=Paper=
{{Paper
|id=Vol-2956/paper42
|storemode=property
|title=Automatically Linking Concepts in Distributed, Cloud-Based Manufacturing Environments
|pdfUrl=https://ceur-ws.org/Vol-2956/paper42.pdf
|volume=Vol-2956
|authors=Damian Drexel
|dblpUrl=https://dblp.org/rec/conf/ruleml/Drexel21
}}
==Automatically Linking Concepts in Distributed, Cloud-Based Manufacturing Environments==
https://ceur-ws.org/Vol-2956/paper42.pdf
Automatically Linking Concepts in Distributed,
Cloud-Based Manufacturing Environments
Damian Drexel[0000−0002−0987−0067]
Research Center Digital Factory Vorarlberg, Vorarlberg University of Applied
Sciences, Hochschulstr. 1, 6850 Dornbirn, Austria
damian.drexel@fhv.at
Abstract. With the digitalisation, and the increased connectivity be-
tween manufacturing systems emerging in this context, manufacturing
is shifting towards decentralised, distributed concepts. Still, for manu-
facturing scenarios manual input or augmentation of data is required
at system boundaries. Especially in distributed manufacturing environ-
ments, like Cloud Manufacturing (CMfg) systems, constant changes to
the available manufacturing resources and products pose challenges for
establishing connections between them. We propose a feature-oriented
representation of concepts, especially from the manufacturing domain,
which serves as the basis for (semi-) automatically linking, e.g., manu-
facturing resources and products. This linking methodologies, as well as
knowledge inferred using it, is then used to support distributed manu-
facturing, especially in CMfg environments, and enhance product devel-
opment. The concepts and methodologies are to be evaluated in a real
world learning factory.
Keywords: Manufacturing · Cloud Manufacturing · Distributed Man-
ufacturing · Reasoning · Ontologies · Feature-Based
1 Problem Statement
Emerging technologies and concepts enabled various changes for the manufactur-
ing domain during the recent years. Initiatives, such as Industrie 4.0, advance the
digitalisation of manufacturing and simplify the distribution of manufacturing
tasks [3, 15]. With increasing connectivity, systems are shifting from centralised,
monolithic to decentralised, modular applications [5].
Systems such as CMfg platforms realise this distribution of manufacturing
tasks. These platforms can be employed by a single manufacturing company
or can integrate manufacturing resources of various, independent companies. In
both cases the information and knowledge managed within these CMfg plat-
forms is highly dynamic. New manufacturing resources, as well as products, are
constantly being added to or removed from the plattform.
Copyright © 2021 for this paper by its authors. Use permitted under Creative
Commons License Attribution 4.0 International (CC BY 4.0).
A representation of knowledge in CMfg is required to describe how a product
can be manufactured using the available resources. Knowledge representations
in manufacturing, like the MAnufacturing’s Semantics ONtology (MASON) [13],
model such connections in different ways, e.g. by describing the manufacturing
process for a product using a combination of resources. In case of hardly dy-
namic manufacturing systems or when products and manufacturing resources
are defined by the same authority these links can be established manually. In a
highly distributed and dynamic system, such as a CMfg platform, the manual
creation of these links may not be viable when a new manufacturing resource
connects to the platform or a new product is added.
The aim of the PhD thesis described in this paper is to establish links between
a product and the manufacturing resources required to produce it, based on a
fitting knowledge representation and by employing reasoning as well as matching
mechanisms. Therefore, major challenges are (1) the creation of a feature-based
representation of products and manufacturing resources, (2) implementing (semi-
) automatic linking mechanisms, and using these representation and mechanisms
to support (3) distributed manufacturing as well as (4) product design.
2 Related Work
The description of the manufacturing domain comprises a research topic present
since multiple decades [2]. Widely used technologies for describing manufactur-
ing resources and for enabling the communication between them, are, for exam-
ple, Open Platform Communication Unified Architecture (OPC UA) and MT-
Connect [4, 9]. In the manufacturing domain, ontologies are commonly used for
knowledge representation. An ontology is “an explicit specification of a concep-
tualization” [8] and enables the integration and interoperability of systems [13,
23]. Topics in modeling knowledge in manufacturing include (1) modeling of
manufacturing resources, (2) modeling of manufacturing systems, (3) modeling
of manufacturing processes, and (4) the adoption of foundational ontologies,
besides others [16].
One way to describe products in a manufacturing setting is by specifying their
parts and the relation between each part [10, 19, 21, 23]. This form of represen-
tation builds upon the hierarchic structure, which is often inherent to complex
systems [20]. Other representations of products in manufacturing use features
to describe them. For example, when designing products tools, like Computer
Aided Design (CAD) software, define objects by their features [1]. Generally
speaking, features can be components of a thing as well as concrete or abstract
properties of it [22]. Analogously, in product design features carry information
about the geometric or physical nature of the product as well as manufacturing
and life-cycle related information [16, 6].
Järvenpää et al. use ontologies to match products and manufacturing re-
sources [11]. The authors define product requirements and manufacturing re-
source capabilities to find matching (combinations of) resources. Manufacturing
tools are an alternative to capabilities for describing what a manufacturing re-
source is able to do. The process of manufacturing a product can be described
by defining on the one hand what tools are required for an operation and on the
other hand what operations are required to machine a raw material to achieve
a geometric entity [13]. Independently of whether capabilities, tools, or other
concepts are used to model a manufacturing resource’s skills, connecting manu-
facturing resources to products by describing the process is common, e.g., as in
the reference ontology for manufacturing proposed by Usman et al. [23].
In contrast to hardly changing manufacturing environments, which are often
considered when developing models like the ontologies described before, in CMfg
products may be defined independently of manufacturing resources. A field with
similar challenges are web services where definitions of web services and require-
ments by service requesters are defined independently [24]. Describing services
has been explored extensively for web services [2]. Here, technologies like the
Web Ontology Language for Web Services (OWL-S) and the Web Service Mod-
eling Ontology (WSMO) are commonly used [12]. Those technologies, especially
OWL-S, are also considered for matching manufacturing resources [17].
A different approach for matching web services, but also for other ontol-
ogy related applications, is ontology matching [7]. Differences in meaning when
representing knowledge, i.e. the semantic heterogeneity problem, is a problem
that can be tackled using this approach [18]. Zhdanova and Shvaiko describe
an approach enhancing the process of ontology matching by elevating it to a
community-driven activity [25].
3 Research Questions
The motivation for the work to be done in this thesis stems from the manu-
facturing domain, especially distributed manufacturing. The research question,
that forms the basis of the thesis, can be formulated in a general way as follows.
RQ How can feature based specifications of concepts representing different views
of the same domain, be automatically linked together by means of reasoning and
classification?
4 Hypothesis
The hypothesis related to the research question formulated before are
– Products as well as manufacturing resources described in a hierarchic, feature-
based way are a suitable representation in distributed manufacturing.
– Reasoning and classification methods can be used on such a representation,
to link products to the manufacturing resources required to produce them.
– The methods to be developed in this thesis are of use for various manufac-
turing related tasks, e.g., product design and production planning.
5 Approach
The goal of the thesis is the exploration and advancement of methods to automat-
ically link concepts based on their features, i.e., to describe which combination
of manufacturing resources are able to create a given product. The developed
methods are meant to be applied to the manufacturing domain to support its
processes, especially in distributed, cloud-based manufacturing environments.
A representation of knowledge about manufacturing resources and products,
based on existing ontologies and standards, is to be created as a foundation for
the other work to be done in this thesis, i.e., automatic linking of concepts. In
order to (semi-) automatically create the connections between manufacturing
resources and products, reasoning and matching mechanisms are to be used.
Utilizing the knowledge inferred this way, combinations of resources are defined,
which are able to produce specific products.
The concepts and technologies used for (semi-) automatically linking a prod-
uct to the manufacturing resources required to produce it are then to be applied
to different fields in the manufacturing domain. The main focus is to support
distributed manufacturing systems, like CMfg platforms. The developed repre-
sentation of manufacturing resources and products are used to bring together
customers and manufactures by finding fitting manufacturing resources in an ever
changing environment. During the design phase of a product, possible manufac-
turing processes can be dynamically created, based on the information derived
from these links, while the designer is still working on the product.
Since a fully automated linking of concepts may not result in a useable on-
tology, the use of manual input has to be considered, as well. In order to include
domain experts into the process of creating these links, a community-driven ap-
proach, similar to the one described in [25], could be applied and used to improve
the overall result.
The main aspects of the thesis are:
– Representation of products and manufacturing resources
– (Semi-) Automatic linking
– Support of distributed manufacturing
– Support of product design
– Real world validation
According to these main aspects, tasks can be identified.
– Development of a feature based representation. The foundation of the
work to be done in this thesis is a representation of knowledge about prod-
ucts and manufacturing resources. For this representation existing models
and standards have to be considered. A first feature-based model may be
developed with a still manual process of linking concepts in mind.
– Automation of the linking process. Using the representation developed
during the first phase of the thesis, the process of linking products and
manufacturing resources is automated. Here the degree of automation is an
important factor. Since a full automation may not be feasible, the possibility
of manual steps has to be explored.
– Manufacturing process creation. Linking the concepts in a (semi-) au-
tomatic way already allows to infer assumptions about whether a product
can be produced or not. This knowledge is then to be used to create a man-
ufacturing process. Additional considerations, e.g. geographic locations of
manufacturing resources, have to be taken into account. Implementing the
creation of manufacturing processes enables (semi-) automatic distributed
manufacturing.
– Support of product design. The model and concepts to be developed
can be used to support product design, as well. To accomplish this the in-
formation covered in CAD models has to be brought into the knowledge
representation introduced in this work. Then it can be deduced whether or
not the product currently designed can be produced with the existing man-
ufacturing resources. The inferred knowledge is then to be returned to the
CAD application in order to give useful feedback to the product designer.
– Real world validation. The methods and knowledge representations to
be developed in this thesis are to be validated in a real world setting. The
model factory of the Digital Factory Vorarlberg provides an environment for
validating the work [14]. The manufacturing process modelled by the model
factory includes a heterogenus set of manufacturing resources, like mills,
transportation systems, and assembly resources. It is one of multiple labora-
tories connected for distributed manufacturing. This is enabled by a CMfg
platform which connects different laboratories. This setup includes various
aspects of distributed manufacturing and offers a real world environment for
the validation of the work in this thesis.
6 Conclusion & Future Work
In recent years concepts and methods, especially related to digitalisation, en-
abled the rise of distributed manufacturing and consequently the distribution of
manufacturing tasks to different manufacturing providers. Applications in this
field, such as CMfg platforms, have to establish connections between products
and manufacturing resources required to produce them. This paper outlines how
such a (semi-) automatic linking can be established and gives a roadmap for the
proposed thesis. While similar work often connects products and manufacturing
resources by the manufacturing process itself, we propose automatically gener-
ating feasible manufacturing processes based on systematically derived connec-
tions of products and manufacturing resources. This not only enables distributed
manufacturing by supporting the constantly changing, heterogenus manufactur-
ing environment that follows from this paradigm, but is also able to support
other manufacturing related processes, such as product design. The work de-
scribed here is still in a very early stage and hence no major preliminary results
are being reported in this paper. Nevertheless, with the tasks outlined in this
paper we aim to create an appropriate representation as a basis for (semi-) au-
tomatically linking concepts from the manufacturing domain. For evaluating the
results of the work to be done in this thesis, the representations and method-
ologies developed are to be integrated into a learning factory, offering real world
applications, as well as various manufacturing resources.
Acknowledgment. The research presented in this paper is partially financed
by FFG-Project No. 866833 “CIDOP”.
References
1. Adamson, G., Wang, L., Moore, P.: Feature-based control and informa-
tion framework for adaptive and distributed manufacturing in cyber phys-
ical systems. Journal of Manufacturing Systems 43, 305–315 (apr 2017).
https://doi.org/10.1016/j.jmsy.2016.12.003
2. Ameri, F., Dutta, D.: An upper ontology for manufacturing service description.
In: International Design Engineering Technical Conferences and Computers and
Information in Engineering Conference. vol. 42578, pp. 651–661 (2006)
3. Anderson, N., Diab, W.W., French, T., Harper, K.E., Lin, S.W., Nair, D., Sobel,
W.: The Industrial Internet of ThingsVolume T3: Analytics Framework. Tech. rep.,
Industrial Internet Consortium (2017)
4. Bauernhansl, T., ten Hompel, M., Vogel-Heuser, B. (eds.): Industrie 4.0 in Pro-
duktion, Automatisierung und Logistik. Springer Fachmedien Wiesbaden (2014).
https://doi.org/10.1007/978-3-658-04682-8
5. Burger, A., Lang, A., Müller, Y.: Mögliche Veränderungen von System-
Architekturen im Bereich der Produktion. In: Industrie 4.0, pp. 57–68. Springer
Fachmedien Wiesbaden (2017). https://doi.org/10.1007/978-3-658-15557-5 4
6. Case, K., Gao, J., Gindy, N.: Feature validation in a feature-based design sys-
tem. ‘Advances in Manufacturing Technology VII ‘ the Proceedings of the Ninth
National Conference on Manufacturing Research pp. 272–276 (1993)
7. Euzenat, J., Shvaiko, P.: Ontology Matching. Springer Berlin Heidelberg (2013).
https://doi.org/10.1007/978-3-642-38721-0
8. Gruber, T.R.: A translation approach to portable ontology spec-
ifications. Knowledge Acquisition 5(2), 199–220 (jun 1993).
https://doi.org/10.1006/knac.1993.1008
9. Jasperneite, J., Neumann, A., Pethig, F.: OPC UA versus MTConnect. Com-
puter&Automation (Sonderheft S2 2015 Control&Drives) S pp. 16–21 (2015)
10. Jørgensen, K.A., Petersen, T.D., Nielsen, K., Habib, T.: Product family modelling
for manufacturing planning. In: Proceedings of the 21th International Conference
on Production Research, ICPR 2011, 1.-4. August, Stuttgart, Germany. Fraunhofer
Verlag (2011)
11. Järvenpää, E., Siltala, N., Hylli, O., Lanz, M.: The development of an ontology for
describing the capabilities of manufacturing resources. Journal of Intelligent Man-
ufacturing 30(2), 959–978 (jun 2018). https://doi.org/10.1007/s10845-018-1427-6
12. Lara, R., Roman, D., Polleres, A., Fensel, D.: A Conceptual Comparison of WSMO
and OWL-S. In: Web Services, pp. 254–269. Springer Berlin Heidelberg (2004).
https://doi.org/10.1007/978-3-540-30209-4 19
13. Lemaignan, S., Siadat, A., Dantan, J.Y., Semenenko, A.: MASON: A Proposal
For An Ontology Of Manufacturing Domain. In: IEEE Workshop on Distributed
Intelligent Systems: Collective Intelligence and Its Applications (DIS'06). IEEE
(2006). https://doi.org/10.1109/dis.2006.48
14. Merz, R., Hoch, R., Drexel, D.: A cloud-based research and learning fac-
tory for industrial production. Procedia Manufacturing 45, 215–221 (2020).
https://doi.org/10.1016/j.promfg.2020.04.097
15. Müller, J.M., Voigt, K.I.: Sustainable Industrial Value Creation in SMEs: A Com-
parison between Industry 4.0 and Made in China 2025. International Journal of
Precision Engineering and Manufacturing-Green Technology 5(5), 659–670 (oct
2018). https://doi.org/10.1007/s40684-018-0056-z
16. Sanfilippo, E.M., Borgo, S.: What are features? an ontology-based re-
view of the literature. Computer-Aided Design 80, 9–18 (nov 2016).
https://doi.org/10.1016/j.cad.2016.07.001
17. Sheng, B., Zhang, C., Yin, X., Lu, Q., Cheng, Y., Xiao, T., Liu, H.: Common
intelligent semantic matching engines of cloud manufacturing service based on
OWL-S. The International Journal of Advanced Manufacturing Technology 84(1-
4), 103–118 (nov 2015). https://doi.org/10.1007/s00170-015-7996-z
18. Shvaiko, P., Euzenat, J.: Ontology Matching: State of the Art and Future Chal-
lenges. IEEE Transactions on Knowledge and Data Engineering 25(1), 158–176
(jan 2013). https://doi.org/10.1109/tkde.2011.253
19. Silverston, L.: The Data Model Resource Book, Volume 1. John Wiley & Sons Inc
(Apr 2001)
20. Simon, H.A.: The Architecture of Complexity. Proceedings of the American Philo-
sophical Society 106(6), 467–482 (Dec 1962)
21. Soininen, T., Tiihonen, J., Männistö, T., Sulonen, R.: Towards a general ontology
of configuration. Ai Edam 12(4), 357–372 (1998)
22. Tversky, A.: Features of similarity. Psychological review 84(4), 327 (1977)
23. Usman, Z., Young, R.I.M., Chungoora, N., Palmer, C., Case, K., Hard-
ing, J.A.: Towards a formal manufacturing reference ontology. Interna-
tional Journal of Production Research 51(22), 6553–6572 (nov 2013).
https://doi.org/10.1080/00207543.2013.801570
24. Wu, Z., Ranabahu, A.H., Gomadam, K., Sheth, A.P., Miller, J.A.: Automatic com-
position of semantic web services using process and data mediation. Tech. rep.,
Kno.e.sis Center, Wright State University (Feb 2007)
25. Zhdanova, A.V., Shvaiko, P.: Community-driven ontology matching. In: European
Semantic Web Conference. pp. 34–49. Springer (2006)