In short
Predictive maintenance is an approach that analyses equipment condition data (vibration, temperature, current, etc.) to anticipate when a failure may occur and plans maintenance accordingly.
Reactive maintenance intervenes after a failure, preventive maintenance at fixed intervals, and predictive maintenance based on the actual condition of the equipment. The goal is to reduce both unnecessary maintenance and unplanned downtime.
Predictive maintenance requires reliable sensor data, historical failure records and a maintenance management system that links the two. AI models are used to detect deviations from normal behaviour early.