Page 29 - Kaleidoscope Academic Conference Proceedings 2020

of a plant, geometric information, or quality data, is often
asset- and domain-specific, and therefore not part of the
common AAS. This asset-specific data is stored in sub-
models. One AAS may refer to multiple sub-models that
store specific data. In the case of a device, sub-models track,
for example, the operation hours or electronic data sheets of
the device. Product AASs contain sub-models that define
recipes for manufacturing the product, quality requirements,
and measured quality properties.

The AAS is being developed by a consortium of partners
from academia and industry, led by the German Platform
Industrie 4.0. The development of sub-models is an ongoing
process, which requires a high degree of domain knowledge.
OPC-UA companion specifications may add an important
value here. Ongoing work aims at integrating these
companion specifications as AAS sub-models. Three types
of AASs are distinguished: Type one AAS are XML or JSON
files containing serializations of AAS and sub-model data.
Type 2 AASs are runtime instances deployed to devices or
servers. They provide http/REST or OPC-UA interfaces to
Figure 3 – Extending the Industrie 4.0 maturity index with communicate with applications. Type 3 AASs are type 2
the degree of digitalization AASs that additionally initiate actions. They communicate
with other AASs to negotiate, for example, orders or
Our matrix illustrates the digitalized production assets maintenance intervals. The languages for type 3 AAS are the
integrated into a digital twin, and exemplifies use cases. The subject of current research.
first level of Industrie 4.0, level 3, “Seeing”, may be applied
to individual machines, to a complete production process, or The AAS is specified in a technology-independent manner.
to a full inter-company supply chain. Digitalization of However, Eclipse Basyx [4] already provides an
individual machines on level 3, for example, yields implementation and a runtime interface to the asset
individual digital twins that support the creation of machine administration shell.
dashboards. Creation of an integrated digital twin for a
production process that also includes products additionally 2.3 Industrie 4.0 solution technologies
enables the documentation of process steps and the creation
of a digital twin to document product quality. If this Existing Industrie 4.0 solutions support the digitalization of
information is to be used across companies or organizational manufacturing and therefore the reaching of the six levels of
units, harmonization of data types and data forms becomes the Industrie 4.0 maturity model, as well as the levels of our
necessary. matrix.

2.2 Industrie 4.0 asset administration shell Industrial Internet of Things (IIoT) platforms [5] focus on
the collection and transport of sensor and machine data. Edge
[2] indicates the need for standardization as one of the major devices connect to proprietary field busses and enable users
challenges for the fourth industrial revolutions. Today, many to access data provided by PLCs through unified APIs. The
devices already provide data, but use proprietary data IIoT platform forwards data from the edge devices through
formats and protocols to structure and present the data. (IP) networks to processing backends that aggregate this data
Digital twins need to aggregate data from different sources. and implement data analysis or visualize data with
Standardized, or at least agreed, data structures are therefore dashboards.
needed.
Lightweight communication protocols such as MQTT (an
The asset administration shell (AAS) [3] is a vendor- OASIS standard messaging protocol for the Internet of
independent digital representation for Industrie 4.0 assets. Things (IoT), see [6]) did originate in the IoT domain, but
The AAS turns production-relevant assets into manageable are also used in industrial domains. MQTT does define a
Industrie 4.0 assets. This includes, e.g., product orders, protocol, but no data models, which are necessary for
manufacturing devices, workers, or processes. The asset supporting end-to-end communication across devices and
administration shell provides a unique identifier for assets device types. Today, the meaning of transmitted data is often
and defines a meta-model for the description of the most device-specific, and data sheets are required to interpret
important asset properties. It also defines a type system and received values. The Open Platform Communications United
enables the semantic enrichment of data to enable a better Architecture (OPC-UA) [7] is an initially client/server-based
understanding of data and the mapping of similar data communication platform for the automation industry. It
structured according to different data models. The topology enables the creation of custom data structures that represent

– xxv –