Modeling a Process Shop Floor Digital Twin in Smart Manufacturing: Best Practices From a Case Study

The smart manufacturing paradigm involves the integration of Internet of Things, big data (BD), and artificial intelligence with digital twins, enabling the design of data-driven architectures. In this context, shopfloor digital twin (SDT) frameworks allow for real-time representation, simulation, monitoring, and prediction of shop floor entities, enabling the smarter and sustainable management of production processes. Different SDT frameworks were provided in the literature, but few empirical experimentations were present in real manufacturing scenarios. This issue is more evident when modeling a BD process-oriented SDT, which is the gap this article aims to narrow. Through a case study, based on the adoption of the hexadimensional shop floor digital twin (HexaSFDT) framework and using both qualitative and quantitative primary empirical data, in this article, we provide an original implementation of a process-oriented and BD-driven SDT in the shopfloor of a manufacturing aerospace company and, specifically, within the critical material-handling process. First, by modeling the material-handling process-oriented SDT, technological issues and domain tips (e.g., key performance indicators) are highlighted by experts for a better management of materials. Second, the case study shows the design and implementation of the big data analytics pipeline and how to close the loop. Third, best practices related to innovative empirical experimentation of HexaSFDT framework are emphasized, supporting companies in understanding the potential of SDT for better performance and more effective decision making.