Edge computing-based unified condition monitoring system for process manufacturing

Condition monitoring is of great practical significance in modern manufacturing. This paper presents the implementation of an edge computing-based unified condition monitoring system, which provides all-round services for condition monitoring. These services include monitoring algorithm design and generation, monitor -ing approaches, and monitoring interfaces. With a visualized interactive interface, the monitoring algorithms can be designed online, generated into executable programs, and then implemented into corresponding edge nodes. The edge computing nodes directly act on the real-time data from production equipment which can improve response time for efficient condition monitoring. During the monitoring process, the parameters of the monitoring algorithms can be adjusted and applied in real time. Besides, the monitoring interface can be configured freely with multiple widgets, including charts and video streaming. The feasibility and practicability of the proposed monitoring system have been demonstrated through a real aluminum cold rolling case in process manufacturing.