The Data Centric Architecture of a Factory Digital Twin
The goal of the Resource Task Network ( RTN ) is to provide a standard description of manufacturing operations in as much detail as necessary to answer questions of interest ( Zentner et al 1998 [ 7 ], Perez et al 2022 [ 16 ]). The RTN provides for data centricity and structure , highlights manufacturing data as a core asset , but is independent of the technology used to implement the FDT . The RTN readily captures bill of material and process flow information but must be extended for specific process physics ( see below ) and specialized activities , such as quality testing and equipment cleaning that depend on manufacturing technology . Incorporation of and addressing process specific physics is a critical factor to the success and differentiation of FDT technology and starts with the RTN process description .
All value-adding processes involve the transformation of one or more inputs into more valuable products . This is true whether the process involves building automobiles , producing chemicals , fermenting biological products , or managing a complex project with many contributing activities . Modeling any nontrivial process requires a formal framework capable of describing the inputs and the process steps that transform them . The RTN has been a standard tool for describing processes for many years . For the sake of concreteness , we describe the use of the RTN for manufacturing processes below , but it is also applicable to organizing project management and other processes that transform abstract inputs over time .
As the name implies , the RTN is a network of nodes that are interconnected via directed arcs . Each material in a process is represented by a material node . Tasks consume one set of materials ( inputs / feeds ) and produce another set ( products ). Figure 2-1 illustrates a simple RTN .
Figure 2-1 : A diagram of a simple RTN .
This network RTN contains a single task T1 that converts material A into material B . Figure 2-2 shows a still simple , but more realistic example of an industrial process with eight tasks and five materials . The RTN can easily describe processes of arbitrary depth ( the number of processing steps from start to end ) as well as process parallelism .
Journal of Innovation 101