Good practice#
Accurate records give better history, easier reporting and better maintenance over time. These are the habits that pay off most.
The four ground rules#
1. Record what is to be done, where, on which object, and who is responsible. Without the object and location, the work ends up outside the history where someone will go looking for it.
2. Document the actual work. Comments, photos, deviations, hours and materials. Not just that the job was done, but what was done.
3. Use your organisation’s method for risk. Follow the risk matrix and assessment method that applies where you work, and check that signing requirements are assigned correctly.
4. For recurring faults: consider whether something should change. If the same problem keeps coming back, the maintenance plan, checklist, risk assessment or instruction probably needs updating.
Write so it’s useful later#
The person reading your record is often you — six months from now, in the middle of a fault hunt.
| Instead of | Write |
|---|---|
| “Done” | “Replaced gasket, checked for leaks after 10 min running” |
| “Fault on the plant” | “Fan V-3 stops under high load, thermal cut-out trips” |
| “Replaced parts” | “Replaced ball bearing 6205-2RS, both sides” |
| “Looks odd” | (photo) + “Discolouration on cable insulation at terminal 4” |
Timing matters#
| ✅ As you go | Details are fresh, photos are possible, the time is accurate |
| ⚠️ Right afterwards | Works, but details slip away |
| ❌ The next day | Reconstruction. Photos are no longer possible |
Checklist before you complete#
- All checklist points approved or marked “not relevant” with a reason
- A comment explaining what was actually done
- Photos documenting the condition or the work
- Hours logged
- Materials and inventory usage recorded
- Deviation reported, if you spotted something
- JSA completed, if the job required one
If you’re in doubt whether something is worth recording — record it. Surplus information costs little. Missing information costs a fault hunt.