a PM list nobody can explain; review it as you load it.

4. 4

   Set a short failure code list

   A few codes for what failed, why and what was done, agreed with the technicians. No "Other" without a note.

5. 5

   Link the critical spares

   Start with the parts that stop the line, linked to their assets, with a reorder point and a maximum.

6. 6

   Train by doing, then review weekly

   Technicians close real jobs in the system from day one. Each week the supervisor checks a sample of close-outs, the backlog and PM compliance with the team, and fixes the data at the source.

Spread to the next line only when the first line's work orders are closed with hours and codes as a habit. The maintenance KPI tracker is a simple way to show the first line's numbers to the rest of the plant.

Common CMMS mistakes

• Buying the CMMS to fix a planning problem. Software stores the plan; it does not write it. Without someone planning and scheduling, the CMMS becomes a faster job book.
• Loading every old PM as it is. Tasks without limits and intervals nobody can explain go into the new system and keep failing there.
• Work orders closed with "Fixed". No hours, no codes, no findings, and a year later no MTBF and no Pareto.
• Breakdowns never entered. The technician fixes it, the line restarts, and the CMMS shows a machine with no failures.
• An asset tree built for accounting, not the floor. If the tree does not match how the line is laid out, downtime will not roll up to the line that lost it.
• Making requests hard. If an operator has to log in on a shared PC in the office, problems go back to the radio and never reach the history.
• Treating PM compliance as the goal. A high number with breakdowns still rising means the tasks are wrong; read it next to MTBF.