Predictive maintenance
Predictive maintenance is the practice of using measurements and trends over time to forecast when a machine part will fail, so it is serviced shortly before that point.
What it means
It builds on condition monitoring by looking at the direction of change, not just a single reading. Rising vibration, creeping motor current or growing temperature, tracked week after week, point to a likely failure date. The team then plans the repair at a convenient time. It needs regular readings, history for comparison and some knowledge of what normal looks like. It is most worthwhile for expensive or critical machines.
Example
A pump's bearing temperature rises from 55 to 72 degrees over ten weeks. Based on the trend, maintenance replaces the bearing at the next Sunday stop for ₹4,200, instead of an unplanned failure during a shift.
Records and checks
- Record each reading with date, machine ID, instrument and the normal range.
- Verify instruments are calibrated and readings are taken at the same point under similar load.
- The technician records trends; the supervisor decides the repair date, and the owner reviews savings.
Common mistakes
- Collecting readings but never plotting them, so the trend is missed.
- Taking readings at different loads or places, which makes them incomparable.