2016 13th International Joint Conference on Computer Science and Software Engineering (JCSSE)

Process Modeling for Industry 4.0 Applications

Towards an Industry 4.0 Process Modeling Language and Method

Roland Petrasch Roman Hentschke

Department of Computer Science AMDIS Media and IT Services GmbH
Thammasat University Berlin, Germany
Bangkok, Thailand [email protected]
[email protected]

Abstract—The term Industry 4.0 derives from the new (fourth) are (semi-) autonomous and can be equipped with advanced
industrial revolution enabling suppliers and manufacturers to computing power or AI (Artificial Intelligence).
leverage new technological concepts like Internet of Things, Big
Data, and Cloud Computing: New or enhanced products and Industry 4.0 comprises the following terms or technological
services can be created, cost be reduced and productivity be concepts:
increased. Similar terms are Smart Factory or Smart
Manufacturing. The ideas, concepts and technologies are not • Embedded Systems (ES) / CPS (Cyber-Physical Sys-
hype anymore - they are at least partly reality, but many tems): Networks of IoT devices that interact physically
software specification and development aspects are still not with its environment, e.g. industrial robots with sensors
sufficiently covered, e.g. standardization, specification and and actors need physical input and provide physical
modeling languages. This paper presents an Industry 4.0 process output. ES and CPS are can also be equipped with digi-
modeling language (I4PML) that is an extension (UML profile tal interfaces. An ATM is an example for an embedded
with stereotypes) of OMG’s BPMN (Business Process Model and system.
Notation) standard. We also describe a method for the
specification of Industry 4.0 applications using UML and I4PML. • Internet / Cloud of Things (IoT / CoT): Physical ob-
jects or components like ES or CPS that contain soft-
Industry 4.0, Smart Manufacturing; Internet-of-Things; IIoT, ware and are connected to a network (Internet connec-
Internet-of-Services; Business Process Modeling Language; tivity) and typically to a Cloud application creating op-
BPMN; Industry 4.0 Process Modeling; I4PLM; UML; Profile portunities for new services through integration of the
physical and the digital world, e.g. automatic monitor-
I. INTRODUCTION ing. The “things” can be any items, e.g. a conveyor belt
in a factory. In the context of Industry 4.0 the term In-
Industry 4.0 is part of the high-tech strategy of the German dustrial Internet of Things (IIoT) is used.
Federal Ministry for Education and Research [1], [2]. The term
Industry 4.0 derived from the 4th industrial revolution and is a • Service-Oriented Architecture (SOA) / Internet of Ser-
technology-oriented concept mainly for the manufacturing do- vices (IoS) / Cloud Computing (CC): Service-oriented
main but can be interpreted more generally and applied to any and cloud-based infrastructures and applications have
value chain organization. advantages concerning scalability, elasticity, reliability,
performance, device and location independence and
The 4th industrial revolution means a paradigm shift in in- more. Service models are for example Infrastructure as
dustry: The first industrial revolution brought the mechaniza- a Service (IaaS), Platform as a Service (PaaS), Soft-
tion of production, the second industrial revolution was about ware as a Service (SaaS). Cloud applications can also
mass production and the third industrial revolution means the be categorized in private, public or hybrid Cloud.
digitization (electronic component, computer and IT. Industry Products for CC are available as open or closed source
4.0 enables suppliers and manufacturers to leverage new tech- software systems [5], [6].
nological concepts like CPS (Cyber-Physical Systems), Inter-
net of Things and Cloud Computing (CC): New or enhanced Important topics like Big Data and Business Intelligence
products and services can be created, cost can be reduced and are also part of Industry 4.0, but in order not to go beyond the
productivity can be increased [3]. scope of this short introduction, the concepts mentioned above
are considered sufficient for the understanding of the idea of
In the discussions about Industry 4.0 the term Smart Facto- Industry 4.0.
ry [4] is mentioned quite often and even if the word smart
might be open to debate, the main point is clear: The idea is to The terms Internet of Things (IoT) and Cloud Computing
create a network with decentralized decisions and technologi- (CC) imply a technology-oriented standpoint, e.g. devices and
cal components with enhanced capabilities that can interact other electronic things play an active role in an environment
with each other and with human in real time. The components and usage context that is connected (to the Internet). But when

978-1-5090-2033-1/16/$31.00 ©2016 IEEE

978-1-5090-2033-1/16/$31.00
Authorized licensed ©2016do
use limited to: UDESC - Universidade IEEE
Estado de Santa Catarina. Downloaded on June 22,2020 at 12:48:25 UTC from IEEE Xplore. Restrictions apply.
 2016 13th International Joint Conference on Computer Science and Software Engineering (JCSSE)

it comes to application scenarios, a sole technology-centric

approach could be counterproductive: IoT devices are not used
for their own sake; they must contribute to a more effective or
a more efficient business process or a better user experience.

II. BUSINESS PROCESS MODELING FOR INDUSTRY 4.0 AP-

PLICATIONS
Business process models (BPM) describe behavioral as-
pects of a system and are usually on the formal requirements or
early design level. All major architectural and methodological
frameworks include process modeling: TOGAF with the archi-
tecture domains (e.g. Business Architecture) and the ADM
(Architecture Development Method) [7], RM-ODP (Reference
Model of Open Distributed Processing) with its Enterprise and
Computational Viewpoint [8], [9], and the FEAF (Federal En-
terprise Architecture Framework) with its Business Reference
Model [10]. In terms of OMG's MDA (Model Driven Architec-
ture) BPM is a PIM (Platform Independent Model) concerning
the design and implementation platforms [11].
Figure 1. Definition of the I4PML Profile: Stereotypes with meta-class
The notation language in this context is the OMG Business assignment and icons (Modeling Tool: MagicDraw, No Magic, Inc.).
Process Model and Notation (BPMN 2.0) [12] with its model-
ing elements like tasks (service, send, receive etc.), events and In order to use I4PML as a modeling language, a method is
data objects. Language extensions for aspects of Industry 4.0 to be defined that provides concrete steps. The following sec-
application are needed, e.g. IoT-aware process models (IoT-A, tion describes such an Industry 4.0 Process Modeling Method
Internet of Things Architecture [13], [14]). Existing extensions (I4PMM).
described in [13] and [15] are used and also a new extension
(for Cloud App) are presented here. This BPMN-based lan-
guage is called Industry 4.0 Process Modeling Language III. A METHOD FOR PROCESS MODELING FOR INDUSTRY 4.0
(I4PML): The following table shows the extensions as icons. APPLICATIONS
This section summarizes our method presented in [15]
TABLE I. ICONS TO BE USED FOR I4PML PROFILE (IoT-Aware Process Modeling) and describes some modifica-
tions and extensions concerning Industry 4.0. The approach
Icon Description Used for Ref. uses OOA/OOD and systems engineering techniques [16],
Mobility Aspect
Partition,
[13] [17], [18]. The fundamental activities of I4PMM are as fol-
Pool, Lane lows:
Actuation Task Task [13]
A. Requirements: Basic functional decomposition
Sensing Task Task [13]
• Requirements modeling leads to a semi-formal descrip-
IoT Device tion of the requirements and can be classified in MDA
Partition [15]
terms as a CIM (Computational Independent Model).
Human Computer Interface Partition,
[15] The result is a hierarchy, table or model of the re-
Task quirements, e.g. as a SysML requirements diagrams
Real/device data object [19].
Data Object [13]
• Use Case modeling is the refinement of the require-
Real world/device data store Central
[13] ments (e.g. with the template of A. Cockburn, [20])
Buffer and also the visual representation with UML use case
C loud App, diagrams [20]. Also data elements are identified and
also as described.
Partition,
public, new
Pool
private or
hybrid Cloud B. Identification of Industry 4.0 aspects and process model-
ing
Cloud App, IoT devices, device data, actuation and sensing • During the analysis of the requirements an assessment
tasks, HCI, and mobility are described on a lightweight, but of the use cases and the associated actions, activities,
formal level as a UML profile for I4PML (s. Figure 1). The devices, data elements etc. takes place. Every element
stereotypes only extend, but don't modify BPMN2, so that the is tagged or marked with possible Industry 4.0 aspects,
complete language is available plus the aspects for Industry e.g. IoT devices and data, Human Computer Interac-
4.0. I4PML modeling patterns, like “IoT device data sync with tions etc. (s. TABLE I).
Cloud app” are helpful during process modeling tasks.

Authorized licensed use limited to: UDESC - Universidade do Estado de Santa Catarina. Downloaded on June 22,2020 at 12:48:25 UTC from IEEE Xplore. Restrictions apply.
 2016 13th International Joint Conference on Computer Science and Software Engineering (JCSSE)

• The creation of process models with the Industry 4.0 PreMa system is part of the supplier’s new ICT-oriented busi-
Process Modeling Language (I4PML) takes the tagged ness strategy of becoming a full-service partner for its clients.
elements from the previous step and leads to a com-
A conveyor belt needs to be monitored and maintained to
prehensive and formal model of the business process.
avoid costly downtime of the production site and maximizing
the safety and the lifetime of the belt. Belt monitoring services
C. Detailed specification of data elements are for example belt surface inspection, belt rip detection, and
• Data elements must be described in detail usually with belt thickness measurement. Several technological components
a UML domain class model and OCL constraints for like conductor loops in the belt, or devices are used including a
the consistency rules. communication system that involves field engineers among
others.
• The domain class model can consider Industry 4.0 as-
pects like IoT device data sets. The process models use The scenarios for emergency and predictive maintenance
these data objects. are complex and involve components like IoT devices, com-
munication networks, software applications, and storage for big
D. Consistency check of all models data. Also analytic and machine learning (ML) subsystems are
used. The following two sections exemplify the specification
• Reviews of the models after each step mentioned activities of I4PMM (only steps A. and B. are shown).
above ensure the correctness and consistency between
the model elements, e.g. the terminological consisten-
cy. A. Requirements: Basic functional decomposition
Figure 4 shows some requirements for the PreMa system. A
• Corrections of one model or particular elements might Systems Modeling Language (SysML) requirements diagram is
have an impact on other element and require a repeti- used for this purpose.
tion of some modeling steps.
The method (I4PMM) focus on the software requirements
analysis and does not cover the complete system design of a
development project. The outcome is a semi-formal and de-
tailed software requirements specification. Non-functional as-
pect like user experience, are omitted here.

IV. INDUSTRY 4.0 PROCESS MODELING: AN EXAMPLE

This section presents an example for applying the I4PMM
using the I4PML. The domain for this example is the conveyor
belt industry: A belt conveyor system consists of pulleys and
an endless belt that transports large volumes of resources, e.g.
earth, pieces of rock or ore. The pulleys are powered and mov-
ing the belt and the material on the belt forward. Long convey-
or can have a length of more than 10 km (see Figure 2). Figure 3. Part of the requirements diagram for the PreMa system (SysML
requirements diagram)

Because of the high cost of a blackout, downtime must be

avoided or minimized. Therefore not only the IoT devices, but
also the communication and workflow modules of the PreMa
system are crucial parts.
The functional decomposition is modeled in the form of a
UML use case diagram [21]: Three subsystems (Belt monitor-
ing, Predictive Maintenance and Emergency) contain use cases
and actors like the field engineer, the site manager and the belt
(s. Figure 4.).
Based on the belt operation data and the knowledge base
with an inference engine and a ML subsystem, the situation is
Figure 2. Conveyor belt at a mining site (source: ContiTech AG, analyzed (use case “analyze belt data”) and – if further actions
www.contitech.de)
are considered necessary – events and prepared information
about the belt status are reported to the field engineer. In the
Predictive maintenance and emergency management for
case where a maintenance cycle is recommended (use case
conveyor belts serves as an example for a typical Industry 4.0
“recommend maintenance”), the field engineer submits this
scenario: In order to provide a better service for its customers,
maintenance recommendation to the site manager (use case
a manufacturer (supplier) for conveyor belts that are used for
“notify event”). But in some cases the field engineer escalates
instance in the mining industry, decides to offer a Predictive
Maintenance (PreMa) Service to its customers (clients). This

Authorized licensed use limited to: UDESC - Universidade do Estado de Santa Catarina. Downloaded on June 22,2020 at 12:48:25 UTC from IEEE Xplore. Restrictions apply.
 2016 13th International Joint Conference on Computer Science and Software Engineering (JCSSE)

the event to the emergency system, because the situation may

result in a shutdown of the site (use case “report incident”).

Figure 4. Part of the Use Case diagram for the PreMa system (SysML
requirements diagram)

A. Identification of Industry 4.0 aspects and process model- Figure 5. Process model for acquisition of belt operation data using the
ing Industry 4.0 Process Modeling Language (I4PML)

The requirements specification allows the detailed analysis The model shows the belt as an IoT device (partition “Con-
of Industry 4.0 aspects. Here are some examples for the PreMa veyor Belt”). Every 10 min. the operation data is to be acquired
system: (timer start event). If this acquisition is successful then an “end
of Data-set (DS)” is received, otherwise a sub-process for error
• IoT device: Actors or subsystems (see use case dia-
handling is triggered (1 min. timeout). The belt operation data
gram) are candidates for being IoT devices. In this case
set is transferred to a cloud application (pool “PreMa Cloud
the conveyor belt with the sensor devices that transmit
App”) that stores it in a device data store and shows notifica-
belt operation data is identified as such a device.
tions to the user if necessary (Non-interrupting conditional
Since the belt as an IoT device is an active and integral boundary event).
part in the system it is to be modeled as a partition in
The example scenario demonstrates the relevance of busi-
the process model.
ness process management when dealing with mission-critical
• IoT tasks: IoT devices typically interact with the envi- data and processes, distributed components, complex work-
ronment. This leads to actuation and sensing tasks. flows, and different user groups.
Sensing of belt operation data is such an IoT device VI. Conclusion
task.
In the context of a software development project user’s
• Real world or device data objects can be identified needs and business goals should be considered first. Therefore
from a use case analysis, e.g. the belt operation data a concept for an industry 4.0 application is necessary before
need to be buffered so that they can be transferred to any technological investments are made and new hard- or
the cloud app. software is introduced or implemented. This concept (require-
• Real world or device data stores persist data from IoT ments specification) includes the description and modeling of
devices. A cloud application needs to store the belt op- business processes and data structures that are aware of new
eration data in such a data store. components, e.g. IoT devices or Cloud Computing applica-
tions, because aspects like security, usability, availability, reli-
• Private cloud application can be accessed only via au- ability need to be taken into consideration in the early stage of
thorized persons or services. PreMa with its sensitive the development and introduction of Industry 4.0 applications
information runs on a private cloud. systems. Since devices and services play an active part in busi-
The next step is the modeling of the business process for ness processes, specifications must represent their characteris-
the acquisition of belt operation data (including the presenta- tics, but not all the technological details.
tion of notifications to the user) in order to measure the belt The Industry 4.0 Process Modeling Method (I4PMM)
status and predict maintenance cycles. The Industry 4.0 Process makes it possible to consider Industry 4.0 aspects like IoT de-
Modeling Language (I4PML) is used and shows aspects like vices on a non-technological level for business process model-
IoT device, Sensing Task and Data Storage as stereotypes ap- ing introducing a modeling notation based on BPMN2. The
plied to BPMN elements (see Figure 5.). advantages of a semi-formal process I4PML model are:
• Model validation and simulation: Formal (machine
readable) models can be easier validated and simulated

Authorized licensed use limited to: UDESC - Universidade do Estado de Santa Catarina. Downloaded on June 22,2020 at 12:48:25 UTC from IEEE Xplore. Restrictions apply.
 2016 13th International Joint Conference on Computer Science and Software Engineering (JCSSE)

than pure drawings. I4PML as an UML profile has a [3] Deutsche Bank Research: Industry 4.0 - Upgrading of Germany’s
formal definition and can be integrated in existing val- industrial: capabilities on the horizon:
http://www.dbresearch.com/PROD/DBR_INTERNET_EN-
idation and simulation tools (s. Figure 6.). PROD/PROD0000000000333571.pdf (visited on March 9, 2015)

[4] Dominik Lucke, Carmen Constantinescu, Engelbert Westkämper: Smart

Factory - A Step towards the Next Generation of Manufacturing. In:
Mamoru Mitsuishi, Kanji Ueda, Fumihiko Kimura (Eds.) :
Manufacturing Systems and Technologies for the New Frontier. The
41st CIRP Conference on Manufacturing Systems May 26–28, 2008,
Tokyo, Japan, pp 115-118
[5] OpenStack, www.openstack.org/
[6] Oracle Corp.: Oracle cloud: Software as a Service,
https://cloud.oracle.com/home (visited on April 10, 2015)
[7] Open Group Architecture Forum: Open Group Standard, TOGAF
Version 9.1, Document Number: G116, 2011
[8] ISO/IEC IS 10746: Information technology — Open distributed
processing — Reference model, 2009 (see also ITU-T X.900,
http://www.rm-odp.net/)
[9] ITU Group (United Nations): Information technology – Open distributed
processing – Use of UML for ODP system specifications.
Recommendation ITU-T X.906, 2014
Figure 6. Process model validation (MagicDraw / Cameo Business Modeler, [10] The White House: Federal Enterprise Architecture Framework, Version
NoMagic [22]) 2, 2013
[11] Object Management Group (OMG): MDA Guide.Version. 1.0.1,
• Process Patterns: Re-use of typical scenarios like omg/2003-06-01, June 2003
“transmission of IoT sensor data with error handling” [12] OMG (Object Management Group), “Business Process Model and
Notation (BPMN)”, Version 2.0.2, document formal/2013-12-09, 2013
is possible with process patterns modeled with I4PML.
[13] S. Meyer et al., “Internet of Things Architecture - IoT-A - Project
A pattern library supports the business analyst and pre- Deliverable D2.2 – Concepts for Modelling IoT-Aware Processes” Doc.
vents modeling errors. Ref.: D2.2, 2012
• Code generation: Transformations of I4PML models [14] EU project IoT-A, http://www.meet-iot.eu (7.10.2015)
can lead to artifacts for a workflow engine or an API [15] R. Petrasch, R. Hentschke, “Towards an Internet-of-Things-aware
Process Modeling Method - An Example for a House Suveillance
for software development. Model-to-Model (MOF System Process Model”, 2nd Management and Innovation Technology
QVT [23]) and Model-to-Text (M2T [24]) transfor- International Conference (MITiCON2015), Bangkok, Thailand,
mation languages and tools are available. November 16-18, 2015
[16] D. J. Mala, S. Geetha, “Object Oriented Analysis and Design Using
A question is how much technological details are necessary UML”. McGraw-Hill Education, 2013
for a process model during the analysis and requirements speci- [17] B. Oestereich, S. Bremer, A, Scheithauer, “Analyse und Design mit der
fication phase, e.g. is it important or essential to include an IoT UML 2.5: Objektorientierte Softwareentwicklung”, 10. Aufl.,
device in the process model? To a certain extend these aspects Oldenbourg Verlag, 2012
need to be considered. This is the case in many other domains, [18] S. Friedenthal, A. Moore, R. Steine, „A Practical Guide to SysML: The
e.g. the user of a car also needs to know some technological Systems Modeling Language”, 3rd edition, Elsevier, 2014
details like filling the tank with gasoline. [19] OMG (Object Management Group), “OMG Systems Modeling
Language (OMG SysML)”, version 1.4, document formal/2015-06-03,
Therefore the abstraction level of I4PML models can con- 2015
sidered to be sufficient: The models hide implementation and [20] A. Cockburn, “Writing Effective Use Cases”, Addison-Wesley, 2001
technological details, but allow user and usage relevant con- [21] OMG (Object Management Group), “OMG Unified Modeling Language
cern to be included as model elements. (OMG UML)”, version 2.5, document formal/2015-03-01, 2015
[22] NoMagic: Cameo Simulation Toolkit for MagicDraw,
https://www.nomagic.com/products/magicdraw-addons/cameo-
REFERENCES simulation-toolkit.html
[1] Bundesministerium für Bildung und Forschung, Zukunftsprojekt [23] OMG: Meta Object Facility 2.0 Query/View/Transformation (QVT),
Industrie 4.0: www.bmbf.de/de/9072.php V1.3 – Beta, 2016
[2] Malte Brettel, Niklas Friederichsen, Michael Keller, Marius Rosenberg : [24] OMG: Model to Text Transformation Language (M2T) 2008
How Virtualization, Decentralization and Network Building Change the
Manufacturing Landscape: An Industry 4.0 Perspective. World
Academy of Science, Engineering and Technology International Journal
of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing
Engineering Vol:8, No:1, 2014, pp 37-44

Authorized licensed use limited to: UDESC - Universidade do Estado de Santa Catarina. Downloaded on June 22,2020 at 12:48:25 UTC from IEEE Xplore. Restrictions apply.