Preventive maintenance and checklists

Planned maintenance, checklists on the shop floor and the mobile app.

Baca dalam Bahasa Indonesia

On this page
  1. Corrective, preventive, predictive
  2. The asset register
  3. Preventive maintenance in practice
  4. Running jobs
  5. In Voltrus
  6. Checklist

In most factories maintenance starts as firefighting: a machine stops, someone fixes it, and the repair is remembered by the technician who did it. The cost shows up as lost production, rush orders for spare parts and overtime, none of which appears in the maintenance budget. Planned maintenance replaces this with a schedule, a record of every job, and checklists that make inspections the same no matter who does them.

Corrective, preventive, predictive

TypeWhen it happensExample
CorrectiveAfter a failureReplacing a burnt motor
Preventive (PM)On a fixed schedule, by time or by usageChanging gearbox oil every 90 days or every 2,000 running hours
PredictiveWhen measurements show wearBearing replaced because vibration readings are rising

The goal is not zero corrective work. It is to move the expensive, disruptive failures into planned jobs done at a convenient time, with parts ready. A common rule of thumb is that a well-run plant spends most of its maintenance hours on planned work.

Two measures tell you whether it is working:

  • MTBF (mean time between failures): running time divided by the number of failures. It should rise.
  • MTTR (mean time to repair): total repair time divided by the number of repairs. It should fall, mostly through having parts and instructions ready.

The asset register

Maintenance starts with a list of what you maintain. Each asset needs at least a code, a name, a location and its status. In an ERP the same asset record serves finance, which depreciates it (see Fixed assets and depreciation), and maintenance, which keeps it running. One record means one history: what it cost, when it was serviced, and when it was disposed of.

Preventive maintenance in practice

A good PM programme has four parts:

  1. A task per asset, with a clear title ("Monthly lubrication, press 3").
  2. An interval, based on the manufacturer's manual adjusted to your experience.
  3. A checklist, so the job is done the same way each time and its results are recorded.
  4. A due list that someone actually reviews every week.

Checklists

A checklist turns "check the machine" into specific steps with recorded answers:

StepAnswer typeExample answer
Guard in place and undamaged?Yes / NoYes
Hydraulic pressure (bar)Number182
Oil level within marks?Yes / NoNo
RemarksTextTopped up 2 L, check seal next visit

A No on a yes/no step is a finding that needs follow-up, usually a corrective job. Numbers recorded every visit become a trend you can watch, which is the first step towards predictive maintenance.

Running jobs

Each maintenance job moves through a simple life: open → in progress → done (or cancelled). Record the start when work begins and done when the machine is handed back to production. Those two timestamps are the raw material for MTTR.

In Voltrus

How it works:

  • A maintenance job has a type, pm or corrective, a due date, an assignee and an asset.
  • Recurring PM is set on the job itself with recurrence days. When a PM job with recurrence is done, a scheduled job opens the next one, due the recurrence interval after the later of the old due date and the day it runs. There is one open occurrence at a time, so a missed PM shows as overdue instead of piling up duplicates.
  • Overdue jobs appear on the dashboard and can trigger a maintenance.overdue webhook or an automation, for example to escalate to a supervisor.
  • Checklist steps are yes/no, number or text. A completed run cannot be edited afterwards, and failed yes/no steps are listed so they can be followed up. The same checklists serve as quality inspection templates.
  • Asset status moves between in service and under maintenance, and every change is in the audit trail.

Honest limits today:

  • There is no separate maintenance plan record: the recurring job is its own template. Keep one recurring job per asset and task.
  • PM is calendar-based only. There are no meter readings or running-hour triggers; for usage-based tasks, convert hours to an average interval in days and review it quarterly.
  • Jobs do not record labour hours, spare parts used or cost. Issue spare parts through a stock movement and note the job in its reason, and record hours in the job notes until this is added.
  • The due list at /maintenance is not in the sidebar; reach it from the dashboard tile or global search (Ctrl/Cmd+K).
  • The mobile app has no barcode scanning; technicians pick the job from the list.

Checklist

  • Register every critical asset with a code that is painted or labelled on the machine.
  • Create one recurring PM job per asset and task, with a checklist.
  • Review the overdue list every week with production, and agree the slots.
  • Turn every failed checklist step into a corrective job.
  • Track MTBF on your five most critical machines, even in a spreadsheet at first.

Run it in Voltrus ERP

Everything in these guides is built in: documents that post to the ledger as they happen, Indonesian tax, MRP and costing, with AI agents that propose and people who approve.

See Voltrus ERP