Digital Twin in Industrial Application – Requirements to a Comprehensive Data Model
Software version
78
Asset configuration and parameters
74
Unique ID
71
Electric components (eBoM) 68
Audit Trail 68
Mechanical components (mBoM)
67
Asset related information and documents
66
Simulation models
61
Vendor information
58
3D visualization
57
0
10
20
30
40
50
60
70
80
90
100
Figure 3: Requirements for contents of Digital Twins (x = % of positive answers, multiple selections possible) (Smart Industrial
Products, pg. 68)
conditions, transformations from K to °C). As
this interpretation or usage of data in gen-
eral can vary from case to case, such instruc-
tions for use are of great relevance.
In the use cases mentioned in the previous
section, we find a recurring pattern in IIoT
applications: the mapping of field data to a
reference in order to extract insights is the
key mechanism for generating benefits in
the business processes. Here, the digital twin
is the essential basis for any monitoring,
analysis or prediction relating not only to the
identification of assets, but also to the provi-
sion of the circumstances and state of the as-
set at any point in time in relation to the data
points. There is obviously a direct correlation
between the precision of such data and the
possible insights that may be gained.
Taking the example use cases described
above together with the survey results, we
suggest the following general reference
characteristics for a digital twin data model:
Assets
The asset is the key object in a digital twin
data model. It uniquely identifies and repre-
sents a physical thing (the classical concept
of the digital twin).
The digital twin data model must not only
contain the meta-description of the data,
but also the rules, conditions and algorithms
that define how to interpret the data in the
context of the asset in question (e.g. border
Components (including history)
An asset may consist of components (me-
chanical, electrical or electronic sub-systems
or functional units) which have to be de-
scribed insofar as they influence the asset’s
- 100 -
November 2019