The intelligent supervision of the industrial system is achieved through Industry 4.0. Thus, artificial intelligent-based approaches are widely constructed by incorporating the latest signal processing and sensor technologies to maintain mechanical equipment health and safety during operations. The advancement in technology has introduced multiple features in the bearing attribute space but fails to generalise the defect diagnosis feature space. As a result, it became complex, hybrid and redundant. Hence, in this work, we applied the meta-learning-based fault diagnosis approach without adopting the process of identifying the dominant feature from the feature space. Therefore, it can address the issue of non-coverage of fault region by moving along the feature space through meta-knowledge. The system provides acceptable performance with raw feature space and presents the alternative to current state-of-the-art approaches.