Service-Oriented Predictive Maintenance for Large Scale Machines Based on Perception Big Data

Large scale machines (LSMs) are always crucial equipments in manufacturing. Maintaining reliability, precision and safety for LSMs is very important. However, LSMs always work under extreme condition and are prone to degradation or failure. Therefore, maintenance is important for them. Compared with preventive maintenance, predictive maintenance is cost-saving. Besides, predictive maintenance is a more sustainable way by reducing failure and enhancing safety. Condition perception is needed in predictive maintenance. Due to the complex structure and large scale of LSMs, the perception data can be characterized as Big Data. Therefore, the storage and processing of Big Data needs to be integrated into maintenance. Considering that LSMs can be distributed all over the word, cloud service can be a proper way to support maintenance in a global environment. In this paper, a framework of service-oriented predictive maintenance for LSMs based on perception Big Data is synthesized to meet those demands. The methodologies are discussed as well. Finally, an industry case is studied to illustrate the implementing of predictive maintenance.