Predictive Maintenance System using motor current signal analysis for Industrial Robot

Predictive Maintenance (PdM) is one of the key enabling technologies in Industry 4.0. The Factories of the Future will adopt highly automated and interconnected environment where predictive fault detection will have an essential role to ensure efficient and reliable industrial operations. Due to their high efficiency and their low cost Cartesian Robots (CRs) represent one of the widely used automation systems in industry. Their movements and efficiency depends on transmission system and its degradation. However not much has been done in terms of PdM for these robots and very few works tries to deal with this problems. Different failures for those kind of robots are attributable to the transmission system. This work details the effect of the transmission system on the robot electrical actuation according to Motor Current Signal Analysis (MCSA) theory. This analysis propose different tools, used in others disciplines for different purposes, to infer features of the faulty condition. By monitoring the motor current of the CR, after a signal preprocessing, a proper fault index have been investigated in order to detect the functionality state of the transmission system. The preliminary results obtained are encouraging compared to classic spectral analysis. The monitoring and analysis have also been extended to the transient state. All the fault detection tests have been carried out directly on the electric drive mounted on a real industrial CR.