See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/347154907

Generating Industry 4.0 Asset Administration Shells with Data from

Engineering Data Logistics

Conference Paper · September 2020

DOI: 10.1109/ETFA46521.2020.9212149

CITATIONS READS

28 198

4 authors, including:

Anna-Kristin Behnert Felix Rinker

Otto-von-Guericke-Universität Magdeburg TU Wien
7 PUBLICATIONS 53 CITATIONS 37 PUBLICATIONS 269 CITATIONS

SEE PROFILE SEE PROFILE

Stefan Biffl
TU Wien
449 PUBLICATIONS 6,034 CITATIONS

SEE PROFILE

All content following this page was uploaded by Anna-Kristin Behnert on 01 February 2022.

The user has requested enhancement of the downloaded file.

 Generating Industry 4.0 Asset Administration Shells
with Data from Engineering Data Logistics
Arndt Lüder, Anna-Kristin Behnert Felix Rinker1,2, Stefan Biffl1
Otto-v.-Guericke University, FMB, IAF, TU Wien, Inst. of Information Systems Eng., 2CDL-SQI
1

Magdeburg, Germany Vienna, Austria

[arndt.lueder|anna-kristin.behnert]@ovgu.de [felix.rinker|stefan.biffl]@tuwien.ac.at

Abstract—Advanced production systems, for example in the providing not only identification and communication
European automotive or steel industries, incorporate Industry capabilities of the I4.0 component to the CPPS but also
4.0 defined structures based on Industry 4.0 components and
their asset administration shells. Several development and providing engineering information on the planned structure
standardization initiatives define structures, behavior patterns, and behavior of the CPPS and runtime data on their
meta models, etc. These initiatives assume the Industry 4.0 asset historical, current, and potential future states and behavior,
administration shell (I4.0AAS) as a digital twin of the Industry and giving access to technical functions of the CPPS [5], [6].
4.0 component to contain all relevant engineering and runtime In addition, CPPS, and thereby Industry 4.0, components
data. However, there is only limited discussion on how to collect
and represent these data effectively and efficiently. In this with their administration shells can be found on various ab-
paper, we discuss the collection and representation of straction levels of production systems. In case of a steel mill,
engineering data as part of the I4.0AAS. We introduce the we can find I4.0AASs related to a complete coil car, a drive
I4.0AAS completion method to facilitate the collection of the chain within the coil car, and the drive of this drive chain.
I4.0AAS engineering data set and its export to an I4.0AAS One key challenge within the realization of the I4.0AAS is
serialization. In the multi-disciplinary engineering of
production systems, the I4.0AAS completion method builds on the effective and efficient collection of the scattered and
the AutomationML standard and on the data logistics between heterogeneously represented information required to give a
engineering workgroups. We evaluate the I4.0AAS completion complete representation of the planned (and eventually
method in a feasibility study with an I4.0 measurement cell. realized) structure and behavior that may differ on the
different abstraction levels. At begin of the physical life time
Keywords—Production system engineering, engineering net-
work, engineering data logistics, Industry 4.0 components of each CPPS this information shall be collected from the
management shell, AutomationML multi-disciplinary process of production system engineering
[3]. Usually the system integrator is responsible for this
I. INTRODUCTION collection. He is the leading stakeholder of the production
Production system owner/users within nearly all system engineering having the responsibility as well as the
developed and developing countries worldwide are facing capability of providing the initial content of the I4.0AAS
common challenges: increasing costs of input factors (during and) at the end of the CPPS engineering process.
including materials and human labor accompanied by Therefore, this paper focuses on the end of the engineering
increasing customer power related to product-oriented process, when the main structuring and initial completion of
requirements. The first challenge leads management to call the I4.0AAS occurs.
for cost reduction usually accompanied by mass production We will use the engineering of a steel mill as motivating
for economy of scale, while the second challenge calls for example. Here the system integrator (represented by
improved flexibility of production. To address both trends, companies like voestalpine or SMS group) is responsible not
production system design and application innovators focus only for engineering of the production system but also for its
on increased digitalization of production system covered by commissioning and ramp-up. Following Industry 4.0 claims
terms like Industry 4.0 in Germany, Industrial Internet after this ramp-up for each Industry 4.0 component the
Consortium in US, Factory of the future in France and UK I4.0AAS has to be completed. Thus the system integrator has
and others [1], [2]. to be able to decide on which information shall be integrated
Key concept of this digitalization is the Cyber-Physical in which structure of the I4.0AAS. This decision need to be
Production System (CPPS) that combines the physical taken for the complete system as well as all involved
production process and its basic control with increased component including coil cars, drive chains, and drives
intelligence implemented on top of the control system, Within CPPS engineering, engineers take engineering
providing the CPPS with additional intelligence and self-* decisions based on appropriate knowledge, input data, and
capabilities [3]. Following the Reference Architecture Model tools that generate output data in various languages and
for Industry 4.0 (RAMI 4.0), a CPPS is represented by formats [7]. These engineering actions form a data logistical
Industry 4.0 components, which contain additional intelli- network, linked by engineering data exchange. Within these
gence implemented by the Industry 4.0 asset administration networks engineering data are represented by engineering
shell (I4.0AAS) [4]. The I4.0AAS is envisioned as means for data models related to one engineering discipline (discipline-
internal data models) and data models related to more than

978-1-7281-8956-7/20/$31.00 ©2020 IEEE 867

Authorized licensed use limited to: Otto-von-Guericke Universitaet Magdeburg. Downloaded on February 01,2022 at 12:31:48 UTC from IEEE Xplore. Restrictions apply.
 one discipline (discipliine-crossing data modells) as objeect, such as ddrive, sensor, rrobot, or machhine, or a logiical
foundation foor data logistics. Usually syystem integrattors are objeect, the asset administratioon shell (AAS S) is a virtual ob-
applying dataa logistic systeems ensuring the engineerinng data jectt enabling deescription of and access tto the asset [6].
exchange wiithin the nettwork. For example e the system Reccently, structuure and use off the Industryy 4.0 componeent1
integrator witthin steel induustry can applly an AutomattionML are,, at least in Euurope, discusseed and agreed [10], making it a
based data loggistic implementation as preesented in [377]. goood starting poinnt for implem mentation approoaches [11].
To fill the I4.0AAS witth engineeringg data, potentiially all T
The Industry 44.0 asset adm ministration shhell (I4.0AAS) is
models appliied within thee engineering process needd to be assuumed to repreesent and makke accessible thhe asset alongg its
seen as a heeterogeneous set of data sources. Thesse data commplete life cyccle following the ideas of IEC 62890 [338].
sources havee to be integraated and com mbined approppriately, Fig.1 shows the structure of ann I4.0AAS [44]. Each I4.0A AAS
leading to twoo main open qquestions: conntains one asseet data object and holds a sset of submoddels.
• Which information shall s be part oof the I4.0AA AS and Whhile the assett data object represents thhe identificattion
how shaall this informaation be structtured? infoormation of thhe linked assett the submodells are intendedd to
• How to design the innformation floow from enginneering reprresent inform mation items that
t are technnically separaated
disciplinnes to the I4.00AAS? fromm each other but together fform a consisttent model of the
Intention oof this paper is to give firrst answers too these asseet of interestt. Each subm model may coontain elemennts,
questions by presenting thhe AutomationML-based I4 I4.0AAS whiich can be prooperties, references, files, aand several moore.
completion method.
m This mmethod focusees on the needss of the Beyyond this struucture, the I44.0AAS speciification requiires
system integrrator, who is responsible ffor the effectiive and cappabilities for iddentifying andd referring to innformation iteems
efficient inteegration of alll relevant enggineering dataa to be andd sets, and for further modellling. For detaiils see [12].
collected in an initial I44.0AAS. Therrefore, the I4 I4.0AAS
completion method
m shall fuulfill the follow
wing requirem ments:
Req.1: Efficieent engineerinng data identifi fication - The m method
shall eenable the systtem integratorr to efficientlyy decide
on whhich data sets tto represent in the I4.0AAS..
Req.2: Efficieent engineeriing data reppresentation - The
methood shall autom mate the represeentation of ideentified
engineeering data to the possible eextent.
Req.3: Efficieent engineerinng data integrration - The m method
shall build on engineering data d collectioon and
Fig. 1. Basic moddel structure of I44.0AAS, based onn [4], [12].
integraation capabillities of thee engineeringg data
logistiics in an enginneering networrk. W
While Plattforrm Industrie 44.0 has definedd the basic strruc-
To reach the intended aim, the papper is structuured as turee of an I4.0A AAS in [4], thhey improved on this structture
follows. Secction 2 review ws the state of the art oof data andd defined first data represeentations for the I4.0AAS S in
management in an engineeering networrk and the status of [12]. Thereby, thhe I4.0AAS sstructure refleects the differrent
I4.0AAS stanndardization. Section 3 mootivates the reesearch neeeds of the Inddustry 4.0 commponent life ccycle. The guiide-
questions andd approach. Section
S 4 desscribes the I44.0AAS linee [12] recomm mends AutommationML for serialization du-
completion method coombining AutomationML A L-based ringg CPPS enginneering, AAS SX for CPPS commissioniing,
engineering ddata logistics and I4.0AAS S serialization with a andd OPC UA foor CPPS operration; and presents represeent-
short represeentative exam mple. Sectionn 5 reports on the tatioons of the I4.00AAS meta model
m for these data formats.
evaluation off the I4.0AAS completion m method in a feaasibility
B. Information loogistics withinn engineering networks
study and lessons learned. Section 6 conncludes the papper and
proposes futuure research. T
The engineerinng of productiion systems (liike any techniical
systtem) is a muulti-disciplinarry engineeringg process [3].. In
II. STATE O
OF THE ART AN
ND I4.0AAS STANDARDIZAT
TION genneral, it can bbe regarded aas a network of engineeriing-
A. Standardiization of Induustry 4.0 assett administratioon shell disccipline-specifiic design deciisions that aree related to each
otheer and, therefo
fore, depend on the exchangge of informattion
In the Induustry 4.0 deveelopment and standardizatioon pro- on taken design decisions. In total, the engiineering netw work
cess, variouss partners, unnder the roof of different organi- designs step by step all neceessary informaation requiredd to
zations like V
VDI, VDMA,, ZVEI, and OPC O Foundatiion, are commmission, rammp-up, and run the productioon system.
involved devveloping an auutomation archhitecture (andd more) T
The sequence and granularrity of enginneering decisions
that faces thhe challengess of the chaanging internnational deppend on the application ccase. Followinng the differrent
markets [1]. Within this approach, thhe focus is on the induustries, prodduction system m complexitty, and control
Industry 4.0 component, thhe automationn system com mponent techhnologies appplied in the pproduction syystems, differrent
and its capabbilities [9]. Thhe I4.0 componnent is based on two
main componnents, the assset and the aasset adminisstration 1
See the definitioon and standardizzation of the Inddustry 4.0 component
shell related to the asset. W While the assset can be a pphysical and its structure at htttps://www.plattfoorm-i40.de.

868
Authorized licensed use limited to: Otto-von-Guericke Universitaet Magdeburg. Downloaded on February 01,2022 at 12:31:48 UTC from IEEE Xplore. Restrictions apply.
 engineering process guidelines have been developed and XML-based data exchange format. AutomationML enables
standardized [13]. They range from general guidelines, such modeling hierarchies of physical and logical production
as the VDI Guidelines 2221 and 22062, over modelling- system components as data objects encapsulating all relevant
oriented guidelines, like SysMod, to industry-driven aspects including (but not limited to) topology, geometry,
guidelines, like the STEP and AutomationML reference kinematics, and logic information. AutomationML is
engineering processes. Nevertheless, they usually form very currently standardized as IEC standard 62714 series.
similar structures with very similar engineering tools and Main model of AutomationML is the topology description
data exchange structures [14]. In addition, production system architecture exploiting the XML data format CAEX4. This
engineering networks have been increasingly influenced by data format enables an object-related approach, where object
and interlinked with other engineering networks. As main semantics (i.e., the identification of modelling concepts) are
impact, product engineering needs to be named [15] as it specified as roles and the semantics of relations between
originates the main requirements to production system objects (i.e., relations between modelling concepts) are
engineering and is the main cause of CPPS engineering specified as interfaces. Both roles and interfaces are defined
complexity [16]. and collected in role/interface class libraries. In addition,
Engineering networks of production systems apply a wide reusable objects acting as modelling templates are provided
variety of data models, data formats, and modelling means by classes of system objects (named system unit classes),
[22], such as Entity–Relationship model (ER model) to defined and collected in system unit class libraries. Finally,
define abstract data models [23], UML [24] and SysML [25] individual objects are modeled in an instance hierarchy as a
to describe models explicitly developed for Model-Driven hierarchy of internal elements, referring to system unit
Engineering, and specialized modelling dialects of the classes they are derived from, role classes defining their
involved engineering disciplines and tools [14]. semantics, and interfaces to link objects with each other or to
To bring these different engineering data models together external information.
in a consistent engineering network, requires appropriate For the modelling of dedicated engineering data exchange
information modelling means that provide capabilities for use cases, AutomationML provides domain-specific
the lossless exchange of engineering information between languages (DSLs) based on a set of use-case-dependent role
engineering tools [26], [27]. and interface class libraries that are part of the Auto-
As shown in [17], AutomationML can be applied for mationML standard [27]. Examples of such AutomationML
modelling engineering data for the data exchange along the DSLs [28] implement the data exchange between electrical
complete engineering chain of production systems. In [18] engineering and PLC programming and [29] connect
and [19], engineering data logistics based on a centralized production system engineering and production planning.
engineering data storage using AutomationML are discussed, With respect to this paper, special emphasis is on
while in [20] a more decentralized data logistics based on AutomationML DSLs for the serialization of I4.0AAS [30],
AutomationML is considered. which utilizes the similarities of I4.0AAS meta model [12]
Beyond AutomationML, there are further data formats that and AutomationML regarding the separation of syntax and
may be applicable for the implementation of an appropriate semantics for the definition of mapping rules between
information exchange within engineering networks. Leading I4.0AAS models and AutomationML models. The mapping
candidates potentially covering all relevant engineering rule set defines a set of roles classes, interface classes, and
disciplines are STEP and XMI. STEP (ISO 10303 standard attributes for representing the I4.0AAS meta model structure
series) is tailored to the representation of general technical elements. Fig. 2 shows the resulting class sets.
systems, but lacks capabilities for easy extensibility and In addition, the mapping rules set defines that
adaptability to changing engineering networks. In contrast, AutomationML element name and identification based on
XMI, as an OMG standard, is easy to extend and adapt, the I4.0AAS element idShort shall be the same and shall
designed for transferring UML and SysML models, but only be used if necessary, I4.0 AAS submodels are
limited regarding modeling engineering data of general represented by appropriate AML InternalElements with the
technical systems. Submodel role, and each SubmodelElement shall be an
InternalElement with an appropriate role and an internal
C. AutomationML
structure matching to the type of SubmodelElement (for
AutomationML is both easy to extend and adapt as well as example containing an interface in case of a reference).
tailored to the domain of production system engineering The defined role and interface classes for I4.0AAS meta
[17]. The AutomationML data format has been developed by model representation can be combined with all other
AutomationML e.V.3 as an open, vendor-independent, and AutomationML DSLs as required, e.g., with the upcoming
2 automation component modelling DSL [31].
See https://www.vdi.de/tg-fachgesellschaften/vdi-gesellschaft-mess-
und-automatisierungstechnik/engineering-und-betrieb-automatisierter-
anlagen?tx_vdiguidelines_guidelinelist[filter][organizations][0]=24&tx_vdi
4
guidelines_guidelinelist[filter][organizations][1]=56&tx_vdiguidelines_guid For details on this structure, refer to the AutomationML whitepapers at
elinelist[filter][searchTerm]=engineering www.automationml.org.
3
See www.automationml.org

869
Authorized licensed use limited to: Otto-von-Guericke Universitaet Magdeburg. Downloaded on February 01,2022 at 12:31:48 UTC from IEEE Xplore. Restrictions apply.
 R
RQ2: How cann the engineerring data logiistics be adappted
to enable aautomated serialization of I4.0A
AAS
AuttomationML?
T
The AutomatioonML based engineering ddata logistics [7],
[8] is a network of data sourcces, sinks, andd an intermediiate
struucture for datta transport, transformation
t n, selection, aand
commbination (seee Fig. 3). For ffilling the I4.00AAS submoddels,
Fig. 3 depicts tthe I4.0AAS as a data ssink and relaated
enggineering activvities as data sources.

Fig. 2. Autom
mationML role annd interface classses for I4.0AAS modelling
m
follow
wing [30].

III. RESEARCH QU
UESTIONS AND APPROACH

Aim of thiis paper is to present an A AutomationML L based

I4.0AAS coompletion meethod to suupport the ssystems Fig. 3. Data logisstics in an engineeering network, baased on [7], [8].
integrator in collecting rellevant engineering data aloong the
CPPS enginneering netwoork and to create an I44.0AAS T
To address R RQ2, we will design extennsion points aand
serialization. This methodd requires dessigning (a) I44.0AAS describe the requuired algorithm
ms.
information sstructuring annd (b) informaation flow (inccluding IV.. AUTOMATION
NML-BASED I4
4.0AAS COMPPLETION METH
HOD
necessary moodel and data ttransformationns).
According to the state oof the art and standardizatioon (see T
This section inntroduces the A
AutomationM
ML based I4.0AAAS
Section II), thhe I4.0AAS mmeta model deffines an abstraact data commpletion methhod enhancinng the Autom mationML baased
structuring bbased on subbmodels and submodel ellements enggineering dataa logistics [77], [8] to coollect I4.0 A
AAS
without definning valid infoormation canddidates to be ccovered subbmodel data annd to serialize I4.0AAS AutomationML.
by submodelss, motivating research
r questtion RQ1. A. AutomationM ML-based Enggineering Datta Logistics
RQ1: Whhich I4.0 assset administrration shell (AAS)
S
Starting point for the designn and implem mentation of each
submodels dooes the system m integrator require to reppresent
enggineering dataa logistics is tthe observatioon that there are
main data eleements in engiineering netwoorks?
twoo main types of data modeels applied wiithin engineerring
As the sysstem integratoor shall be ablle to effectiveely and
netwworks, data mmodels appliedd within an enngineering dissci-
efficiently ideentify engineeering data to be
b representedd within
plinne and/or tool and data moddels that are appplied to transsfer
an I4.0AAS, it must bee possible too select apprropriate
enggineering infoormation betw ween engineeering discipliines
discipline-intternal as well as discipline-ccrossing data models
andd/or tools [7]. All these enggineering moddels form dom main
for integration within the I4.0AAS.
as data sourcee candidates fo
specific languagees (DSLs) thaat fulfill the special needss of
Following VDI 3695 [39] it is assum med in this papper that
theiir use more orr less well. Buut they were not
n developed for
the engineeriing network aapplied by the system integrator is
an engineering ddata logistics covering the exchange of the
supported byy an engineeriing data logistics that impllements
commplete set oof engineeringg data relevvant within the
the informattion exchangee between innvolved enginneering
production system m engineeringg [32].
activities byy realizing the artifact exchange bbetween
To reach such an overall lannguage, severaal approaches are
T
engineering ttools. Hence, to address RQ Q1, we will consider
posssible [33]. Wiith the AML12 data logisticcs one promissing
the structure and behaviorr of an enginneering data loogistics
appproach has beeen described in [7], [8]. T This approachh is
and derive caandidates for I4.0 AAS subbmodels. As tthe I4.0
based on the application of AutomationML and the
AAS compleetion method sshall be basedd on AutomatiionML,
commbination of AutomationM ML DSLs. E Each local, ii.e.,
results from m standardizzation activitties around Auto-
enggineering domaain and/or toool specific moddel is represennted
mationML annd the Industryy 4.0 componeent will be appplied.
by an AutomattionML DSL L with approopriate role aand
The resulting set of submmodels is the ffoundation to discuss
inteerface classes as well as sysstem unit classses derived frrom
the enhancem ment of enginneering data logistic approoaches,
themm (named AML2 A in [88]). They ccover the ussual
resulting in the I4.0 AA AS completioon method ffor the
enggineering conccepts applied within the respective moddels.
automated ggeneration of an I4.0AA AS AutomattionML
As the engineeriing disciplines and tools hhandle aspectss of
serialization.
the same techhnical objects, these domain d speccific
enggineering conccepts are not independent ffrom each othher;

870
Authorized licensed use limited to: Otto-von-Guericke Universitaet Magdeburg. Downloaded on February 01,2022 at 12:31:48 UTC from IEEE Xplore. Restrictions apply.
 moreover, it is possible tto find a conncept represennted in the AutomationM ML DSL of tthe data sink and exports the
several discipplines and toolls. These com
mmon conceptss can be dataa from the datta logistics. Seecond, relevannt data is seleccted
modelled in the AutomattionML overaall language (named ( folllowing the intternal data moodel of the daata sink tool aand
AML1 in [8]) by instantiatting the system m unit classess of the trannsformed by an adapter too the tool-speecific file form mat
domain specific engineeriing concepts as InternalEllements readd by the tool. F
Fig. 5 shows tthese process ssteps.
within one coommon system m unit class (seee Fig. 4). F
For completingg an I4.0AAS, the I4.0AAS S or a tool ablee to
gennerate the I4.00AAS shall bee regarded as a data sink. Thhus
the named structuure of an enginneering data loogistics is a vaalid
canndidate for the automated geeneration of ann I4.0AAS.
B. Design of I44.0AAS Submoodel Candidatees
T
To address R RQ1, this suubsection idenntifies candiddate
subbmodels for ann I4.0AAS.
T
The integratedd engineering objects moddelled within the
enggineering dataa logistics folllowing the ccommon conccept
ideaa described iin Subsectionn IV.A preserrve the differrent
AuttomationML D DSLs (see Fiig. 6). In thiss example, thhree
proprietary (reppresenting fu functional, m mechanical, aand
elecctrical engineeering) and onee standardizedd AutomationM ML
DSL L (representinng automationn engineeringg following [228])
are combined. Each DSL L results in a separrate
InteernalElement w with the domaain-specific atttributes involvved
andd the domain-sspecific role cllasses as semaantic indicatorss.

Fig. 4. Autom
mationML DSLs within engineerinng data logistics,
based on [7], [8].

Exploiting these local AutomationM ML DSLs aand the

AutomationM ML overall lannguage in enggineering dataa logis-
tics can be esstablished in thhe following w
way (see Fig. 55).
A source tool exports engineering data followiing the
discipline- annd/or tool-speecific data moodel. A flexibble tool
adapter utilizing the sysstem unit claasses of the Auto-
mationML D DSLs, as presennted in detail in [34], can trranslate
the exported data from thhe tool-specific data formatt to the Fig. 6. Object reppresentaton withinn the engineering data logistics,
AutomationM ML DSL of the engineeering domainn. The based on [77], [8].
importer of thhe data into thhe data logistiics, maps the data to
C
Comparing this data structurre with the Suubmodel structture
the domain related part of the AuttomationML overall
of the I4.0AAS S shows clearr similarities.. Each Internnal-
language annd, thereby, may merge and aggregaate the
Element reflectiing one AuutomationML DSL can be
imported datta with engiineering data from furtheer data
connsidered as a Submodel witth Submodel elements of tyype
sources. Finaally, collected data forms ann integrated mmodel of
Prooperty. This leeads to the firrst set of Subm
model candidaates
the complete system with aall relevant enggineering dataa.
formmed by the AutomationM ML DSLs andd related to the
enggineering datta models applied withhin engineerring
netwworks, i.e., thee data modelss applied withiin an engineerring
disccipline and daata models appplied to trannsfer engineerring
infoormation betw ween engineeriing discipliness.
P
Plattform Induustry 4.0 also ddiscusses Subm model candidaates
[35]: Nameplate for basic assset descriptionn, Documents for
describing and referring too relevant documents like l
certtificates and m manuals, Servvice for referrring to techniical
servvice to the aasset, Identifification for a detailed, m more
Fig. 5. Enginneering data logisstics, based on [7]] and [8].
purrchase related,, asset descripption, and Devvice-Descriptiion-
Filees for descrribing and rreferring to various devvice
To importt the data to the data sinnk tool, the reverse descriptions to the asset. A All these Subbmodels can be
process appliies. First, an eexporter mapss the relevant data to idenntified as speecific engineeering related disciplines (llike

871
Authorized licensed use limited to: Otto-von-Guericke Universitaet Magdeburg. Downloaded on February 01,2022 at 12:31:48 UTC from IEEE Xplore. Restrictions apply.
 Identificationn and Servicee) or as infoormation sets to be • Add role claass AssetAdmin
nistrationShel
ll
RoleClassLibb/Submodel
exchanged beetween discipllines (like Doccuments and D Device- • Add role-rellated attribut
tes and comple
ete them
Description-F Files). Thus thhey can be inttegrated easilyy in the following [330]
o Attribut
te idShort equ
uals InternalE
Element name
engineering ddata logistics cconcept (see S
Subsection IV.A). (i.e. vi
iew / Automati
ionML DSL name
e)
The cooperrative activitiees of VDMA aand OPC Founndation o …
Stepp3: For each AAttribute of t
the InternalEl
lement
for developinng OPC UA coompanion speecifications rellated to reprresenting a viiew (an Automa
ationML DSL) d
do:
Industry 4.0 ccomponents [336] can be reggarded as thirdd source • Instantiate a new InternaalElement with
h the role
AssetAdminisstrationShellR
Role ClassLib/
/Property
of Submodels. In most casses, they definne an OPC UA A node • Add to the nnew InternalEl
lement the fol
llowing
set applicablee as interface to
t control speccial types of aassets at attributes aand complete t
them
o Attribut
te idShort equ
uals InternalE
Element name
runtime. Thiss node set cann be seen as R Runtime DSL for the o Assign value
v equal to
o the value of
f the attribut
te
related asset type that haas to be handdled also withhin the o …
END
engineering by mapping the OPC U UA data poiints to
U
Using Algoritthm 1 to trannslate the enngineering objject
engineering oobjects. Thus OPC UA datta points can also be
reprresented in F Fig. 6 leads tto an I4.0AASS AutomationM
ML
regarded as a data sink of the engineerinng data logistiics (see
seriialization as ddepicted in Figg. 7.
Subsection IV V.A).
In summarry, valid canddidates for thhe Submodels within
I4.0AAS are formed by thhe AutomationnML DSLs rellated to
the engineerring data moodels applied within enginneering
networks, i.e.., the data moddels applied w
within an enginneering
discipline annd data moddels that are applied to ttransfer
engineering information bbetween engiineering disciplines.
Hence, if the system integrrator has to deecide on the daata sets
to be integratted in an AAS he has to consider all relkevant
AutomationM ML DSLs aand select bbased on thee data
application annd exchange use
u cases of innterest.
C. Improveed Engineeringg Data Logistiics – Serializaation Fig. 7. I4.0AAS AutomationML
A seerialization exam
mple
To answeer RQ2 the necessary adaptations to the
D. AutomationM ML based I4.00AAS completiion method
Subsection IV.A need
engineering ddata logistics ddescribed in S
to be identifieed and detailedd. To reach the intended AuutomationML based I4.0A
T AAS
Fig. 6 shoows how thee engineeringg data logistiics can commpletion methhod the descriibed buildingg blocks shall be
provide AutoomationML daata models, w where each Subbmodel appplied by system
m integrators iin the followinng way.
is representedd by an AutoomationML D DSL related Innternal- Inn a first sttep they havve to identify the relevvant
Element. Com mparing this sstructure withh the AutomattionML AuttomationML D DSLs to be inntegrated withiin the I4.0AA
ASs.
guideline [330], there iis a need to transform m the Forr each of themm the related engineering tools need to be
AutomationM ML data m model towaards the I4 I4.0AAS inteegrated withinn the AutomaationML baseed data logistics
AutomationM ML serializatioon. This call ffor an export adapter described in [377], [7], [8] neext by develooping the relaated
from the enggineering dataa logistics thaat acts as automated apppropriate AutoomationML roole and interfaace class librarries
I4.0AAS gennerator (see Figg. 5: adapter inn red color). andd system unitt class librarries as well as the requiired
This adaptter shall add the necessaryy role and innterface adaapters (see Fiig.5). As third step an aadapter realizzing
classes defined in the AuttomationML gguideline [30]] to the Alggorithm 1 neeed to be impleemented. If alll three steps are
AutomationM ML elements and translate the AutomattionML donne the configgured and exxtended AutoomationML ddata
attributes to IInternalElemeents with the role Property with a logiistics can directly
d be applied to create I4.0A AAS
special attribuute structure. Algorithm 1 ddescribes the pprocess AuttomationML sserializations.
for Submodeelelements off type Propeerty. For alll other V. EVALLUATION
Submodelelem ment types thee process is aanalogous. Thiis leads
to the I4.0AAAS generation algorithm (seee Algorithm 1). T
To evaluate the appliccability of the presennted
AuttomationML based I4.0A AAS complettion method, a
Algorithm 1: I4.00AAS generationn from engineeringg data logistics. feassibility studyy was conduucted for thhe use case of
gennerating I4.0 asset
a administtration shells for an ultrasoonic
Step1: For eaach InternalEl
lement represe
enting an asse
et
administratioon shell do: meaasurement cell and its compponents, to be applied in a stteel
• Add rolee class AssetA
Administration
nShell mill. Therefore, tthe engineerinng network of a leading systtem
RoleClasssLib/AssetAdm
ministrationSh
hell
• Add rolee-related attr
ributes and co
omplete them inteegrator of steel mills has beeen analyzed annd reproducedd on
followinng [30] lab scale, reflectiing the engineeering of steel mill componeents
o Attri
ibute idShort equals Intern
nalElement nam
me
o … trannslated the to a measuremennt cell. Withinn this replicatiion,
Step2: For eaach InternalEl enting a view (an
lement represe enggineering artiffacts coming from the dom mains functioonal
AutomationML DSL) do:
enggineering, mecchanical enginneering, electrrical engineeriing,

872
Authorized licensed use limited to: Otto-von-Guericke Universitaet Magdeburg. Downloaded on February 01,2022 at 12:31:48 UTC from IEEE Xplore. Restrictions apply.
 and control engineering have been aggregated into one issues for research. This paper made a first step in this
engineering logistics as depicted in Fig. 7. research direction.
The feasibility study started with the consideration of the This paper presented an extension of an existing
existing engineering data logistics and the discussion of the engineering data logistics following [7] and [8] as foundation
applicability of the different domain-specific and domain- for combining production system engineering tools in an
crossing engineering data models involved in this enginee- engineering network using AutomationML based domain-
ring data logistics as candidates for representation in I4.0 specific languages. Enhancing this engineering logistics with
AAS submodels. In discussions with domain experts, it a special engineering adapter will result in a capability to
became obvious that the data models represented in the automatically generate an I4.0AAS AutomationML
engineering data logistics can be an efficient starting point serialization following the current standardization [12].
for I4.0AAS generation as they represent all engineering In addition, this paper identified the AutomationML DSLs
information required by data sinks within engineering and related to the engineering data models applied within
run time of a CPPS. This confirmation validated the engineering networks as candidates for specifying the
fulfilment of RQ1 by the AutomationML based I4.0AAS relevant Submodels as structural parts of the I4.0AAS. The
completion method. Nevertheless, it is not guaranteed that nature of these data models is not fixed. They may range
the existing data exchange capabilities are complete. from requirement over function, mechanical, and electrical,
In the second the part of the feasibility study, a software to control engineering and may also contain simulations and
component, which implemented the I4.0AAS generation of much more.
Algorithm 1, exported I4.0AAS AutomationML seriali- Both results together result in the presented
zations from the engineering data logistics for a complete AutomationML based I4.0AAS completion method, enabling
measurement cell, a drive chain within this cell, and different a step towards the simplification of the practical use of
automation devices. Checking the result of the serialization I4.0AAS within production system automation architectures
confirmed that Algorithm 1 enables translating engineering by avoiding the, still necessary, manual model design.
data collected in an AutomationML based engineering data Main benefit of this approach is its simplicity and focus on
logistics, following [7] and [8], can be effectively and standardized technologies and domain specific languages,
efficiently translated to I4.0AAS AutomationML serializa- while its main drawback is its dependency on
tions fulfilling RQ2. AutomationML. Facing the increasing importance of
Reflecting the results of the feasibility study, the require- AutomationML within engineering data exchange and the
ments to the AutomationML based I4.0AAS completion increasing standardization of AutomationML DSLs, the cost
method listed in the introduction can be evaluated. The focus of the drawback will diminish in the near future.
on the data models reflected within the engineering data The presented results related to the applicability of Auto-
logistics and, thereby, to the applied AutomationML DSLs mationML DSLs within this approach will call for new stan-
strongly simplifies for the system integrator deciding on dardization initiatives to be discussed within the Industry 4.0
relevant data set to be represented in an I4.0AAS. Especially community. It becomes more and more critical to identify
as they these AutomationML DSLs represent standardized the most relevant engineering disciplines and data models
needs of engineering data sinks within the engineering applied within the different production system life cycle
network. Thus Req.1 is fulfilled. The applied phases and to develop AutomationML DSLs based on them.
AutomationML based engineering data logistic The AutomationML standard is able to support the consis-
automatically integrates the engineering data transmitted tent development of such domain specific languages [27].
while the exporter implementing Algorithm 1 automatically Moreover the topic of engineering data security becomes
translates the engineering data to an I4.0AAS AutomationML more and more critical.
serialization. Only the required data logistic integration Finally, the authors of this paper will develop support
structure needs to be parameterized to the intended use case methodologies for interested system integrators assisting
of engineering data. Thus also Req.2 and Req.3 are fulfilled. them within the implementation / application of the
presented approach.
VI. CONCLUSION AND FUTURE WORK
An essential part of the intended digitalization of ACKNOWLEDGMENT
production systems is the Industry 4.0 component with its The financial support by the Christian Doppler Research
asset administration shell. Especially the Industry 4.0 asset Association, the Austrian Federal Ministry for Digital and
administration shell (I4.0AAS) [4] is seen as the main Economic Affairs and the National Foundation for Research,
carrier of information required to describe, access, and Technology and Development is gratefully acknowledged.
control an Industry 4.0 component.
REFERENCES
While the structure, behavior, and software imple-
[1] acatech, “Recommendations for implementing the strategic initiative
mentation of the I4.0AAS are well developed, the definition INDUSTRIE 4.0,” April 2013, Report, http://www.plattform-
of the content for an I4.0AAS and the methodology to col- i40.de/sites/default/files/Report_Industrie%204.0_engl_1.pdf.
lect and provide this content to an I4.0AAS are still open

873
Authorized licensed use limited to: Otto-von-Guericke Universitaet Magdeburg. Downloaded on February 01,2022 at 12:31:48 UTC from IEEE Xplore. Restrictions apply.
 [2] C. Manzei, L. Schleupner, R. Heinze: Industrie 4.0 im internationalen Engineering Network Challenges, 4th AutomationML user conference,
Kontext, VDE Verlag, 2016 (in German). Oct. 2018, Gothenburg, Sweden, Proceedings, www.automationml.org.
[3] S. Biffl, D. Gerhard, A. Lüder: Multi-Disciplinary Engineering for [21] S. Biffl, M. Eckhart, A. Lüder, T. Müller, F. Rinker, and D. Winkler.
Cyber-Physical Production Systems, Springer, 2017. 2018. “Data Interface for Coil Car Simulation (Case Study) Part II -
[4] Plattform Industrie 4.0: Details of the Asset Administration Shell - Part Detailed Data and Process Models.” Technical report CDL-SQI-M2-
1 - The exchange of information between partners in the value chain of TR03. TU Wien.
Industrie 4.0, https://www.plattform-i40.de/I40/Redaktion/DE/ Down- [22] A. Lüder, N. Schmidt, K. Hell, H. Röpke, J. Zawisza: Description
loads/Publikation/2018-verwaltungsschale-im-detail.html, April 2019. means for information artifacts throughout the life cycle of CPPS, In:
[5] DIN Spec 91345: Reference Architecture Model Industry 4.0 (RAMI Multi-disciplinary engineering for cyber-physical production systems:
4.0), Beuth publisher, April 2016. data models and software solutions for handling complex engineering
[6] Wagner C., Grothoff J., Epple U., Drath R., Malakuti S., Grüner S., projects - Cham: Springer-Verlag, S. 169-183, 2017
Hoffmeister M., Zimmermann P.: The role of the Industry 4.0 Asset [23] P.P.S. Chen: The entity-relationship model—toward a unified view of
Administration Shell and the Digital Twin during the life cycle of a data. ACM Transact. on Database Systems (TODS), 1(1), 1976, pp.9-
plant. In: Proceedings of the 2017 IEEE Conference on Emerging 36.
Technologies Factory Automation (ETFA 2017), Limassol, Cyprus, [24] J. Rumbaugh, I. Jacobson, G. Booch: Unified modeling language
2017. reference manual, the. Pearson Higher Education, 2004.
[7] A. Lüder, K. Kirchheim, J.-L. Pauly, S. Biffl, F. Rinker, and L. [25] L Delligatti: SysML distilled: A brief guide to the systems modeling
Waltersdorfer: Formalizing the information logistics within engineering language, Pearson Education, 2014.
networks, IEEE 17th International Conference on Industrial Informatics [26] R. Drath, M. Barth, A. Fay: Offenheitsmetrik für Engineering-
(INDIN), June 2019, Helsinki, Finland, Proceedings. Werkzeuge: Die Fähigkeit zur Interoperabilität bewerten, atp-edition,
[8] A. Lüder, J.-L. Pauly, F. Rinker, and S. Biffl: Data Exchange Logistics Heft 9/2012, S. 46-55.
in Engineering Networks Exploiting Automated Data Integration, 24th [27] A. Lüder, N. Schmidt, R. Drath: Standardized information exchange
IEEE Conference on Emerging Technologies and Factory Automation, within production system engineering, In: Multi-disciplinary
Sep. 2019, Zaragoza, Spain, Proceedings. engineering for cyber-physical production systems: data models and
[9] DIN Spec 91345: Reference Architecture Model Industry 4.0 (RAMI software solutions for handling complex engineering projects - Cham:
4.0), Beuth publisher, April 2016. Springer-Verlag, S. 235-257, 2017
[10] Ministry of Economy and Finances (France), Alliance Industrie du [28] AutomationML association: Application Recommendation AR 001E –
Futur (Italy), Federal Ministry for Economic Affairs and Energy Automation Project Configuration, 2018,
(BMWi) and Plattform Industrie 4.0 (Germany): The Structure of the https://www.automationml.org/o.red/uploads/dateien/1542365565-
Administration Shell: TRILATERAL PERSPECTIVES from France, ApplicationRecommendation_AutomationProjectConfiguration_V1.1.0
Italy and Germany, 2018, https://www.plattform-i40.de/PI40/ .zip
Redaktion/DE/Downloads/Publikation/hm-2018-trilaterale-coop.html [29] AutomationML association: Application Recommendation AR 003E:
[11] Oztemel, E., Gursev. S.: Literature review of Industry 4.0 and related Provisioning for MES and ERP – Support for IEC 62264 and B2MML,
technologies, J Intell Manuf 31, 127–182 (2020). 2018, https://www.automationml.org/o.red/uploads/dateien/
https://doi.org/10.1007/s10845-018-1433-8. 1542365399-AR_MES_ERP-1.1.0.zip.
[12] Plattform Industrie 4.0: Details of the Asset Administration Shell - Part [30] AutomationML association: Application Recommendation - Asset
1 - The exchange of information between partners in the value chain of Administration Shell Representation (AR AAS), 2019,
Industrie 4.0 (Version 2.0), https://www.plattform- i40.de/PI40/ https://www.automationml.org/o.red/uploads/dateien/1574258013-
Redaktion/DE/Downloads/Publikation/Details-of-the-Asset- AR_004E - Assset Administration Shell Representation 20V1_0_0.zip
Administration-Shell-Part1.html, November 2019. [31] AutomationML association: Rationals of Automation component
[13] A. Lüder, M. Foehr, L. Hundt, M. Hoffmann, Y. Langer, St. Frank: classification, 2020, https://www.automationml.org/o.red/uploads/
Aggregation of engineering processes regarding the mechatronic dateien/ 1575989332-WP Compo – AutomationComponent
approach, 16th IEEE International Conference on Emerging Library_v1_0_0.zip
Technologies and Factory Automation (ETFA 2011), Toulouse, France, [32] S. Biffl, A. Lüder, F. Rinker, L. Waltersdorfer, and D. Winkler: Quality
September 2011, Proceedings-CD. Risks in the Data Exchange Process for Collaborative CPPS
[14] A. Strahilow, H. Hämmerle: Engineering Workflow and Software Tool Engineering, IEEE 17th International Conference on Industrial
Chains of Automated Production Systems, In: Multi-disciplinary Informatics (INDIN), June 2019, Helsinki, Finland, Proceedings.
engineering for cyber-physical production systems: data models and [33] F. Ekaputra: Ontology-based data integration and knowledge change
software solutions for handling complex engineering projects, Springer management in multi-disciplinary engineering environments,
Publisher , pp 207-234, 2017. Technische Universität Wien, Dissertation, 2018,
[15] K. Paetzold: Product and Systems Engineering/CA* Tool Chains, In: https://repositum.tuwien.ac.at/obvutwhs/download/pdf/3078031?origin
Multi-disciplinary engineering for cyber-physical production systems: alFilename=true
data models and software solutions for handling complex engineering [34] A. Lüder: Designing tool chains based on common concepts - a data
projects, Springer Publisher , pp 27-62 integration approach – an advanced workshop, 5th AutomationML
[16] M.R. Mehr, S. MirRashed, M. Mißler-Behr, A. Lüder: Analyzing the PlugFest, Proceedings, https://www.automationml.org/o.red/uploads/
causes and effects of complexity on different levels of automobile dateien/1579772563-5th AutomationML Plugfest Presentations.zip
manufacturing systems, 21th IEEE Conference on Emerging [35] Platform Industry 4.0: Standard repository of information for the Asset
Technologies and Factory Automation (ETFA), Sep. 2016, Berlin, Administration Shell, 2020, https://github.com/admin-shell/io.
Germany, Proceedings. [36] VDMA: OPC UA companion specifications, 2020,
[17] A. Lüder , N. Schmidt: AutomationML in a Nutshell, in B. Vogel- https://opcua.vdma.org/companion-specifications#catalogs
Heuser, T. Bauernhansl, M. ten Hompel (Ed.): Handbuch Industrie 4.0, [37] A. Lüder, J-L. Pauly, K. Kirchheim, F. Rinker, S. Biffl: Migration to
Bände 1-4, VDI Springer Reference, 2020. AutomationML based Tool Chains –incrementally overcoming
[18] F. Himmler, M. Amberg: Data Integration Framework for Engineering Network Challenges, 5th AutomationML user conference,
Heterogeneous System Landscapes within the Digital Factory Domain, Gothenburg/Sweden, October 2018, Conference proceedings,
Procedia Engineering, Volume 69, 2014, pp. 1138-1143. https://www.automationml.org/o.red/uploads/dateien/1548668540-
[19] R. Mordinyi, S. Biffl: Versioning in cyber-physical production system 17_Lueder_Migration-ToolChains_Paper.pdf
engineering: best-practice and research agenda, Proceedings of the First [38] International Electrotechnical Commission: IEC 62890 - Life-cycle
International Workshop on Software Engineering for Smart Cyber- management for systems and products used in industrial-process
Physical Systems, Florence, Italy, pp. 44-47, 2015. measurement, control and automation, Beuth publisher, April 2017.
[20] A. Lüder1, J-L. Pauly, K. Kirchheim, F. Rinker, S. Biffl: Migration to [39] Verein Deutscher Ingenieure. Richtlinie 3695: Engineering von
AutomationML based Tool Chains – incrementally overcoming Anlagen - Evaluieren und Optimieren des Engineerings. VDI-Verlag,
Düsseldorf, 2009.

874
Authorized licensed use limited to: Otto-von-Guericke Universitaet Magdeburg. Downloaded on February 01,2022 at 12:31:48 UTC from IEEE Xplore. Restrictions apply.
View publication stats