Predictive maintenance means not too early, not too late, ideally planned and accompanied by the sensible use of human resource. Prescribed maintenance intervals may be too late due to more intensive use of a machine and the equipment might already be on the verge of failure.
Too early maintenance, on the one hand, may send in the mechanic unnecessarily, and on the other hand, as a result of that, he might also have to return to the same plant after a shorter time if it really gets critical. Here the working hours of the mechanic or technician is not used meaningfully.
The predictive maintenance means that the ideal time for a maintenance measure can be determined. This is possible when systems are equipped with sensors that report wear and workload on the components. Ideally, an IoT online connection compares specified runtime data and real usage data to determine the ideal time for efficient maintenance for a cost and resource efficient operation.