A Quality-Gated, Event-Triggered Edge–Cloud Framework for Collaborative Predictive Maintenance of Three-Phase Induction Motors

Three-phase induction motors are widely used in industrial pumps, fans, compressors, conveyors, and production machinery. Their failures can interrupt processes, increase maintenance costs, and create secondary equipment damage. Machine-learning studies have reported promising results for vibration-based classification and sensorbased monitoring, while recent work has explored self-supervised and federated learning for noisy, distributed motor data. However, these contributions address different tasks and do not by themselves establish the performance of a single deployable Edge–Cloud system.