The advance of the Industrial Internet of Things has largely increased the availability of shop floor data for software applications in production management. Digital Twins utilise shop floor data to provide synchronised virtual representations of production resources, e.g. by applying Data analytics such as Machine Learning techniques. A main barrier to take advantage of the potential of Digital Twins is the personnel effort and skills required for development, which can be lowered by ready-to-use approaches. This article proposes a corresponding software infrastructure. Prebuilt and configurable Digital Twin components are allocated to suitable platforms and linked by suitable data connections. To enable interoperability with production management applications, the standard of the Asset Administration Shell is utilised. Establishing resource-specific shop floor data connection and transformation is facilitated by a shopfloor integration platform, which provides configurable components. The software infrastructure is demonstrated in a laboratory environment, where Digital Twins of a manufacturing machine and an assembly robot entail Machine Learning models to predict and provide capacity parameters to a discrete event simulation.
Luber et al. (Thu,) studied this question.