Vol-3194/paper5
Jump to navigation
Jump to search
Paper
| Paper | |
|---|---|
 edit
| |
| description | |
| id | Vol-3194/paper5 |
| wikidataid | Q117344903→Q117344903 |
| title | A Multi-Perspective Data Model for Cyber Physical Production Networks |
| pdfUrl | https://ceur-ws.org/Vol-3194/paper5.pdf |
| dblpUrl | https://dblp.org/rec/conf/sebd/BagoziBR22 |
| volume | Vol-3194→Vol-3194 |
| session | → |
A Multi-Perspective Data Model for Cyber Physical Production Networks
A Multi-Perspective Data Model for Cyber Physical
Production Networks
Ada Bagozi1 , Devis Bianchini1 and Anisa Rula1
1
University of Brescia, Brescia, Italy
Abstract
Recently, the research on data management is moving towards the design of data models for Cyber
Physical Production Networks. They constitute ecosystems where cyber and engineered physical
elements record data (e.g., using sensors), analyse them using connected services (e.g., over cloud
computing infrastructures) and interact with human actors using multi-channel interfaces, going beyond
the boundaries of a single enterprise and spanning over the whole production network. In this paper, we
propose a conceptual data model that relates the digital counterpart of a product with data collected over
the phases of the product lifecycle, with special attention on the manufacturing process, and with data
gathered to monitor the shop floor machines used during the production. This enables the development
of multi-perspective data services that implement advanced business goals at the production network
level, such as production scheduling, energy efficiency, product and process monitoring. We provide an
architecture of the proposed solution and its preliminary validation in an ongoing industrial research
project.
Keywords
Multi-perspective data model, service-oriented architectures, Cyber Physical Production Network, Indus-
try 4.0
1. Introduction
The Industry 4.0 revolution is shifting the innovation focus from traditional crafts and engineer-
ing, such as mechanical engineering, material science and plants construction, to immaterial
assets, like knowledge extraction and data analysis and exploration. The latter ones are still
mainly focused on the design of Digital Twins of machines or parts in isolation, as virtual
representations of physical assets in a Cyber-Physical System. Recently, research on Digital
Threads is gaining momentum, conceived as the cyber side representation of a product and the
context where the product is designed, produced, used and maintained, to enable the holistic
view and traceability along its entire lifecycle [1, 2].
However, this vision is focused on the digital representation of the product only, while its
full potential is still far from the sphere of awareness. Heterogeneous and not interconnected
data comes from the shop floor level, all along the product lifecycle, to ensure the production of
costly and complex products, high product quality levels, long lasting functioning, less frequent
and efficient maintenance activities and scalable performance over time. A pressing demand
has been raised to process and integrate those data by exploiting service-oriented architectures
SEBD 2022: The 30th Italian Symposium on Advanced Database Systems, June 19-22, 2022, Tirrenia (Pisa), Italy
$ ada.bagozi@unibs.it (A. Bagozi); devis.bianchini@unibs.it (D. Bianchini); anisa.rula@unibs.it (A. Rula)
© 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR
Workshop
Proceedings
http://ceur-ws.org
ISSN 1613-0073
CEUR Workshop Proceedings (CEUR-WS.org)
�and web-based systems, leading to the notion of Cyber Physical Production Network (CPPN).
A CPPN is an ecosystem where cyber and engineered physical elements collect (e.g., using
sensors) and analyse data using connected services (e.g., over cloud computing infrastructures).
It allows the interaction with human actors using multi-channel interfaces, going beyond the
boundaries of a single enterprise and spanning over the whole production network.
In this paper, we propose an information model that relates the digital counterpart of a
product with data collected over the phases of the product lifecycle, with special attention
on the manufacturing process, and with data gathered to monitor the shop floor machines
(work centers) used during the production. The information model is represented according
to multiple perspectives, namely the production process, the industrial assets and the product,
used to organise the data collected along the product lifecycle. The approach also enables the
development of a portfolio of data-oriented services on top of the model, properly tailored on
the perspectives that can be composed to implement advanced supply chain functionalities
such as production scheduling, energy efficiency, product and process monitoring. Finally, we
provide an architecture of the proposed solution and its validation in an industrial research
project, to demonstrate its effectiveness for a particular case of the remote monitoring service
to detect near real-time anomalous events.
The paper is organised as follows: in Section 2 the multi-perspective model is described;
Section 3 introduces data-oriented services design; in Section 4 implementation and experimental
validation are discussed; comparison with related work is presented in Section 5; finally, Section 6
closes the paper.
2. Multi-perspective information model
The proposed multi-perspective information model is shown in Figure 1. The model is organised
according to the following perspectives.
Product perspective. The product perspective concerns product configuration over the
Product Lifecycle Management (PLM). Each product (in turn described through product details)
is composed of a set of parts which are identified by a part code and are hierarchically organised
into different BoM, depending on the specific phase of the product lifecycle that is being
considered (e.g., Engineering BoM, Manufacturing BoM). Part codes belonging to different
hierarchies can be connected to each other, e.g., in order to manage changes propagation over
the product lifecycle. We focus on the MBoM, which concerns the manufacturing process and
reports the distinction between: (i) part codes that must be produced internally, connecting to a
Production Order (PO); (ii) part codes that must be produced externally, by one of the suppliers
of the supply chain, connecting to a Contract Work Order (CWO); (iii) part codes to be bought
and then assembled in the final product, connecting to a Purchase Order. The final product,
corresponding to the root of the BoM hierarchy, is associated with the PO as a Sales Order item.
Process perspective. In the process perspective, the main entity is represented by the
production phase, that is in turn organised according to a hierarchy, where each phase is
composed of a set of sub-phases, reaching a level of detail ranging from macro to micro
industrial processes. Production phases implement: (i) the PO of the final product, connected to
the Sales Order item; (ii) the PO of one or more components of the final product, connected to
� Product perspective
{ JSON
part_of { …
“partcode_ID”:…;
…
… “partcode_ID”:…;
Product } …
}
Partcode
(E-BOM)
part_of
Product aggregated
measure
…
Partcode …
(M-BOM)
Process perspective
buy
SalesOrder SalesOrder
make_external { JSON
(SO) Item { …
“prod_order”:…;
…
“phase_ID”: …;
“prod_order”:…;
… “phase_ID”: …;
ContractWorkOrder } …
PurchaseOrder }
PO_owner (CWO)
Client make_internal
ContractWorkOrder …
PO_supplier Item
ProductionOrder
(PO)
PurchaseOrder Process aggregated
PO_supplier Item measure
subphase_of
ProductionPhase
…
ProductionPhase
Execution
Asset perspective composed_of
{
JSON
Resource { …
“workcenter_ID”:
… …;
…
“workcenter_ID”: …;
WorkCenter } …
}
Asset aggregated
measure
Employee Equipment Software …
…
Figure 1: Multi-perspective information model of the Cyber Physical Production Network.
the CWO item or to the Purchase Order item and corresponding to the point of view of one
of the suppliers in the supply chain. In this way, the information model brings together the
viewpoints of all the supply chain actors. In this case, the hierarchy of process phases is focused
on the production step in the PLM, thus corresponding to the MBoM. However, this modelling
can be seamlessly extended to the other phases of the PLM (e.g., engineering, maintenance) and
the other kinds of the BoMs.
Asset perspective. The asset configuration perspective includes the resources involved in the
realisation of the product (machinery, equipment, information systems, human resources). Work
centers are the machines used in the manufacturing process and are hierarchically organised
according to the RAMI 4.0 Reference Architectural Model for Industry 4.0 [3]. A work center or
a resource can be involved in many production phases during its lifecycle. A production phase
can be executed and distributed on several work centers and may consume/require different
resources.
Parameters. Parameters are distinguished among aggregated measures (e.g., KPI) and data
streams collected as-is from the field (e.g., sensors data acquisition). In case of aggregated
measures a new entity is created for each measure containing records with the calculated value,
a time reference (depending on the granularity of the aggregation, e.g., monthly/weekly/daily
�and so forth) and relationships with the target entities of the multi-perspective data model. The
data stream collected from a vibration sensor on a specific work center or component is an
example of this kind of parameter.
3. Data-oriented services design
A service repository is designed on top of the multi-perspective data model. Services are
categorised according to different service types and over the three perspectives of the product,
process and assets. Concerning service types, we distinguish among: (i) collect services, used
to transfer data from the physical side of the production network towards the data storage
on cloud; (ii) monitor services, used to proactively identify/predict warnings/errors about the
phenomena occurring on the physical production network [4]; (iii) dispatch services, used to
share data across the actors of the production network; (iv) display services, exposed to visualise
data on the dashboard. We formally define a multi-perspective data-oriented service as follows:
Definition 1 (Data-oriented service). A data-oriented service 𝒮𝑖 is described as a tuple
𝑆𝑖 = ⟨𝑛𝑆𝑖 , 𝑡𝑆𝑖 , 𝑃𝑆𝑖 , 𝑢𝑟𝑙𝑆𝑖 , 𝑚𝑠𝑖 , 𝐼𝑁𝑆𝑖 , 𝑂𝑈 𝑇𝑆𝑖 ⟩ (1)
where: (i) 𝑛𝑆𝑖 is the name of the service; (ii) 𝑡𝑆𝑖 is the service type, namely: collect, monitor, dispatch,
display; (iii) 𝑃𝑆𝑖 is the set of perspectives on which the service is focused (product, process, assets);
(iv) 𝑢𝑟𝑙𝑠𝑖 is the endpoint of the service for it invocation; (v) 𝑚𝑠𝑖 is the HTTP method (e.g., get,
post) used to invoke the service; (vi) 𝐼𝑁𝑆𝑖 is the representation of the service input; (vii) 𝑂𝑈 𝑇𝑆𝑖 is
the representation of the service output. We denote with 𝒮 the overall set of data-oriented services.
Services in 𝒮 are implemented as RESTful services; therefore, they are also described through
HTTP method (e.g., post, get, put and delete) implemented in the service, corresponding to
the Create-Read-Update-Delete actions, and the resource that is being created, read, updated or
deleted, respectively. A multi-perspective service 𝑆𝑖 deals with different types of data, based on
its input/output parameters. Among the inputs/outputs of 𝑆𝑖 we could mention entities from
the process, product and assets perspectives of the data model. Moreover, input/output data can
be parameters (i.e., calculated measures or data-streams collected from the physical side of the
production network). Services dealing with data-streams are called data-intensive services and
need advanced techniques to deal with data-streams. For instance, a predictive maintenance
service is a data-intensive monitor service if it detects concepts drifts from massive data streams
collected on the monitored machines. Service inputs/outputs are represented in JSON format
and modelled in terms of information model entities and their relationships.
4. Approach implementation in a real case study
The approach presented in this paper guided the implementation of the architecture presented in
Figure 2. The proposed architecture is composed of four main parts: (i) the production network,
including the process phases and actors involved in each phase; (ii) the services repository,
integrating the set of services aimed at implementing the objectives of actors in the supply
�Figure 2: Approach architecture.
chain; (iii) the data repository, containing parameters values coming from different sources and
owned by different actors; (iv) a multi-perspectives dashboard, for detailed exploration of data
according to the multiple perspectives.
For the production network, we considered the real case study about the production of
valves to be used in deep and ultra-deep water applications. Valves are placed in prohibitive
environments and, once installed, are difficult to remove and maintain over time and require
high quality levels. Valves supply chain involves several actors to produce different sub-parts of
the valve: the valves producer, who represents the OEM of the production process; the forger,
who is in charge of transforming the raw metals into workable pieces; the mechanical supplier,
who is in charge of modelling pieces provided by the forger to be assembled in the final product;
the product quality test company, who is in charge of performing quality tests on the final
valves. The parameters values are collected from the production network and stored in data
repository, thanks to sensors and IoT technologies, or from legacy systems owned by involved
actors (e.g., ERP, PDM). Data streams are saved within a NoSQL database (e.g., Product Data,
Process Data and Asset Data) as JSON documents using MongoDB technology. Each JSON
document represents a record of collected measures within a set of parameters and includes the
timestamp and a reference to the perspective in which the measures have been collected. Supply
chain actors have different goals, that are achieved through the deployment of combinations of
data-oriented:
• Production Scheduling Services, that interact with the schedulers of each actor
involved in the production process and support the actors in scheduling the production
in an efficient way;
• Energy Efficiency Services, to monitor the efficiency of energy harvesting assets
in the supply chain;
• Product Quality Services, in charge of correlating anomalies detected on the prod-
uct along its lifecycle with anomalies raised on work centers involved in the production
process;
�Figure 3: Comparison between correct and wrong answers in the usability experiment.
• Remote Monitoring Service, to monitor the work centers in order to detect anomalies
that may lead to higher consumption or breakdown/damage of the monitored system or
production process failure in the near future; to this aim, the IDEAaS approach, described
in [4], has been integrated within the architecture; IDEAaS is a novel system used to
detect anomalies within data streams in Industry 4.0 applications, by applying incremental
clustering and data relevance evaluation techniques to manage volume and velocity of
incoming data; the IDEAaS approach can benefit from the partition of incoming data
streams according to different perspectives or contexts; for this reason, the application
of IDEAaS techniques has been fostered through its integration on top of the multi-
perspective information model of the CPM, as demonstrated in the next section on
experimental results.
These services are strongly interleaved over a supply chain. For example, a peak in energy
consumption does not always mean a low efficiency issue. Indeed, such peak may be due to a
dense production scheduling. By correlating the output of the production scheduler with the
energy efficiency optimisation module, these cases can be properly detected. For this reason,
the mechanical supplier relies on the service, which is used for the anomaly detection. Similarly,
problems raised during product quality tests may be correlated with anomalies raised on the
machines of one of the participants in the production process. To implement all these services,
acquisition and visualisation of data under the three perspectives of product, process and
industrial assets are required for all the actors of the supply chain.
Finally, display services populate a web-based dashboard, that the actors use for data explo-
ration. From the home page of the dashboard, it is possible to start the data exploration by
following one of the three perspectives, namely, product, process and industrial assets. Each
perspective brings to a UI component (tile) implemented using ReactJS libraries: i) the product
synoptic tile allows an exploration from the product perspective; ii) the process phases tile allows
an exploration from the process perspective; iii) the working centers tile allows exploration
from the industrial asset perspective. The management of different access permissions through
display services leads to tiles customisation. Therefore, it is possible through the dashboards to
configure what to show and what to hide for each actor according to his/her role.
Preliminary validation of the approach. We performed a usability experiment with 13
participants grouped in three categories, with increasing expertise on the use of data visualisation
tools, such as synoptic and dashboard, to explore data relate to industrial processes. The users
�have been requested to perform a set of exploration actions, that included multiple perspectives,
(e.g., What are the problems, if any, that occurred on industrial assets during the production
of the valve VSS000799?). The experiment is significant, since the experimental setup and
the participants are representative of the considered industrial context, and reliable (in this
particular context, poor reliability is linked to users’ variability). The experiment comprised
four steps: test preparation, its introduction to participants explaining the purpose of the test, its
execution, in which the examiner observes without influencing the participant, and debriefing,
to collect through questionnaires the participants’ opinions about the perceived difficulty of
the experiment. Figure 3 shows the correctness of the answers, having a high rate of 89%,
thus demonstrating the effectiveness of the multi-perspective approach in supporting data
exploration in industrial environments.
5. Related Work
The modelling approach addressed in this paper has been compared with the literature about
the design of information models for managing the data flow from the physical world and the
cyber one, where solutions such as Digital Twins and Digital Threads have been investigated.
Authors in [5] describe a model-based approach (and a corresponding web-based GUI) to
compose predefined building blocks, abstracted as Smart Services, connected each other and
hierarchically organised. In [6] authors proposes an architecture where users specify a goal and
take advantage of technologies such as digital twins to automatically compose the corresponding
physical processes. Ontologies have been also proposed in [7] to face interoperability issues.
In these papers the focus is on CPS within a single production line. Authors in [8] propose a
generic architecture based on digital twin that relies on four layers (physical, network, virtual
and application layer). In [9] authors model Digital Twins for product customisation. An
information model is proposed to provide data about product, process, plan, plant, resource.
The approach provides five services: production planning, automated execution, real-time
monitoring, abnormal situation notification and dynamic response. In our information model,
process and plan are associated to the different concepts of production phases and phase
execution. Moreover, resources are properly organised within plants in a hierarchical way and
services are designed on top of such multi-perspective model. The modern version of Digital
Thread can be seen as a step forward with respect to Digital Twins. In [1] the authors focus
on the use of models for designing smart products along their lifecycle, being agnostic about
the specific technologies, and binding to specific implementations of such features only when
needed. A case study for the design of Universal Robots UR3 controller is proposed in [10] within
a model-driven integrated development environment independently from any implementation
programming language, operating system, or runtime platform. The focus in the latter papers
is on the smart product, whose representation evolves during the product lifecycle, but no data
is collected on the design, production or maintenance stages. Similarly, the notion of Digital
Thread proposed by commercial solutions such as PTC Windchill is implemented as a sequence
of interleaved BOMs (e.g., EBOM, MBOM, as-built, as-maintained) without collecting data on
the process and on the work centers during the product lifecycle. Our approach, although
mainly focused on the production phase, could be fruitfully extended also during the other
�stages of the product lifecycle, such as the design or the maintenance ones.
6. Concluding Remarks
In this paper, an information model has been proposed that relates the digital counterpart
of a product with data collected over the phases of the product lifecycle, according to three
perspectives, namely the production process, the industrial assets and the product. The approach
goes beyond the boundaries of a single actor, embracing all the participants in the production
network, thus including also the definition of a portfolio of services on top of the model and
data and services access rules for different actors involved in the supply chain. Future efforts
are required to test the proposed solution in an intensive way when the massive collection of
data will start in the project. Furthermore, a methodology is being developed to extend the
portfolio of services to other kinds of manufacturing production networks and with proper
access control policies to manage the authorisation of supply chain actors on data and services.
References
[1] T. Margaria, A. Schieweck, The Digital Thread in Industry 4.0, in: IFM 2019, 2019, pp.
3–24.
[2] P. Loucopoulos, E. Kavakli, N. Chechina, Requirements engineering for cyber physical
production systems, in: CAiSE 2019, volume 11483, Springer, 2019, pp. 276–291.
[3] M. Hankel, B. Rexroth, The Reference Architectural Model Industrie 4.0 (RAMI 4.0), in:
ZVEI, 2015.
[4] A. Bagozi, D. Bianchini, V. D. Antonellis, M. Garda, A. Marini, A relevance-based approach
for big data exploration, FGCS 101 (2019) 51 – 69.
[5] D. Stock, D. Schel, T. Bauernhansl, Middleware-based Cyber-Physical Production System
Modeling for Operators, Procedia Manufacturing 42 (2020) 111–118.
[6] T. Catarci, D. Firmani, F. Leotta, F. Mandreoli, M. Mecella, F. Sapio, A conceptual ar-
chitecture and model for smart manufacturing relying on service-based digital twins,
in: E. Bertino, C. K. Chang, P. Chen, E. Damiani, M. Goul, K. Oyama (Eds.), 2019 IEEE
International Conference on Web Services, ICWS 2019, Milan, Italy, July 8-13, 2019, IEEE,
2019, pp. 229–236.
[7] D. Stock, D. Schel, Cyber-Physical Production System Finger-printing, Procedia CIRP 81
(2019) 393–398.
[8] H. Zhang, Q. Yan, Z. Wen, Information modeling for cyber-physical production sys-
tem based on digital twin and automationml, The international journal of advanced
manufacturing technology 107 (2020) 1927–1945.
[9] K. Park, J. Lee, H. Kim, S. Noh, Digital twin-based cyber physical production system
architectural framework for personalized production, International Journal of Advanced
Manufacturing Technology 106 (2020) 1787–1810.
[10] S. Bosselmann, M. Frohme, D. Kopetzki, M. Lybecait, S. Naujokat, J. Meubauer, D. Wirkner,
P. Zweihoff, B. Steffen, DIME: A Programming-Less Modeling Environment for Web
Applications, in: ISoLA 2021, 2016, pp. 809–832.
�