In-service equipment failure leads to considerable financial loss due to the interruption of production or service lines in different industries. Scheduled maintenance is a traditional solution to prevent unplanned equipment outages however it could be an expensive and less effective solution. Predictive maintenance algorithms monitor the deterioration of the equipment over time and improve maintenance schedules based on the equipment working conditions.
At CIRC, we are solving this problem by developing a novel autonomous monitoring system which 1. Collects the electrical/mechanical properties of the equipment 2. Learns normal and abnormal signatures in system behavior 3. Notifies the operators based on the intensity of anomalous behavior.