As the name implies, predictive maintenance techniques assess the condition of equipment to determine when maintenance will be needed. Unlike preventive maintenance, in which service is performed at regular, set intervals regardless of equipment performance, predictive maintenance can save costs, because service is only performed when it’s needed. Predictive maintenance generally includes three phases: surveillance, diagnosis, and remedy.
In the surveillance phase, equipment conditions are monitored to identify potential problems. In the diagnosis phase the root cause of the issue is identified, and in the remedy phase corrective action is taken to correct the problem.
The surveillance and diagnosis phases in particular lean heavily on data collection and assessment. A smart building’s automation hardware and sensors can be used to gather the necessary data, which can be delivered to smart building software. The software can then utilize trends to figure out and “predict” when maintenance may be needed, and trigger an alarm to notify a facility manager.