Software-defined Cloud Manufacturing in the Context of Industry 4.0

In the practice of "Cloud Manufacturing (CMfg)" or "Industrial Internet", there still exist key problems, including: 1) big data analytics and decision-making in the cloud could not meet the requirements of time-sensitive manufacturing applications, moreover uploading ZettaBytes of future device data to the cloud may cause serious network congestion, 2) the manufacturing system lacks openness and evolvability, thus restricting the rapid optimization and transformation of the system, 3) big data from the shop-floor IoT devices and the internet has not been effectively utilized to guide the optimization and upgrade of the manufacturing system. In view of these key practical problems, we propose an open evolutionary architecture of intelligent CMfg system with collaborative edge and cloud processing capability. Hierarchical gateways near shop-floor things are introduced to enable fast processing for time-sensitive applications. Big data in another dimension from the software defined perspective will be used to decide the efficient operations and highly dynamic upgrade of the system. From the software system view, we also propose a new mode - AI-Mfg-Ops (Al-enabled Cloud Manufacturing Operations) with a supporting framework, which can promote the fast operation and upgrading of CMfg systems with AI enabled monitoring-analysis-planning-execution close loop. This work can improve the universality of CMfg for real-time fast response and operation & upgrading.