IIC Journal of Innovation 12th Edition | Page 20

Digital Twin + Industrial Internet for Smart Manufacturing: A Case Study in the Steel Industry Based on our experience, the feedback first and foremost comes from product quality, followed by energy consumption, material supply, equipment conditions and up- stream and down-stream sub-processes. This feedback helps determine the best set of engineering parameters for meeting product quality, lower energy and material consumption and achieve a higher production rate. from a fleet of assets that are well-organized in association with each other. The required level of complexity in analytics is clearly increased as a result (Figure 3). The fragmented data silos—as well as the absence of a systematic description in the digital space of complex production environment analytics found in many manufacturing environments today— together present a great obstacle to achieve in-depth collaborative analytics. In other words, we need a systematic approach to represent the real world in the digital space and facilitate these sophisticated analytics. On the other hand, to realize data-driven optimization, the analytics become more involved—progressing from descriptive to diagnostic, predictive and prescriptive. Its scope also expands from analyzing a single asset (e.g. in the case of predictive maintenance) to a fleet of assets (e.g. in a production line, or even across production sub-processes such as sintering and casting in an iron-and-steel manufacturing process). This type of analytics relies on data collected D IGITAL T WIN The concept of digital twin has garnered increasing attention in the recent years because it can be used to systematically describe the real world, including physical assets and logical processes, in the digital space. Figure 3: Increasing Analytics Complexity IIC Journal of Innovation - 15 -