A Hybrid Method integrating Industry 4.0's Energy Digitization
AGOUZZAL KAWTAR, ABBOU AHMED
Department of Electrical Engineering, Mohammed V University of Rabat
Mohammadia School of Engineers, MOROCCO
Abstract: - Industrial firms must face important environmental challenges (greenhouse gas emissions, energy
efficiency) and business imperatives. For this sector to accomplish a lower-cost energy transition, digitalization
is a key lever.
Today, the fourth industrial revolution is constructing a forward-thinking first industry known as industry 4.0,
which combines many developing technologies to produce digital and efficient solutions.
In this paper, we examine the impact of Industry 4.0 on the evolution of a new simulation modeling paradigm
embodied by the concept of Digital Twin.
To begin, we will discuss the industry 4.0 paradigm, its history, current state of development, and its impact
on the development of the simulation modeling paradigm.
A needs-based approach can result in the faster, deeper, and more extensive implementation of efficient systems.
Furthermore, we present the methodology's multiple case studies and discuss several research and
development projects involving the modeling of automated industrial processes that have been presented in recent
scientific publications.
The lack of tools, however, is not a problem because the current generation of general-purpose simulation
modeling tools provides adequate integration options. However, to build on several physical levels of the
integrated model system, close collaboration between academia and industrial partners is required to demonstrate
industry and scientific community acceptance of the new analog modeling paradigm.
Adoption and development of relevant morality in a needs-based process can lead to more efficient industrial
automation implementation that is faster, deeper, and more extensive.
Keywords: - Industry 4.0, Digital Twin, SME (Small and Medium-sized Enterprises), energy digitalized,
simulation and modeling; automated modeling
ICT: Information and communication technologies, OEMs: original equipment manufacturers, CPS: cyber-
physical system, CPPS: physical production systems, DES: discrete event simulation, SQL: Structured Query
Language, XML: Extensible Markup Language, ABM: Agent-Based Modeling, CMSD: Core Manufacturing
Simulation Data, UML: Unified Modeling Language, MES: Manufacturing Execution System, ERP: Enterprise
Resource Planning
Received: June 21, 2021. Revised: June 26, 2022. Accepted: July 22, 2022. Published: August 24, 2022.
1 Introduction
Digital transformation is the primary lever for
optimizing a sustainable economy in an industrial
context. Most industries currently only use 10% of
their data, preventing them from taking meaningful
action [8].
As a result, the first and most important step in the
industry's digital transformation is to collect and
manage all data, and it is necessary to be able to
retrieve and store all of this information in a single,
highly secure location.
The use of models of real or fictional systems, or
processes to better understand, or predict the
behaviors of a modeled system or process is known
as simulation modeling in engineering can reduce
costs, and consumption, shorten development cycles,
improve product quality, and greatly simplify
knowledge management.
This article will describe the developments of
Industry 4.0 and the Fourth Industrial Revolution that
have led to new simulation modeling paradigms,
embodied by the concept of digital twins, and will
validate the adoption of new simulation modeling
paradigms in the industrial and scientific
communities through multiple case study by
presenting real cases of Industry 4.0 method and
technology application [15].
WSEAS TRANSACTIONS on SYSTEMS
DOI: 10.37394/23202.2022.21.17
Agouzzal Kawtar, Abbou Ahmed
E-ISSN: 2224-2678
157
Volume 21, 2022
Today, the position of SMEs (small and medium-
sized enterprises) in Industry 4.0 is of particular
concern, because the level of automation of SMEs is
generally low and their funds are limited, but SMEs
represent an important part of the creation of jobs and
value [8].
Information and communication technologies
(ICT) and automation technologies are completely
integrated into the factory of the future. All systems,
including R&D, as well as business partners,
suppliers, original equipment manufacturers
(OEMs), and customers, are networked and
consolidated [8].
According to KPMG, manufacturing networking
and transparency enable a paradigm shift from
"centralized" to "local" production.
Manufacturing now employs "embedded
systems," which collect and transmit specific data. A
central computer in the future factory coordinates the
intelligent networking of all subsystems into a cyber-
physical system (CPS) capable of increasing
independence [15].
CPS refers to the networking of various integrated
software systems that collect and transmit data. A
paradigm shift from "centralized" to "local"
production is thus taking place, and a central
information system manages an intelligent network
while taking physical factors into account, such as the
capture of needs via man-machine interfaces that
allow independent process management [8].
The close interaction of the physical and virtual
worlds represents a fundamentally new aspect of the
manufacturing process known as "CPS" physical
production systems. Because many SMEs are still
using older proprietary systems, the emergence of
Industry 4.0 networking and integration standards, as
well as open standards architectures, has the potential
to benefit them.
SMEs can become temporary production
networks with precisely calculated value-added
contributions, allowing them to significantly reduce
their production management efforts while meeting
much higher market demands [8].
Furthermore, additive manufacturing and flexible
machines enable the production of very small series
and personalized products at unit costs previously
only possible in large series.
3 Green technology trends in Industry
4.0
Large battery-powered sensor networks can now
be built using new energy-efficient green computing
technologies. A major shortcoming is the general
lack of standards. Many aspects of Industry 4.0
technology are already in place, but some areas still
require binding international standards. Industry 4.0
is currently a concept rather than a product or service
that can be purchased [10].
This is due in part to the imprecise definition of
"Industry 4.0" and customers' exaggerated
expectations. What is certain is that Industry 4.0
necessitates the use of products such as industrial and
management software (CAD, virtual simulation
tools, etc.), processes, and devices (Ethernet,
robotics, relays, motors drives, sensors, switches,
etc.). These devices necessitate specialized
knowledge of information and communication
technologies (ICT) and automation technologies,
posing both a challenge and an opportunity for future
workforce educators and trainers [10].
4 The modern simulation paradigm
Connectivity of a simulation model in the modern
simulation paradigm typically involves integration
with a static database of business variables, a user-
friendly front end, and additional decision support
tools such as expert systems or group decision
support systems. Simulation has primarily been used
to create stand-alone solutions with limited scope and
lifespan. However, as computer simulation has
permeated various areas of business processes, the
need to link simulation models used in different parts
of an organization has arisen [5].
Furthermore, the trend in simulation development
has shifted from purely analytical and optimization-
oriented models to the integration of simulation
models into decision support tools that will be used
on a regular basis. A common distributed simulation
system, for example, can be built by integrating
models of various parts of an organization to perform
large-scale business system simulations, providing a
complete view of the modeled organization.
As a result, design requirements have shifted from
stand-alone models accessible only to simulation
experts to models that can be connected to, and even
modified by, user-friendly interfaces or other
applications [7].
Figure 1 depicts a diagram of such a system.
WSEAS TRANSACTIONS on SYSTEMS
DOI: 10.37394/23202.2022.21.17
Agouzzal Kawtar, Abbou Ahmed
E-ISSN: 2224-2678
158
Volume 21, 2022
2 The position of SMEs (Small and
Medium-sized Enterprises) in Industry
4.0
Figure 1: Schematic of a typical simulation modeling
based DSS system
5 The new "Digital Twin" concept
The best guess is that the new simulation
modeling paradigm is the concept of the "Digital
Twin," which we will discuss in the following
chapters.
The Digital Twin concept extends the use of
simulation modeling to all phases of the product life
cycle, where products are first developed and tested
in detail in a virtual environment, and subsequent
phases use the information generated and collected
by previous phases [11].
Manufacturers must shift away from traditional
design processes and practices that used a "build and
modify" approach, and instead adopt a more systemic
approach that has been an essential part of the design
process in the aerospace and automotive industries.
for a long time
To accelerate the development of model and scenario
releases, algorithms that create or modify simulation
models based on model input data can be developed.
This is especially useful for large simulation models
and when the variants of the model are prepared by
an algorithm, such as an optimization algorithm.
However, automated model creation and
modification necessitates that the model structure is
modifiable by an algorithm without the need for
manual intervention.
These four points summarize the main changes in the
simulation and modeling paradigms during the
transition from an autonomous simulation-based
decision support system to the Digital Twin:
▪ Integration and connectivity into a larger IS
(manufacturing or integrated management
software)
▪ A multi-level/resolution holistic approach,
including physical modeling, is used to model
the system.
▪ Several aspects of the simulation model
necessitate a high level of detail and a low level
of abstraction.
▪ Model creation and modification are fully
automated (data-driven) [11].
6 The Digital Twin simulation
A digital twin is a natural result of using a system
design approach to product development, and it can
be easily integrated into the final product for training,
online diagnostics, performance optimization, and
other purposes.
The functional model interface is supported by the
majority of industrial automation platforms as a
means of integrating Twin so that it can run in
parallel with the real machine.
Engineers can use simulation software to create a
virtual prototype of the machine design directly from
the CAD representation and integrate it as a Digital
Twin as a functional mock-up unit on their real-time
platform (FMU) [12].
According to Goossens (2017), the cost of
developing numerical models of multidisciplinary
systems has decreased significantly in recent years
due to the introduction of powerful and user-friendly
mathematical system modeling tools and general-
purpose modeling tools like MATLAB and Analogic.
Identifying and addressing signaling issues early in
the Twin’s development process is critical. Figure 2
depicts an example of a cyber-physical production
system incorporating Digital Twin simulation
modeling.
A Digital Twin simulation model of a process or
part of a process has several potential uses in an
organization:
▪ Without the expense of a dedicated training
simulator, an online digital twin allows an
operator to train on a virtual machine until they
have the skills and confidence to operate the real
machine. Using an in-line Digital Twin speeds up
the learning process while reducing the risk of
machine damage.
▪ A digital twin can be used to identify potential
problems with its real counterpart by combining
optimal control and model predictive control
techniques with advanced machine learning
capabilities. By detecting a drift between machine
performance and model behaviour, a high-fidelity
physical model running in parallel with the real
machine can immediately indicate a potential
malfunction of the real machine. The data can be
Business
System
Simulato
r
Scenarios
GSS
Scenario
Rank
Business
Database Simulation
results
ES
Simulation
Model
Agouzzal Kawtar, Abbou Ahmed
E-ISSN: 2224-2678
159
Volume 21, 2022
used to shut down and repair the faulty machine,
or the model can be used to provide a strategy for
compensating for a drop in performance without
slowing or stopping production.
▪ An integrated digital twin would serve as the
foundation for increasing machine self-
awareness, allowing it to optimize its own
performance for given duty cycles, diagnose and
compensate for non-catastrophic faults, and
coordinate operation with other machines while
requiring minimal operator intervention.
The business system simulator in such a system
employs a Digital Twin model of the business
process.
The digital twin is used to provide a detailed,
dynamically updated digital representation of the
actual business process to the business intelligence
toolkit (eg, a manufacturing plant).
Process data is collected in real-time by the array
of sensors and intelligent machines in the business
process, stored in the enterprise database, and then
transmitted to Digital Shadow.
The operation of the Digital Master model is
adjusted based on the data contained in the Digital
Shadow, enabling online optimization and decision
support, as well as process automation control,
thereby creating a control feedback loop, which is the
foundation of cybernetics.
Figure 2: Schematic of a system implementing the new
simulation modeling paradigm: a Cyber-Physical Production
System incorporating the Digital Twin, CPS
7 Methodology
Because SMEs are the slowest to adopt the
industry 4.0 paradigm, we chose a case aimed at
developing Industry 4.0 approaches and methods that
can be implemented in SMEs.
Finding case studies through reports of previous
studies allows for the exploration and understanding
of complex issues and can be regarded as a strong
research method, particularly when a thorough and
in-depth investigation is required.
In this case, we present a comprehensive study
aimed at exploring and reasoning about the global
phenomenon of the new simulation modeling
paradigm based on the Digital Twin concept, and
attempting to draw conclusions about the
phenomenon, particularly about the adoption of the
new simulation. The paradigm of modeling within
the framework of industry projects [17].
8 The case for adopting the new
simulation modeling paradigm
Implementing the new simulation modeling
paradigm and Industry 4.0 remains a significant
challenge for researchers and businesses. There are
new ways, however, to demonstrate the integration of
built models into general-purpose simulation
modeling tools, automate their construction and
modification, and implement these solutions without
major financial investments, which is a very
appealing prospect, particularly for SMEs.
In this chapter, we will look at how the new
simulation modeling paradigm is being implemented,
from the creation of a new high-level modeling
automation methodology to the creation of a digital
twin concept for SMEs [17].
9 Case 01: Automated creation of XML
templates
The use of a new method of automated DES
model building, which uses customer order data
obtained via SQL queries to modify the Extensible
Mark-up Language (XML) file containing the
simulation model, thereby changing the default
model structure, is the focus of this paper.
The paper describes the methods and outcomes of
a manufacturing process optimization project. The
authors built a model that reflects current
manufacturing processes and allows them to test
optimization methods using discrete event simulation
(DES) [22].
Due to a large number of products and
manufacturing processes, they developed an
automated model building method that builds an ad
hoc simulation model using customer order data and
the manufacturing process database.
Process
Automa
-tion
Business
process
Decision support tools:
machine learning
optimization …
Digital twin
Online
business
databas
e
Feedback loop
Digital
shadow
Digital
Master
WSEAS TRANSACTIONS on SYSTEMS
DOI: 10.37394/23202.2022.21.17
Agouzzal Kawtar, Abbou Ahmed
E-ISSN: 2224-2678
160
Volume 21, 2022
The model and method were put to the test in the
optimization task, which involved reducing product
travel distance through changes in factory layout and
employing a new heuristic optimization method
based on force-oriented graph drawing.
9.1 Description of the problem
Creating a static simulation model that covers all
possible products (i.e. 30,000) that could be included
in customer orders is not feasible because it takes
approximately 15 minutes to build a process model
for each product, and a model containing 30,000
processes also exceeds the memory limits of the
modelling tool used (Analogic,
http://www.anylogic.com/).
Manually modifying the simulation model can
take a long time, especially if a large number of
model variants must be built. In Analogic, the
simulation model is typically built by dragging and
dropping various blocks and connections onto the
canvas. Instead, for each set of open commands, an
ad hoc model-building method has been developed.
The method involves editing the XML file [19].
9.2 Results
Because orders change on a regular basis, the
authors created a method and an application in Java
that builds the model automatically from a template,
a database of technical procedures, and a database of
open or running procedures.
Based on the ordered products and technical
procedures, only the necessary machines are placed
in the model. Analogic saves models as standard
XML files, making it simple to manually or
algorithmically modify the model.
The XML simulation model file of Analogic
stores information on standard and user-defined
blocks and agents, connectors between blocks,
statistical monitors, input readers, output writers, and
so on.
Data is stored in a tree structure as elements
(nodes). An element may have several attributes that
describe the element's type as well as any parameters
that describe the element's properties. The attributes
may contain several lines of programming code
describing how the block operates in various
situations and states.
The Java application modifies the XML code to
modify the machine data and all other relevant
abstract objects connected to the machine blocks,
such as connectors, sources, and sinks. The Java
application reads the blocks in the template file and
copies them based on the input data. The following
procedure is used to add a new element (block) to the
model:
▪ Locate an XML node that represents a template
block.
▪ Make a copy of the node and connect it to the
original node's parent.
▪ Modify the copied block's data (block name,
position on the canvas, block properties, part of
the programming code, etc.)
▪ After that, the resulting XML structure is saved
in a new logical file. Figure 4 depicts the
transaction role of products and carts, which
consists of four main elements:
▪ Analogic environment simulation model of the
basic manufacturing process.
▪ The layout of the machines and paths was
generated from the factory's AutoCAD model.
▪ A Java program that generates the Analogic
XML model from a template file.
▪ Microsoft Excel as a bridge tool for storing and
analyzing input and output data
Figure 3: Schematics of a system implementing automated DES
modeling
10 Case 02: Modelling a virtual factory
based on standards
Jain and Le Chevalier describe a method and
proof of concept for automatically generating virtual
factory models from manufacturing configuration
data in standard data formats like XML.
In this context, the virtual factory represents a
multi-level high-fidelity simulation. A number of
initiatives, such as Smart Manufacturing and Industry
4.0, have identified modeling and simulation as
critical to the advancement of manufacturing.
Agouzzal Kawtar, Abbou Ahmed
E-ISSN: 2224-2678
161
Volume 21, 2022
Our proposals include the use of simulation at
multiple levels within manufacturing, with models
ranging from highly detailed physics-based models
of the manufacturing process to high-level DES and
SD-based supply chain models.
10.1 Description of the problem
Currently, developing a Digital Twin requires
significant resources and expertise, which limits
access to large corporations to the detriment of
SMEs.
The automatic generation of data-driven models
has the potential to reduce the need for expertise,
allowing simulation to be used more frequently.
The proposed method extends the scope of the
generated model to a virtual factory model rather than
a single-level model, and it complements existing
automation solutions by proposing the use of
standard data formats for the input data describing the
manufacturing system in question.
10.2 Results
Analogic is used as a simulation modeling tool in
the proof-of-concept method, with multi-level
modeling implemented using Java code for the
process model, agent-based modeling (ABM) for
the machine level, and DES models for the
cell/process chain level, a concept similar to that
described in (Rodi & Kan- du, 2015).
The integration of the various modeling methods
is accomplished using the Analogic tool's native
capabilities, resulting in a hybrid model. Historically,
hybrid modeling was accomplished by linking
models using intermediary software solutions [20].
Figure 4 depicts the proposed method's scheme. The
following describes how the proposed automatic
generation works:
1. Read configuration data from the manufacturing
system using a standard-format interface.
2. Retrieve information from machine parameters
and process levels.
3. Assemble the logical network at the plant or cell
level using the process plan data as input.
4. Connect the plant or cell level logic network to
each individual machine and process.
5. Create the template by using the corresponding
templates in the library.
6. Create an installation layout based on
configuration data information, with links to the
logical network.
7. Run the model with user-specified parameters
such as resolution level and output format via the
run interaction [21].
The data-based modeling interface works with
data in CMSD format, which is based on XML. A
Java parser has been created to browse a CMSD file
for the machine shop and collect the information
required to build the corresponding virtual factory
model automatically.
The goal is to generate a virtual factory model
automatically using data from the actual factory in
the applicable standard formats, with the option of
generating output data streams in other applicable
standard formats [22].
By generating a multi-resolution model and using
standard input file formats, the automatic generation
of the virtual plant model is intended to go beyond
previous efforts involving the automatic generation
of single-level plant simulations [23].
Figure 4: Standard data format-based modeling automation
system schematic
11 Case 03: Digital Twin for SMEs
Uhlemann et al. (2017) introduce a notion for
integrating a Digital Twin of the manufacturing
system within SMEs. Their concept is feasible
because it ensures adequate data quality while
minimizing investment costs and without
jeopardizing the benefits of the Digital Twin and
CPPS [2].
Their concept includes the proposed database
structure as well as guidelines for implementing the
Digital Twin in SMEs' production systems. Our
additional concept of the Digital Twin for a
production process enables the coupling of the
production system with its digital equivalent as a
basis for optimization with the shortest possible delay
between data acquisition and the creation of the
Digital Twin [4].
This enables the development of a physical cyber
production system, preparing the way for powerful
Agouzzal Kawtar, Abbou Ahmed
E-ISSN: 2224-2678
162
Volume 21, 2022
applications. Multimodal data acquisition and
evaluation must be performed to ensure maximum
agreement between the cyber-physical process and
its real model [5].
11.1 Description of the problem
Industry 4.0 methods are still underutilized in
manufacturing operations.
Furthermore, the authors describe the following
difficulties in realizing the Digital Twin as a
necessary prerequisite for a CPPS:
▪ Manual motion data acquisition is widely
used, despite the fact that it conflicts with the
required real-time availability.
▪ Manually acquiring motion data snapshots
limits the simulation's potential.
▪ A central information system is required in
conjunction with decentralized data
acquisition.
▪ Internal implementation of Industry 4.0
concepts is frequently inadequate.
▪ Slow data acquisition standardization in
production systems impedes the
implementation of agile and adaptable
systems.
▪ Data acquisition standardization has not yet
been achieved.
▪ The high costs of new IT environments are
impeding vertical industry 4.0
implementation.
▪ The coupling of simulation and optimization
is insufficiently guaranteed to fully exploit
near-real-time models, and there are data
security concerns.
▪ The collection of motion data, in conjunction
with data on employee activity and the
position and use of production machines, has
enormous potential for CPPS
implementation.
▪ Existing time-dependent position data
sources and databases are insufficient,
particularly in SMEs with a low level of
automation.
▪ A complete picture of the production system
can only be obtained with additional
information on employee and production
resource movement.
11.2 Results
The described concept is novel in comparison to
the approaches prevalent in large corporations, which
are focused on full automation. Because the
production database in SMEs is highly heterogeneous
and frequently insufficient for the realization of the
Digital Twin. Sensor tracking provides information
on the routes and positions of production employees
as well as large, highly mobile production equipment.
The necessary technologies and tracking systems
based on sensors, as well as extensive program
libraries for the implementation of machine vision,
are available on the market, making the proposed
concept feasible.
Sensor-based tracking should provide information
on the routes and positions of production workers as
well as large, mobile production equipment such as
forklifts, whereas image recognition can detect and
identify product types in production as well as
smaller machines.
Figure 5 depicts the digital system diagram and
the matching concept.
Because of the low level of digitalization of
manufacturing data in SMEs, automated collection of
machine data is not envisaged.
Furthermore, in the presented concept, the
collection of detailed machine data is not required.
The concept's innovative aspect is the incorporation
of widely adopted and commercially available
components that are already available as stand-alone
solutions.
Figure 5: Concept of the CPPS through the Digital Twin in
SMEs
12 Case 04: Automated modeling based
on ERP (Enterprise Resource
Planning) data
Kirchhof (2016) describes a practical case in
which entire simulation models of complex, large-
scale production in a just-in-time manufacturing shop
are automatically created from an automotive
company's SAP, ERP, and MES systems to support
operational planning goals and reduce operational
logistics risks.
Agouzzal Kawtar, Abbou Ahmed
E-ISSN: 2224-2678
163
Volume 21, 2022
12.1 Description of the problem
The time and effort necessary to manually create
and maintain such a comprehensive model are
frequently too great. Manufacturing systems
simulation has long been characterized as facing
substantial hurdles due to automatic model creation,
the resulting decrease in problem-solving cycles, and
the requirement for a higher level of data integration.
Operational manufacturing simulation models
depend on a lot of input data since they require a high
level of modeling. The utility of simulation for
operational planning purposes would be greatly
increased by the autonomous and on-demand
development of type manufacturing simulation
models from, for example, business data sources.
12.2 Results
Lean manufacturing practices are widely used in
the automobile sector, forcing businesses to strike a
balance between cost-effective inventory reduction
efforts and operational hazards like stock-outs on the
assembly line. In order to analyze and implement
countermeasures, the simulation model's goal is to
serve as an early warning system and identify
probable stock-outs before they happen. As a result,
the model's scope includes every internal logistic
operation of the business, from the selection of parts
in the warehouse to consumption on the assembly
line.
The SIMIO simulation program was used to
create a general simulation model of the factory floor.
A customized SIMIO extension was created to
facilitate modeling automation. It enables the
modification of a blank model by arranging certain
instances of the modeling parts in accordance with
the input data. By automatically positioning, linking,
and parameterizing the predefined model items in the
model, the addon may create whole flow shop models
[13].
The company's SAP and MES systems are used to
extract the data required to create the model using
customized data software that pulls the necessary
information straight from the databases of the
corresponding systems. The SAP system supplies the
necessary data, including workstation specifics,
routings, BOMs, shift schedules, production orders,
stock levels, machine master data, etc., to model the
production line. The manufacturing line and the
production process are covered in full by the MES
system.
The MES system offers comprehensive data on
production sequences, the status of ongoing and
planned production, and production order- and
workstation-specific production progress. The
resulting simulation system assists the company's
planning staff in averting logistical issues and
production hiccups by being fully integrated into the
operational planning process and IT architecture of
the business. When compared to a manual approach,
automating the models greatly shortens the problem-
solving cycle. The approach that is being given works
well for complex, large-scale models [27].
13 Case 05: Methodology for
automating high-level modeling
The MES system offers comprehensive data on
production sequences, the status of ongoing and
planned production, and production order- and
workstation-specific production progress.
The resulting simulation system assists the
company's planning staff in averting logistical issues
and production hiccups by being fully integrated into
the operational planning process and IT architecture
of the business.
When compared to a manual approach,
automating the models greatly shortens the problem-
solving cycle. The approach that is being given works
well for complex, large-scale models [9].
13.1 Description of the problem
The automation of analysis, training, and
solutions is the key obstacle. As the PAs already have
in-depth knowledge of civilian transportation
networks, they can provide directions alongside the
PAs on mobile devices. However, even if the PAs
had access to the company's information systems,
they would not be able to perform the duties
necessary for an experimental design environment
because the systems would not be complete and the
PAs would not be familiar with the products,
processes, resources, and proprietary facilities of an
arbitrary company.
Following the paradigm of separating system
description from system analysis, the authors made
an attempt to automate engineering workflows by
creating system descriptions in an ostensibly simpler
way and then automatically transforming them into
the semantics and syntax of a specific analysis
language as needed. The authors, however,
discovered that none of the system description
languages investigated for production systems was
adequate and made the decision to figure out a means
to support a multitude of related but distinct
Agouzzal Kawtar, Abbou Ahmed
E-ISSN: 2224-2678
164
Volume 21, 2022
languages. The lack of a language for discrete event
simulation analysis was another barrier [10].
13.2 Results
The approach outlined by Thiers et al. (2016)
places the majority of the transformation intelligence
in the model-to-model transformations themselves
but takes a small step forward by maximizing the use
of model library blocks, which are executable
versions of the linking abstraction model elements, to
construct simulation models. A middle phase known
as the Bridging Abstraction Model will be created to
allow the translation of requirements into the model
structure.
The Bridging Abstraction Model is an abstract
design that captures the fundamental similarities
shared by all discrete event logistics systems,
including supply chains, warehousing and
distribution, transportation and logistics, and health
systems.
A system model must be as inventive as necessary
for accessibility, the transition abstraction model
must be as abstract as possible for robustness and
reusability, and efficiency depends on how simple it
is to create and maintain mappings between the two.
The transition abstraction model is introduced to
mediate this fundamental tension between the
concrete and the abstract. [11].
The uniqueness of the methodology is found in its
approaches and instruments that deal with the main
research problems needed to make it applicable to
systems engineering, as follows:
• The Bridge Abstraction Metamodel is an explicit,
analytically neutral, machine- and human-readable
metamodel that encapsulates the fundamental
alienations that underlie all discrete event logistics
systems.
• Model-to-model transformations: The process has
evolved from applying UML stereotypes to
declarative specifications in generic transformation
languages, such as QVT and ATL, into a tailored
model-to-model transformation language and engine
(see Figure 3).
• Being aware of the range of production system-
related queries and the analyses that can address
them. This difficulty is closely related to in-depth
expertise in the subject.
- A new difficulty will emerge if and when the
methodology's implementation enables a sufficient
number of questions about the feasibility to be
resolved.
- How to ask the proper questions in order to make
effective use of a "question-answer genie," keeping
in mind that this would need documenting higher-
level processes (diagnosis, continuous improvement,
de-identification, etc.) and the questions posed at
each stage of their implementation.
The proof-of-concept software created to support the
model also provides answers to the following
engineering-related questions: - What is the
(anticipated) duration (gross cycle time) of a specific
(job)?
- What is the typical throughput for doing a particular
(task) on a regular basis?
- What is the anticipated throughput of producing a
specific (product) in a particular (facility)?
- How many resources at a minimum are required to
support a specific throughput?
The concept of the new modeling automation process
is depicted in Figure 6.
Figure 6: Novel modeling automation methodology described by
Thiers et. al (2016)
14 Description of results
The Digital Twin for SMEs represents the most
effective approach to implementing a new paradigm
of alternative approaches that have been established
for the fully automated advancement of simulators in
correlation with industrial systems in order to
develop models of system components at different
levels, especially in SMEs.
Agouzzal Kawtar, Abbou Ahmed
E-ISSN: 2224-2678
165
Volume 21, 2022
The scenarios studied in this article allow us to
draw the conclusion that both large and small
businesses are implementing the concept of digital
twinning, while there are important distinctions
between the issues they encounter and the strategies
and tools they employ to address them. The creation
of standardized procedures and architectures that
would permit integration into their R&D processes
and current ERP and MES solutions is a problem for
major research companies. SMEs are more focused
on using affordable, off-the-shelf simulation
modeling tools and commercially available sensors to
build proprietary automation that would allow them
to implement some of the concepts of energy
efficiency than on the purchase or development of
automation technologies, likely because of the
abundance of resources available.
The only method that could construct the
manufacturing process model for a given set of orders
in a reasonable amount of time was automated data-
driven model development.
Industry 4.0 techniques are currently underutilized in
production settings. On the one hand, contemporary
publications are addressing the issue of inconsistent
definitions of Industry 4.0. On the other hand, there
are widespread issues such as a lack of standards and
uncertainty over the economic rewards while dealing
with the need for often sizable investments.
Aspects of simulation-based engineering and
decision support systems can be developed using
unique solutions made possible by the research
reported in this paper. These methods automate
model development and solution identification. The
techniques and solutions made available allow for the
development of Industry 4.0 automation solutions
using the Digital Twin concept and generally
accessible sensor technologies, as well as the
automation of general-purpose simulation modeling
tools using data and standards from ERP/MES.
The process of moving from design to production is
one that involves many business considerations. The
research findings that have been presented here help
this process by allowing designers to understand how
their decisions will affect production considerably
earlier in the scheduled design cycle than is now
achievable.
The multilevel model still faces difficulties with
model validation, which can be difficult for
automatically constructed models, particularly for
multilevel models. The effects of both inherent and
extrinsic uncertainty must be taken into account
when validating each simulation model.
All physics-based process models must have their
application areas established and be evaluated against
actual machine operations. As a result of the multi-
level model's stacking of validity concerns over
multiple levels, one level of this type of model
depends on the outcomes of another. Before the
virtual factory and other multi-resolution models can
be utilized to help industrial decision-making, the
impact of the stacking of uncertainty must be
recognized and measured.
In order to create integrated multi-level system
models, the adoption of a new simulation modeling
paradigm in the research environment necessitates
tighter collaboration with industrial partners and
diversification of researchers' knowledge. The
examples given illustrate that there are sufficient
integration choices in the current generation of
general-purpose simulation modeling tools, therefore
the lack of tools is not a problem. Additionally, a
variety of approaches have been developed for the
automatic development of simulation models that
correlate to industrial systems. Barlas and Heavey
provide a comprehensive summary of these
approaches (2016). The ideas of Industry 4.0 and
Digital Twin provide academics with a new incentive
for greater collaboration and knowledge transfer
between study fields, as multi-level modeling
necessitates the integration of models created using
various approaches and tools.
References:
[1] Dejan Gradišar, Gašper Mušič, Automated
Petri-net modeling for batch production
scheduling, Petri Nets – Manufacturing and
Computer Science, 2012, pp. 3-26
[2] Harrell. C. R., Hicks. D. A. (1998), Simulation
software component architecture for simulation-
based enterprises applications, Proceedings of
the 1998 Winter Simulation Conference, 1998,
pp. 1717-1721
[3] Jain. S, Lechevalier. D, Standards based
generation of a virtual factory model,
Proceedings of the 2016 Winter Simulation
Conference, 2016, pp. 2762-2773
[4] Kannan. R. M, Santhi. H. M, Automated
construction layout and simulation of concrete
formwork systems using building information
modeling, Proceedings of the 4th International
Conference of Euro Asia Civil Engineering
Forum 2013 (EACEF 2013), 64, 2013, pp. C7-
C12
[5] Kirchhoff. P, automatically generating flow
shop simulation models from SAP data,
Agouzzal Kawtar, Abbou Ahmed
E-ISSN: 2224-2678
166
Volume 21, 2022
15
Conclusion
Proceedings of the 2016 Winter Simulation
Conference, 2016, pp. 3588-3589
[6] Kljajić. M, Škraba. A, Simulation Approach to
Decision assessment in Enterprises, Systems
Theory Simulation, (2002), pp.199-210
[7] Lattner. A, D. Bogon, T. Lorion, Timm. I. J, A
knowledge-based approach to automated
simulation model adaptation Spring Simulation
Multiconference,2010, pp. 200-207
[8] Borshchev. A The Big Book of Simulation
Modelling, AnyLogic North America, C5-C8,
(2013)
[9] Conner. W. R, Automated Petri net modelling of
military operations, National Aerospace and
Electronics Conference NAECON, 1990, pp.
624-627
[10] Forrester. J. W, Industrial Dynamics,
Cambridge MA MIT Press. 1961
[11] Goldsman. D, Nance. R. E, Wilson. J. R, A brief
history of simulation, Proceedings of the 2009
Winter Simulation Conference, 2010, pp. 310-
313
[12] Goossens. P, Industry 4.0 and the Power of the
Digital Twin., 2017, from Design News
Direct,2017, pp. 5-3
[13] Marolt.M, Lenart. G, Maletič. D, Kljajić,
Borštnar. M, Pucihar. A Business model
innovation insights, from a multiple case study
of Slovenian SMEs, 49(3),2016, pp. 161-171
[14] Merriam. S. B, Case Study Research in
Education, A Qualitative Approach. San
Francisco: Jossey-Bass, 1998
[15] Robson. C, Real World Research, A Resource
for Social Scientists and Practitioner
Researchers Oxford, 1993
[16] Rodič. B, Kanduč. T, Optimisation of a complex
manufacturing process using discrete event
simulation and a novel heuristic algorithm,
International Journal Of Mathematical Models
and Methods in Applied Sciences, 2015
[17] Rodič. B, Kljajić. M, Mobile agents and XML
for distributed simulation support, Simulation-
based decision support, 2005,
pp. 490-498
[18] Rosen. R, von Wichert. G, Lo. G,
Bettenhausen. K. D, (2015). About The
Importance of Autonomy and Digital Twins
for the Future of Manufacturing, IFAC-Papers
Online, 2015, pp.567-572
[19] Schwab. K. The Fourth Industrial Revolution,
Geneva: World Economic Forum, 2016
[20] Stark. R, Kind. S, Neumeyer. S, Innovations in
digital modelling for next-generation
manufacturing system design. CIRP Annals -
Manufacturing Technology,2017
[21] Thiers. G, Graunke, Christian. M, Automated
production system simulations using
commercial off-the-shelf simulation tools,
Proceedings of the 2016 Winter Simulation
Conference, 2016, pp. 1036-1047
[22] Yin. K, Case Study Research and
Applications: Design and Methods, London:
SAGE. (2017)
[23] Zainal. Z, Case study as a research method,
Journal Kemanusiaan, 2007, pp.1-6
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the Creative
Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en_US
Agouzzal Kawtar, Abbou Ahmed
E-ISSN: 2224-2678
167
Volume 21, 2022