Key points
- One work order is one job on one asset. Two jobs on one sheet cannot be tracked, costed or closed separately.
- Most teams need only three types to start: preventive (planned before failure), reactive (fixing a fault) and other.
- A work order moves through a short lifecycle: open, in progress, on hold, completed and closed. Every step should be visible to everyone.
- The closing half matters as much as the opening half. A job closed without what was done and what was used leaves no history.
What goes on a work order
A work order has two halves. The first is written before the job, by whoever plans it. The second is written during and after the job, by whoever does it. Paper forms and software differ in layout, but the fields are almost always the same.
| Field | Filled in | Why it is there |
|---|---|---|
| Number | At creation | So the job can be referred to without ambiguity |
| Asset | At creation | So the history, downtime and cost land on the right machine |
| Title and description | At creation | What the problem is or what the job is, in plain words |
| Type | At creation | Separates planned work from fixing faults in every report |
| Priority | At creation | Tells the team what to do first when everything is urgent |
| Assigned to and due date | At creation or triage | Makes one person responsible for one deadline |
| Photos and attachments | Before and after | A picture of the fault saves a paragraph and a wasted trip |
| Work done | During and after | What was actually found and fixed, which may differ from the plan |
| Parts used and time spent | During and after | Feeds stock levels and the true cost of the asset |
| Sign-off | At completion | Records who accepted the work, where your site requires it |
Types of work order
Maintenance textbooks list many work order types: preventive, corrective, emergency, inspection, predictive, safety, project. In practice a long list makes people guess, and a report built on guesses is worth little. Most teams get clearer numbers from three:
| Type | What it means | Typical examples |
|---|---|---|
| Preventive | Planned before anything fails, on a schedule of time, hours or distance | 250 hour service, monthly filter change, annual safety check |
| Reactive | Raised because something is wrong: a fault, a breakdown, a failed inspection item | Hydraulic leak, door that will not close, alarm on a pump |
| Other | Work that is neither, kept separate so it does not distort the other two | Installing a new machine, moving equipment, a modification |
The split that matters most is planned against reactive. If you can see how much of your team's time goes to each, you can tell whether your preventive maintenance programme is working: over time, a programme that works shifts hours from the reactive column to the preventive one. Urgency is better recorded as priority than as a separate emergency type, so an urgent preventive job and an urgent repair are still counted correctly.
The work order lifecycle
A work order changes state as the job moves. The names differ between systems, but the steps are the same everywhere:
- Open. The job exists and has an asset, a description and ideally an owner. It may come from a work request, a breakdown, a failed inspection item, a due preventive service, or be created by hand.
- In progress. Someone has started. From this point the asset may be out of use, so the people who depend on it should be able to see that.
- On hold. The job cannot continue: waiting for a part, a permit, a contractor or access. Recording why it is blocked is what turns a stuck job into a solvable one.
- Completed. The technician has finished and recorded what was done, the parts used and any required signatures.
- Closed. A supervisor has accepted the record. A closed work order is no longer edited; if the fault returns, a new work order is opened and linked to the old one.
Small teams often merge completed and closed into one step, and that is fine. What should never happen is a work order disappearing: a job that was cancelled is marked cancelled, with a reason, so the record of what was planned and why it stopped survives.
How to write a work order a technician can follow
Most bad work orders fail the same way: the technician arrives and cannot tell what the problem is, where it is, or what done looks like. A few habits fix most of that.
- Lead with the symptom, not the guess. "Loader loses lift power when hot" is more useful than "replace hydraulic pump", which may be wrong.
- Name the asset by its number, not "the big pump" or "the one by the gate". If assets carry a QR label, scanning it fills this in.
- Add a photo. One picture of the leak, the warning light or the broken part is worth more than any description.
- Say what done means. "Leak stopped and machine back in service" is a finish line. "Look at the leak" is not.
- State anything that must happen first. A permit, an isolation, a part that must arrive. Put it above the instructions, not below them.
- Give it one owner and one date. A job assigned to "the workshop" belongs to no one.
Here is a worked example. The machine and details are invented, but the shape is what a technician needs to see:
| Field | Example |
|---|---|
| Title | Wheel loader loses lift power when hot |
| Asset | WL-014, wheel loader, Yard B |
| Type and priority | Reactive, high |
| Description | Operator reports the boom lifts slowly after about an hour of work. Normal when cold. Photo of the hydraulic oil sight glass attached. |
| Before work begins | Park on level ground, lower the bucket, isolate the machine |
| Done when | Cause found and fixed, lift tested hot, machine back in service |
| Assigned to, due | One named technician, tomorrow |
Work order, work request and job card
Three words get mixed up constantly. They describe three different moments in the same job:
- Work request: anyone saying something needs attention. It is an observation, not a commitment. See work request vs work order.
- Work order: maintenance committing to the job, with an asset, an owner and a date.
- Job card (or job sheet): the record of carrying it out, the work done, the parts, the hours, the signatures. Many systems keep it as the second half of the work order rather than a separate document, which avoids two records of one job disagreeing.
What work orders tell you once you have enough of them
A single work order gets one job done. A few hundred of them, recorded against the right assets, answer the questions managers actually ask:
- Backlog and its age: how many jobs are open, and how long the oldest has been waiting.
- Planned against reactive: the share of hours spent preventing failures compared with fixing them.
- Repeat faults: assets that keep coming back with the same problem, which is where a root-cause fix pays off.
- Cost per asset: parts and labour added up per machine, which is the honest basis for a repair or replace decision.
- Reliability: with breakdowns recorded too, the same history gives you MTBF, MTTR and availability.
None of this works on paper beyond a handful of jobs, and it barely works in a shared spreadsheet, because nobody can add up parts or downtime per asset from a pile of sheets. That is the point at which teams move to a CMMS.
Work orders in Solmex
In Solmex every work order belongs to one asset and uses the three types above: preventive, reactive and other. Work orders can be raised by hand or from a work request, a breakdown, a failed inspection item or a due preventive service, and they stay linked to where they came from. A technician opens the job on a phone, records the work done, the parts used and a photo of the finished job, and presses one button to finish; the time spent is filled in from when the work started and stopped. Each company decides who has to sign off, if anyone. A blocked job shows the reason, so a missing part is visible to the whole team rather than to whoever happened to ask.
Work orders, breakdowns and work requests are on every plan, including the free Starter plan. See work order management for the details.