Technology/IT ¦ IDM
Figure 1: Exemplary data model of a “growing” digital twin along the production process
digital twin is considered the "poorest
in data" reflecting only one production
step. The last digital twin in the production
chain is the one with the most
and actual product data. However, all
twins are needed to reflect the production
process and allow traceability. Figure
1 illustrates such "growing" digital
twins using an exemplary data model
from pudding production.
Implementation into an
existing pudding production
The development steps from raw production
to a complex DTM were applied
to two use cases – the first is an
aroma supplier and the second one is
a dairy plant for yoghurt and pudding
production. The used DTM system is
based on the Siemens’ cloud platform
MindSphere, which is able to collect
data points from different MES or ERP
systems and to make them available
May 2020 ¦ international-dairy.com · 23
are not just applicable to production
plants. They can also represent a wide
variety of use cases like public transportation
systems and even whole cities.
In case of the dairy plant in focus of
this article, the benefits of using digital
twin technology are a data-based
optimization of an existing production
process and a simplified data transfer,
internal and along the entire value
chain starting with raw materials (e.g.
aroma substances) up to packaged
pudding cups. The enormous amount
of potentially usable data is the basis
for tracking and tracing of food within
the production process and between
the stakeholders involved along the
complete value chain.
Standardization of processes
and model structuring
To reach comparability between different
use cases in the DTM system, the
respective production processes have to
be analyzed and structured according
to relevant standards. Here, the standards
of the International Society of Automation
(ISA) offer plenty of possibilities
3. In this project, especially the ISA
88 and ISA 95 are applied. Initially, the
classification into work centers, work
units, equipment modules, and phases
is applied to design comparable, standardized
structures in the DTM. Using
these standards, a physical model of
each production process is constructed
where each work unit has its own
digital twin. The key aspect is that the
existing physical model only contains
equipment that gets in direct contact
with the product. Equipment modules
and phases are categorized – according
to ISA 88 – in each digital twin unit.
Often these are sensors and valves that
collect measurement data or connect
individual plant components (and thus
their digital twins).
The definition and structuring of
process-related digital twins always
require a detailed analysis of the existing
production process, based on
given production plans, SCADA plans
(Supervisory Control and Data Acquisition)
and other information from the
ERP system (Enterprise Resource Planning).
All available data points must
be collected and structured according
to their prioritization in production.
Based on this, a second model – the
data model – is designed. This model
consists of digital twins, which feed
time-dependent data points into the
model from the beginning to the end
of production. This means that the first
Figure 2: Structure of the Digital Twin Management system and interconnections
between individual stakeholders via the IoT platform MindSphere.
/international-dairy.com
/international-dairy.com
