Predictive Maintenance is an advanced technique that uses data analysis and condition monitoring to anticipate equipment failures before they occur. Unlike preventive maintenance (which is based on fixed schedules), predictive maintenance is based on the asset's actual condition.
The premise is simple: "The equipment will tell us when it needs attention." Through the use of sensors, IoT (Internet of Things), and algorithms, anomalies are detected—such as vibrations, temperature increases, or irregular electrical consumption—that precede a failure.
Key Difference: Preventive vs. Predictive
It's common to confuse them, but the difference is economic and operational:
- Preventive (Calendar): You change the car oil every 10,000 km, whether it's dirty or not. It's safe, but you may be wasting the useful life of a part that still worked.
- Predictive (Condition): You change the oil only when a sensor analyzes its viscosity and alerts you that it has lost properties. You maximize the resource, intervening at the exact moment: neither too soon (unnecessary expense) nor too late (breakdown).
The Role of IoT and Sensorization
For outdoor assets, predictive maintenance is possible thanks to sensorization:
- Lighting: Remote management systems alert if a luminaire consumes more current than it should (indication of imminent driver failure).
- Solar PV: Monitoring of strings and inverters detects performance drops or abnormal temperatures before the equipment fails.
- Water: Flow meters that detect micro-leaks by analyzing nighttime pressure patterns.
What the product does, and what it does not
Predictive maintenance, artificial intelligence models and connections to IoT platforms, remote management systems or SCADA are outside the scope of Maptainer. What it does contribute to a condition-based strategy is this:
- Analytics on history: MTBF by asset type, MTTR by criticality, availability, and the ratio of preventive to corrective work, calculated on what was actually executed. They describe what has already happened; they do not anticipate what will happen.
- Meter readings with thresholds: each measuring point carries its admissible values, an out-of-range reading is flagged on the spot, even without coverage, and, if the point requires it, the system opens an incident.
- Incident creation via the API: if the organization already has a monitoring system, it can create incidents in Maptainer through the public API, which handles them with the same process as any other: SLA, work order, planning, and closure. Detection belongs to the external system; Maptainer manages the intervention.
On that basis, the preventive plan is reviewed with real data, and alerts generated by other systems reach the work order without being copied by hand.