Digital Twin Framework Development for Apparel Manufacturing Industry PDF Free Download
1 / 123/123
100%
Digital Twin Framework Development for Apparel
Manufacturing Industry
A Thesis
Submitted to the Faculty of Graduate Studies and Research
In Partial Fulfillment of the Requirements
For the Degree of
Master of Applied Science
in
Industrial Systems Engineering
University of Regina
By
Mohammed Didarul Alam
Regina, Saskatchewan
February, 2023
©Copyright, 2023: Md. D. Alam
UNIVERSITY OF REGINA
FACULTY OF GRADUATE STUDIES AND RESEARCH
SUPERVISORY AND EXAMINING COMMITTEE
Mohammed Didarul Alam, candidate for the degree of Master of Applied Science in Industrial
Systems Engineering, has presented a thesis titled, Digital twin framework development for
apparel manufacturing industry, in an oral examination held on January 12, 2023. The
following committee members have found the thesis acceptable in form and content, and that the
candidate demonstrated satisfactory knowledge of the subject material.
External Examiner:
Supervisor(s):
Committee Member:
Committee Member:
Chair of Defense:
Dr Niamat Hossain, Arkansas State University
Dr Golam Kabir, Industrial Systems Engineering
Dr Sharfuddin Khan, Industrial Systems Engineering
Dr Mohammad Khondoker, Industrial Systems Engineering
Dr Abdul Bais, Electronic Systems Engineering
i
ABSTRACT
Apparel manufacturing, being a labor-intensive industry, is evolving into a more
complicated, challenging, and dynamic business as a result of fast-changing fashion trends,
increased variety, and increased personalization of product demands. Quick and optimized
decision-making is extremely important to overcome these barriers. It is necessary to have
a mechanism that can ensure real-time visibility of the production process and provide
assistance in decision-making.
Digital Twin (DT) provides real-time visibility. It keeps all production data sources
in connection, allowing for fast analysis without affecting physical setup. Although most
DT ‘research focused on the machine level and automated industries, this has the potential
of applying to apparel manufacturing industries as well. This research contributes in
several ways to the current literature. 1. This study identifies the research gap and
demonstrates its application to apparel manufacturing industries. 2. It develops a
methodology describing step-by-step guidance for applying apparel manufacturing plants.
3. A case study is instantiated to authenticate the methodology. Using the suggested idea
and technique, the case study creates a DT of a sewing assembly line. It collects real-time
data and carried out simulations dynamically to reduce bottleneck operations. It also
communicated production downtime with users instantly to reduce downtime by
responding quickly.
Keywords: DT; Apparel Manufacturing; Sewing Assembly Line; Production Efficiency;
Production Downtime.
ii
ACKNOWLEDGMENTS
Special thanks to my respected supervisor, Dr. Golam Kabir, whose presence has always
been a strength to do my study. Undoubtedly, this dissertation would not have been
possible without his help and guidance.
I also acknowledge the support I received from the Faculty of Graduate Studies and
Research of the University of Regina for providing the awards, scholarships, and positions
to support the students. They have been providing me with awards, scholarships, teaching
assistantships, and many more.
I am also grateful to my fellow mates, and friends for their support in different parts of my
life, and mostly to my family for their love, which always inspires me to achieve greater
things in my life.
iii
DEDICATION
To my mother, Monowara Begum, my father, MD Abul Kashem, my wife, Reduana Kafil,
who is taking care of my two lovely sons, Aadheen Abdullah and Arham Abdullah.
I am also grateful to my elder brother, Dr.A.N.M Jane Alam for his continuous support
throughout my life. I am also thankful to My father -in-law, Engr. Kafil Uddin, and my
mother-in-law, Farida Begum, for their support.
My thankful appreciation to my extended family members, relatives, my friends, and well-
wishers. Thank you for your endless love, support, and encouragement.
iv
TABLE OF CONTENTS
ABSTRACT ....................................................................................................................i
ACKNOWLEDGMENTS ............................................................................................ ii
DEDICATION ............................................................................................................. iii
List of Tables ................................................................................................................ vi
List of Figures ..............................................................................................................vii
List of Abbreviations ................................................................................................. viii
CHAPTER 1: INTRODUCTION ................................................................................. 1
1.1 Background and Motivation ...................................................................................1
1.2 Statement of the Problem .......................................................................................4
1.3 Research Objectives and Contributions ..................................................................6
1.4 Organization of the Thesis .....................................................................................7
CHAPTER 2: LITERATURE REVIEW ...................................................................... 8
2.1 Current Methods and Research Gap .......................................................................9
2.2 DT Concept ......................................................................................................... 11
2.3 DT History & Current Works ............................................................................... 11
2.4 Digitalization in Apparel Manufacturing .............................................................. 21
CHAPTER 3: PROPOSED METHODOLOGY ........................................................ 27
3.1 Introduction ......................................................................................................... 27
3.2 DT Concepts for Apparel Manufacturing Plants ................................................... 28
3.3 Methodology of Developing DT for Apparel Manufacturing ................................ 39
CHAPTER 4: DIGITAL TWIN IMPLEMENTATION IN AN APPAREL
MANUFACTURING ................................................................................................... 45
4.1 Introduction ......................................................................................................... 45
4.2 Current Methods and Practice of Assembly Line Balancing ................................. 45
4.3 Objective of the Case Study ................................................................................. 49
v
4.4 Case Study ........................................................................................................... 50
4.4.1 Observable Manufacturing Elements (OME) ................................................. 54
4.4.2 Digital Elements ............................................................................................ 55
4.4.2.1 Process Map ............................................................................................... 60
4.4.2.2 Data Collection and Processing .................................................................. 62
4.4.2.3 Simulation Execution ................................................................................. 64
4.4.3 Integration ..................................................................................................... 79
4.4.4 Model Validation .......................................................................................... 80
CHAPTER 5: RESULT AND DISCUSSION ............................................................. 82
5.1 Implication of the Study ....................................................................................... 84
5.1.1 Implications in Theory .................................................................................. 84
5.1.2 Implications in Practice ................................................................................. 86
CHAPTER 6: CONCLUSION .................................................................................... 88
6.1 Limitations of the Study ....................................................................................... 90
6.2 Future Work......................................................................................................... 91
Reference ..................................................................................................................... 92
APPENDIX ................................................................................................................ 103
vi
LIST OF TABLES
Table 2. 1 Summary of recent research with contribution and focused area 27
Table 2. 2 Research contribution in apparel manufacturing digitalization. 34
Table 4.1 Operation Bulletin of the Product. 60
Table 4.2 Downtime collection of Day-1 67
Table 4.3 Downtime incorporation into Cycle Time (CT) 69
Table 4.4 Cycle time incorporated with Downtime of day-1 70
Table 4.5 Comparison between actual output vs the simulated output. 74
Table 4. 6 Cycle Time (CT) of day 1 after DT implementation 78
Table 4.7 Difference between actual production out and after DT implementation 83
vii
LIST OF FIGURES
Figure 3. 1 Elements of DT and their integration ........................................................... 36
Figure 3.2 Illustration of Apparel Manufacturing DT ..................................................... 38
Figure 3. 3 Development method of DT for apparel manufacturing ................................ 40
Figure 4.1 A sewing assembly line (Source: A apparel manufacturing company in
Bangladesh) ................................................................................................................... 51
Figure 4. 2 An overview of the DT for the case study .................................................... 53
Figure 4.3 Sketch of the final product produced from the assembly line. (Source: A
apparel manufacturing company in Bangladesh) ............................................................ 59
Figure 4.4 Process map of the woven shirt sewing assembly line. .................................. 61
Figure 4.5 Simulated view of the OME .......................................................................... 68
Figure 4.6 Simulated output of Day-1 ............................................................................ 69
Figure 4.7 Bottleneck operations of Day 1. .................................................................... 73
Figure 4.8 Bottleneck operations after DT implementation of Day-1.............................. 76
Figure 4.9 Production output at DT implemented scenario of Day-1 .............................. 76
Figure 4.10 Bottleneck operations of Day-11. ................................................................ 77
Figure 4.11 Bottleneck operations of Day-11 after DT implementation. ......................... 77
Figure 4.12 Simulated Production output of Day-11 ...................................................... 78
Figure 4.13 Production output after DT implemented scenario of Day-11. ..................... 78
Figure 5.1 Comparison of the production output among different scenarios. .................. 83
viii
LIST OF ABBREVIATIONS
BS Bundle System
BI Business Intelligence
CT Cycle Time
CDT Cognitive Digital Twin
CPS Cyber Physical System
DT Digital Twin
DTS Digital Twin shopfloor
DTPD DT driven product design
FG Finished Goods
IoT Intranet of Things
OME Observable Manufacturing Elements
PLM Product Lifecycle Management
PBS Progressive Bundle System
RFID Radio Frequency Identification
RAMI Reference Architectural Model Industries
TQM Total Quality Management
VSM Value Stream Mapping
WIP Work-in-Process
3D Three-dimensional
1
CHAPTER 1: INTRODUCTION
1.1 Background and Motivation
To satisfy customers in a competitive market, manufacturing organizations must
consistently bring innovation, enhance quality, and reduce cost and lead times. They can
accomplish these objectives thanks to digitization. Factory data, communication methods,
and common interpretations are all crucial for achieving such digitalization and integration
inside a manufacturing factory. In turn, a free flow of communication between various
systems including software and hardware often depends on the correct implementation of
tested techniques (Luo et al., 2019). One of the biggest problems facing manufacturing
facilities is the exploitation and use of modern technologies in production (Cwikła & Foit,
2017). As a result, the industrial facilities are pressing for a change in strategy. At all levels,
these facilities are attempting to modify a number of planning and management decisions,
leading to the introduction of smart manufacturing systems. Zheng et al. (2019) suggest
that industrial uses of intelligent design, intelligent planning, intelligent machining,
intellectual monitoring, and intelligent control are all parts of smart manufacturing systems.
Any of the decision-making hierarchy levels may use artificial intelligence,
machine learning, deep learning, Radio Frequency Identification (RFID) or sensor-based
automation, big data analytics, the internet of things (IoT), and more to make it efficient
and intelligent. The emergence of the fourth level of industrial automation, known as
Industry 4.0, is marked by the increasing use of smart manufacturing. Industry 4.0,
sometimes referred to as the "Fourth Industrial Revolution," "smart manufacturing,"
"industrial internet," or "integrated industry," is a term used to describe a contemporary
2
age characterized by the development of cyber-physical systems based on the fusion of
heterogeneous data and knowledge. Numerous technologies and related concepts are
included in Industry 4.0 (Lu et al., 2016). The fundamental elements that provide the
standard quality for digital connection and communication between physical and digital
components might be referred to as Industry 4.0, key enabling technologies and
characteristics of today's digitalized manufacturing. The market for intelligent products,
three-dimensional (3D) printing, and autonomous cars is dominated by Industry 4.0
(Ivanov et al., 2019). Industry 4.0, however, lacks a systematic structure that might be
easily distinguished. It differs significantly between industries. When it comes to labor
intensive industries where many activities are needed to carry out manually, digitalization
is even more challenging. The manufacture of clothing involves a lot of manual labor.
Currently, industry 4.0 is a work in progress. However, the route for industry 4.0 is being
shaped by all of the smart production technologies. One of the related ideas that is assisting
the industry is the DT. Conceptually, digital twinning has been put out as an approach to
data-driven smart manufacturing that aims to achieve deep cyber-physical integration by
digitally representing physical elements. The idealization of DT, as opposed to pure
simulations, is often searching for dynamic optimization technologies that provide real-
time system reflections, interaction between real and virtual worlds, and automated model
evolution with new data feed (Tao et al., 2018).
The clothing industry faces greater obstacles than ever before to achieve anticipated
profits. This is due to the continuous increase in total manufacturing costs and asking for
further price reductions from customers. One of the primary causes of this predicament is
the inability to effectively digitalize processes to compensate for rising labor costs. In
3
addition, recent fashion trends of customization compelled the manufacturer to produce a
wide range of products with low quantities of each style (Liu et al., 2020). This resulted in
decreased efficiency due to the increased number of style changeovers in production
lines. Frequent changeovers resulted in line imbalances that cannot be properly balanced
within a short span of the style's life cycle. The performance of workers is inconsistent
throughout the life cycle, making it more difficult to achieve line balancing. In addition,
production downtime caused by defective products, long time to set up machines for a
specific process, operator training for a particular process, machine breakdown, and slow
response to repair has a negative impact on efficiency (Flores et al., 2021).
The efficiency of the production lines is crucial to the success of a garment
manufacturing company. Line balancing, preventative maintenance, rapid changeover,
optimizing the arrival time of raw materials, material flow, inventory management,
facilities layout, off-line simulation, lean manufacturing, lean six sigma and data
integration are typical methods for increasing manufacturing plant efficiency (Rajput et al.,
2018; Flores et al., 2021). From the perspective of a production manager, it often takes a
significant amount of time to collect all the data, analyze it, and respond appropriately.
During this period, the production line may switch to a different style. Even though a great
deal of study has been conducted on this issue, none of it helps them attain the requisite
level of effectiveness. First, these current tools are oversimplified, since they do not
account for real-world constraints or uncertainties present on the ground. Consequently,
the procedures cannot be simply duplicated or modified. Furthermore, these methods lack
the flexibility to analyze uncertain scenarios. The "simulation" approach is often employed
on the factory floor as a solution to these issues. Plants often accept simulation as a fast
4
and easy way to solve problems, but it is rarely employed for major choices. In the absence
of real-time data collection and an autonomous decision-making process, manufacturing
managers are not being able to take swift action to mitigate the negative impact (Jung et
al., 2022).
Consequently, Data is collected by a company not only to carry out simulation but
also to use it to increase efficiency that can account for real-time conditions. The
simulation's outcomes must be more realistic so that users can visualize them. There is
more to DT than just monitoring the performance of remote equipment and activities. The
information that is gathered by DT as well as the value that is offered by DT may and can
be utilized to advance the operational excellence of organizations (He & Bai, 2021). DT
provides businesses with visibility and knowledge into the operation of their devices at the
most fundamental level. The information collected by RFID or sensors enables firms to
have the necessary components or personnel on hand before issues become unmanageable.
The acquired data may be used to forecast the future, allowing businesses to alter their
business operations to be more intelligent. In its most basic form, DT can be considered as
a management modalities that may be applied to enhance the quality of business decisions
(Latif & Starly, 2020).
1.2 Statement of the Problem
The key to a manufacturing company's success is its ability to function effectively.
If efficiency and productivity do not steadily increase, it would be impossible for a clothing
manufacturer to remain competitive in the market. For this reason, management should
place a high focus on improving production processes before anything else. Tracking real-
time production data and being aware of the machine's state enables production managers
5
to make efficient decisions. A DT can provide production managers with this level of
visibility into potential decision possibilities. Nevertheless, industrial businesses are unable
to use DT due to several implementation obstacles. Fuller et al. (2020) identified the
following challenges:
1. The first significant obstacle is the overall IT infrastructure.
2. The second obstacle is the relevant data needed for a DT. It should be enriched and
pristine data sent in an uninterrupted, unbroken flow.
3. Privacy and security concerns related to DT provide a hurdle. The issues related to trust
arise from both an organizational and a user perspective.
4. Due to the lack of a standardized method for modeling, the next obstacles in all kinds of
DT development are associated with the modeling of such systems.
5. From the original concept through the simulation of a DT, a standardized strategy,
whether physics-based or design-based, is required.
6. In terms of modeling a DT, ensuring that domain-specific information is available and
provided to every level of development and function is a further challenge stemming from
the demand for consistent usage.
Labor-intensive manufacturing processes present unique challenges for DT
creation. To create effective DT for factories, developers need to take a methodical
approach. There aren't enough real-world instances of DT being used to address problems
with complicated factory layouts. Because permitting DT will boost growth in
manufacturing facilities, addressing these difficulties is crucial.
For the development of DT, labor-intensive manufacturing processes provide
distinct obstacles. The creation of efficient DT for apparel factories requires a deliberate
6
approach from developers. There are not enough instances of DT being employed to solve
difficulties with apparel manufacturing processes in the real world. Because allowing DT
would increase the expansion of industrial facilities, it is essential to solve these obstacles.
1.3 Research Objectives and Contributions
This study intends to establish a framework for implementing DT in an apparel
manufacturing facility. Using this research as a guide, any plant manager or stakeholder
may change a garment manufacturing facility into a DT by resolving numerous difficulties
related to particular problem situations. The primary objectives of the study are given
below:
● To perform a literature review to identify the implementation challenges in apparel
manufacturing industries
● To develop a framework for implementing the DT for apparel manufacturing
industries.
● To perform a real-life case study to illustrate the effects of DT implementation on
a clothing assembly line.
The primary contribution of this study is a tested and authenticated approach for the DT
at the clothing making facility. The statement of the problem has been responded through
the below contributions:
1) The research shows that a DT approach has to be institutionalized for the apparel
making industry. This study exhibits application of DT at the apparel making
production floor.
2) The literature has paid little attention to the DT approach in the
apparel manufacturing setting. Industry and solution providers are seeking a
7
standardized technological strategy to enable the DT. In response, a development
technique for DT has been presented.
3) In this research, a case study is carried out to validate the presented methodology by
addressing real-life issues. The outcomes of the case study demonstrated that the
overall improvement reinstates the benefits of DT applications.
4) The study described how the real-time status, integration, and simulation impact the
DT development in manufacturing plants.
5) The real-life case study illustrates the integration of current manufacturing processes
within an apparel manufacturing facility. In addition, it provides a mechanism based
on suggestions which is capable of proactively identifying possible bottlenecks in
order to enhance the efficient utilization of resources.
1.4 Organization of the Thesis
The research is divided into six chapters. The problem along with the previous
literature and DT of a manufacturing facility is outlined in the first chapter. The second
chapter examines the existing concept of DT and the resulting research gap. A development
technique with detailed stages is presented for DT of clothing making plants in Chapter 3.
The fourth chapter presents a case study illustrating the integration of several operations,
equipment and machines, and operators of garment making into the current arrangement.
Results and outcomes of the case study are analyzed and discussed in the fifth chapter.
Finally, the last chapter underlines the research's conclusions and limitations and suggests
future research.
8
CHAPTER 2: LITERATURE REVIEW
An abundance of research has been done on how to improve the garment industry. Yet,
these factories are inefficient. Inflexible and lacking in visibility, traditional shop floor
equipment and software cannot keep up with the demands of the modern business world.
Production facilities have a significant influence on profit or loss, customer satisfaction or
annoyance, streamlining activities with precision or devolving into chaotic surges of haste
projects, unanticipated postponements, and material shortages. For this reason,
management should set production facility enhancements as their top priority when
establishing goals for digital efforts (Manglani et al., 2019). All the functional areas need
to be connected well for efficient operation. Focusing on a particular function may help to
achieve the functional objectives but the overall enterprise goal may not be able to achieve.
Putting the company's broad strategic goals in context is essential.
As consumer choices are changing rapidly, apparel manufacturers must achieve the
ability to produce fashionable clothing products at affordable prices with the right quality
and lead time. These put manufacturers in a challenging situation to efficiently produce the
products, reducing the non-productive time from the manufacturing processes. The DT can
play a crucial role in this circumstance. As the DT represents the physical world virtually,
it enables the production managers with real-time information and the status of the
manufacturing operations and equipment (Zheng et al., 2019). DT has the potential to
identify the bottleneck in manufacturing operations dynamically, and enhances the
enterprise's overall efficiency and functionality by providing recommendations.
This chapter takes a look at the literature from the perspective of an apparel
manufacturing facility. Current technologies or methods, their limitations, and potential
9
future developments are all investigated. The research gap and the DT's function in filling
it are both highlighted in this chapter.
2.1 Current Methods and Research Gap
The apparel manufacturing industry has been evolving since its creation. With the
development of the industrial revolution, this industry also adopted many concepts of those
revolutions as well. In the past, each department at the clothing factory only employed one
kind of machine. In contrast, modern plants must contend with a widening variety of
activities, requirements, and complications owing to modern society's fast evolving culture,
political world, and economy have a significant impact on the industrial surroundings,
particularly consumer sectors like textile and apparel (Bae & May-Plumlee, 2005).
Previously, the production Bundle System was followed in the production line, which is
the first bulk clothing manufacturing method. Progressive Bundle System (PBS) is another
kind of Bundle System. Skill centers are formed by clustering PBS workstations. A
collaborative sewing system is known as a Modular Production System, where a garment
serves as the work unit. To prevent the movement of component bundles, the process is
supplied with single-ply components for a single garment (Lin et al., 2002). The idea of
Method Time Management first appeared in 1948 and was gradually adopted by businesses
over the years that followed. The manufacturing sector felt the effects of the Toyota
Production System, an approach to managing production, in the 1980s. Total Quality
Management (TQM) is a management concept that emphasizes catering to the needs and
wants of customers while also paying close attention to the efficiency, effectiveness, and
adaptability of production. At all levels most of the manufacturing firms of the world have
been profoundly impacted by the manufacturing thinking arising from Lean, TPS and
10
TQM. Although these approaches are efficient and straightforward, they lack specific
instructions. These methods are fairly straightforward and holistic in nature. The methods
can enable global policymaking, but they do not facilitate visualization or improvisation at
the machine level.
Algorithm-based approaches to problem formulation include things like
Aggregated Single Objective Algorithms, Simulated Annealing, Genetic Algorithms,
Response Surface Methodology, and Stochastic Optimization. Although they have been
utilized to address a variety of production issues, their suitability to a wide range of settings
has not been given much thought. Further, not all of these methods have been demonstrated
in real-world examples. There is a tendency for case studies to be unduly simplistic.
Making customization for a specific industry is challenging. Too much time and resources
are needed to run these algorithms, and they aren't well visualized (Latif & Starly, 2020).
Therefore, in the real manufacturing facility, the circumstances are always evolving,
uncertainty always emerges, and complexity is constantly growing, none of these
algorithm-based solutions have had much of an impact.
Manufacturing plants use simulation. This validates product life cycle behavior. It
also facilitates cost-integrated design and manufacturing modeling at concept design. This
technology simplifies digital manufacturing simulation. Simulation can solve
manufacturing plant issues easily. Simulation is too basic. It just analyzes the ideal
condition when it generates the output without concern. It skips inter-process uncertainty
and assumes an ideal scenario. It doesn't have a vital section where every decision takes
effort and modification. Thus, a simulation combining machine learning and predictive
analysis can transform the game. Here, the innovative concept of the DT can contribute a
11
vital part. DT has the potential to close the distance between the current requirements of
industry and the technologies that are accessible. The DT has a wide range of applications
and a great deal of potential for integration. It is easy to do, can be adapted, and is quite
convenient (Latif & Starly, 2020).
2.2 DT Concept
The DT is the digital and virtual counterpart of a physical setup which can be used
to imitate the physical elements for multiple objectives, utilizing real-time
synchronization of the sensed data originating from the field. For a manufacturing plant,
the digital representation of the manufacturing plant, also known as the simulation of the
production facility, covers everything from layout, machines, the material flow, and the
rate at which parts arrive, the rate at which equipment breaks down, etc. In a manufacturing
plant, a DT is a dynamic profile of the process that offers acumens built on real-time and
practical data from various process and business levels. For example, the virtual
representation for DT can be provided by a simulation of a manufacturing plant that
incorporates data collected in real time. It will be possible to build a full DT of a
manufacturing plant by incorporating AI-based recommendations, the combination of
various software to share information, and process optimization.
2.3 DT History & Current Works
University of Michigan demonstration to plant for a Product Lifecycle Management
(PLM) center introduced the DT. The first executive PLM courses at Michigan employed
this conceptual paradigm in 2002. In 2003, Professor Grieves first proposed the definitive
concept of DT that was merely a virtual depiction of a real thing during that period. Until
12
2012, NASA released a mechanism named “Modeling, Simulation, Information
technology, and Processing”, which presented a succession of fresh viewpoints for the
concept’s progress. In the meantime, DT contains principally the physical replicas in the
physical environment, 3D computer-generated models in digital settings and data
connecting both places. In 2013, Air force of US introduced the notion of digital thread,
which is developed based on simulation and modeling techniques to flow data and
information during entire product life cycle including designing product, utilization of
process, manufacturing, and maintenance of equipment. During the period of 2014 to 2017,
DT is adopted and generalized by large firms and promoted swiftly by General Electric,
Siemens and Dassault. Gartner identified DT as top ten strategic science and technology
(Grieves & Vickers, 2016; Zhang & Zhu, 2019).
The DT notion is viewed by Boschert and Rosen (2016) as a modification of current
system modeling and methodologies of simulation based on real product usage data and
information. In its infancy, DT encompassed the entire product life cycle, from start to end.
Thus, it provided a prototype-based, comprehensive description of the current and future
life cycle stages of a product. Tao (2017) proposed the conceptual model and specific
operation mechanism of DT Shopfloor (DTS), and predictive maintenance and monitoring
a machine's condition for a CNC. In another study, Tao et al. (2018) developed a
framework of DT-driven product design (DTPD), and described the concept of using this
framework to redesign a bi-cycle. Qi and Tao (2018) examined the data sources, data
processing, and data applications of big data and DT in manufacturing. The similarities and
contrasts between them in manufacturing are investigated from both a general and a
statistical standpoint. The integration of the DT, big data, and services to promote smart
13
manufacturing is illustrated. In another research, DT is analyzed on the basis of big data in
managing product lifecycles. It is stressed and researched how to produce and use
convergent cyber-physical data to better serve product lifecycle in order to drive more
efficient, intelligent, and sustainable product design, manufacture, and service (Tao et al.,
2018).
Wang and Wang (2019) designed the concept of DT-based system for used
electrical and electronic equipment that is designed to assist manufacturing and
remanufacturing processes over the whole of the product's life cycle, beginning with
product design and ending with product recovery. Within the context of a welding
manufacturing line, a suggested DT application framework for product lifecycle
management is presented. Later, the DT concept seeks to maximize product utilization and
manufacturing plant servicing and maintenance activities. During product service and
maintenance, condition monitoring and operations/service support utilizing all of the
available historical data and models are the primary focuses of this endeavor such
as manufactured and maintained state (Zheng et al. 2019).
Liu et al. (2019) developed the framework and mechanism of DT to enhance the
performance of the production system of sheet material with regards to throughput time,
working inventory between machining and assembly. Botkina et al. (2018) presented an
analysis about the DT representing a cutting tool in a separate publication. DT is shown as
a digital reproduction of a physical tool, including the instrument's information flows, data
management, and data format and structure, as well as additional application and
productivity analysis options. DT has been utilized in a limited capacity in the field of CNC
machines. Luo et al. (2019) presented a modeling method of DT for CNC machine tools.
14
Later on, a DT framework is presented to aid the installation process of maintenance of
CNC machines in a large-scale manufacturing facility by O’Sullivan et al. (2020). Liu et
al. (2019) worked on marine diesel engines and offered a comprehensive execution
procedure of DT technique for the critical components of it.
Some researchers worked to improve the performance of robots and its operations.
Park et al. (2019) devised and implemented a DT in order to overcome the challenges
associated with many product groups, dynamic scenarios of customized production, and
dispersed manufacturing systems that include heterogeneous manufacturing systems.
Another study of them developed the DT framework of a cyber physical logistics system
for supply chain control, and this for automobile parts to verify the effectiveness (Park et
al., 2020). Additionally, DT with horizontal coordination for reinforcement-learning-based
production control is applied in a re-entrant job shop in another research (Park et al., 2022).
A DT architecture that permits the interchange of data and information between a remote
emulation or simulation and its physical counterpart is proposed. The architecture consists
of various layers, including a local data layer, an IoT Gateway layer, cloud-based
databases, and a layer with emulations and simulations. This architecture provides a real-
time, service-based infrastructure for both vertical and horizontal integration. For the
purpose of evaluating the architecture, it was implemented for a tiny, but typical,
component of a physical manufacturing system (Redelinghuys et al., 2020). A framework
of DT is presented for sustainable intelligent manufacturing (He & Bai, 2021). In another
paper, a conceptualization of the robots and their respective active and passive auxiliary
equipment that are automatically positioned in a digital plant model using information
obtained from the currently active configuration (Braun et al., 2020; Talkhestani et al.,
15
2020). The connection of DT with a Cyber Physical System (CPS) in a closed-loop method
that shares both data and models is one of the ways in which a DT framework might
facilitate cognitive application development. A case is provided that demonstrates the
prospects for applying DT-driven machine learning, and the advantages that arise when the
idea is used in industrial practice are discussed. This is done in conjunction with the
presentation of the case (Alexopoulos et al., 2020). DT was built for real-time optimization
of motion planning of robotic cell to save energy consumption (Vatankhah Barenji et al.
2021)
Researchers also focused on applying DT in various manufacturing areas, though
Lattanzi et al. (2021) found lack of clear guidance and instruments for bringing the ideals
of the DT into actual production settings after reviewing ideas of DT, and their current
implementation scope in practical industrial processes.
A methodology known as a DT is proposed by Nikolakis et al. (2019) as a
component of a broader CPS, with the goal of improving the planning and commissioning
stages of human-based industrial processes via the use of simulation-based methods. The
principles, framework, configuration and operating method, as well as the real-time data-
driven operations control of DT-CPPS are discussed in another study (Ding et al., 2019).
An adaptive, simulation-based, DT was developed to improve managing and carrying out
operations; maximizing the use of available resources; minimizing delays in service;
adhering to strict deadlines, and also concluding that DT needs to be use case specific (Latif
& Starly, 2020).
A reference architecture for Cognitive DT (CDT) is suggested, and it draws
inspiration from the Reference Architectural Model Industries 4.0 (RAMI4.0) as well as
16
other current architectural styles. Several use cases for CDT and its accompanying
technologies are discussed. Implementation barriers of CDTs are identified. These are
interoperability issues, issues of project administration, data privacy/security, and
intellectual property protection and so on (Zheng et al., 2021). Polini and Corrado (2020)
presented a DT tool in 2020 to enable the lightweight construction of assemblies made of
composite material. This is accomplished by creating the geometrical deviations of the
pieces that are manufactured. A data interchange takes place between the created virtual
model and the physical operations in order to bring the produced digital representation up
to date with regard to the geometric deviation of the in-process components and the
assembly result. Another research focused on DT modeling of product, process and
operation are offered to demonstrate the improvement of production efficiency (Bao et al.,
2019). Another paper used a DT in an aircraft parts production workshop with less
computational costs. It proposed multi-fidelity simulation-based optimization method is
well-applied in solving large-scale problems and outperforms other simulation-based
optimization methods (Zhang et al., 2022)
DT also applied in other manufacturing industries as well such as musical
instrument factory and pharmaceutical industry. Lu et al. (2021) makes a contribution to
the existing body of research on DT modeling by putting up a theoretical framework that
combines Value stream mapping (VSM) with DT. The simulation-based technique was
applied to the VSM analysis, and this work presents a formal methodology for doing so. In
the Pharmaceutical industry, a framework to improve processing times of sample and
utilization rates of equipment and staff who are engaged in analytical tasks was developed
by Lopes et al. (2020). Another paper demonstrated that the DT acts as a channel for
17
attaining cyber-physical amalgamation via bidirectional interface, data analytics, and
connection of information silos across the whole product cycle. This is accomplished
throughout the entirety of the product development process (Onaji et al., 2022).
18
Table 2. 1 Summary of recent research with contribution and focused area
Authors
Research
Type
Research Contribution
Focused
Industry
Technology
Used
Boschert
and Rosen
(2016)
Review an
existing a
product
Analyzed life cycle
assessment of a
mechatronics product.
General
Manufacturi
ng
Simulation
Tao (2017)
Concept
design
Conceptual model and
specific operation
mechanism of DTS.
CNC
machine
3D modeling
software
Tao et al.
(2018a)
Development
of a
Framework
A framework of DTPD,
and described the concept
to redesign a bi-cycle
Redesign of
a Bi-cycle
3D modeling
software
Qi & Tao
(2018)
Review
Analyzed similarities,
contrasts and integration of
Big data and DT.
General
Manufacturi
ng
Graphics
design
software
Tao et al.
(2018b)
Concept
design
Use of convergent cyber-
physical data to manage
product lifecycle in DT
General
Manufacturi
ng
Graphics
design
software
Wang and
Wang
(2019)
Concept
design
DT based system to tackle
the connectivity and
integration issue along the
WEEE
recycling/remanufacturing
chain.
Electrical
and
electronic
waste.
3D model in
Web
Zheng et al.
(2018)
Framework
and case
study
An application framework
of DT for product lifecycle
management is proposed in
a welding production line.
Welding
production
line
3D model &
OPC
protocol
Liu et al.
(2018)
Framework
and case
study
The performance of the
system is greatly
improved.
Automated
flow-shop
manufacturi
ng system.
3D modeling
Botkina et
al. (2018)
Concept
design and
case study
DT concept is developed
for monitoring the
condition and supporting
the operations.
Cutting tool
making
CAD, CAM
Luo et al.
(2019)
Concept
design
Demonstrated the DT
concept in the era of CNC
machine tools for fault
prediction and diagnosis.
CNC
machine
tool
3D modeling
O’Sullivan
et al. (2020)
Framework
A DT framework for
improving maintenance
strategy.
CNC
machines
Sensor
19
Liu et al.
(2019)
Methodology
Suggested methodology to
real-time process planning
evaluation and
optimization
Marine
diesel
engine
RFID, OPC
interface
Park et al.
(2019)
Concept
design and
case study
DT concept is developed
and adopted for integrated
monitoring, tracking, and
decision-making support
functions.
Robot in a
Micro smart
factory.
3D, Sensors,
Simulation
Park et al.
(2020)
Framework
and case
study
Supply chain planning was
carried out according to
customer orders and
responded resiliently to
actual fluctuation.
Automobile
parts
industry
Simulation
Park et al.
(2022)
Framework
and case
study
A production control
model with a DT
application was designed
for Re-entrant Job shop
production control.
Lithium-ion
battery
aging work
center
Visual
studio,
Python
Redelinghu
ys et al.
(2020)
Framework
and case
study
Developed an architecture
to provide a real-time,
service-based
infrastructure for both
vertical and horizontal
integration.
Robotic
gripper in
assembly
lines.
Sensor,
Simulation
He and Bai
(2021).
Concept
design
Framework is developed
for intelligent
manufacturing
General
Manufacturi
ng
Visual studio
Braun et al.
(2020)
Concept
design
Presented a concept for
automated layout
generation
Robot in
manufacturi
ng industry.
Visual studio
Alexopoulo
s et al.
(2020)
Framework
Framework for vision
based recognition of parts’
orientation using
simulation of DT models
Industrial
warehouse.
Visual studio
Vatankhah
Barenji et
al. (2021)
Framework
and case
study
DT was built for real-time
optimization of motion
planning of robotic cell to
save energy consumption.
An
industrial
robot used
for
assembly
purposes.
3D,
Simulation
Lattanzi et
al. (2021)
Review
Examined the practical
implementation status of
DT.
Production
environmen
ts
visual studio,
3D
20
Nikolakis et
al. (2019)
Framework
and case
study
Virtual representation of
warehouse task and
evaluate cycle time
Industrial
robot in
warehouse
3D, Sensors,
Simulation
Ding et al.
(2019)
Framework
Framework developed for
real-time data-driven
operations control of DT-
CPPS.
Manufacturi
ng shop
floor.
visual studio,
3D
Latif and
Starly,
2020)
Methodology
and case
study
DT integrates, virtualizes,
and replicates physical
objects to improve
operation sequence.
Missile
manufacturi
ng industry
3D, Python
Zheng et al.
(2021)
Framework
and case
study
Framework for DT based
product smart
manufacturing
Aero engine
blade
manufacturi
ng
RAMI4.0
Polini and
Corrado
(2020)
Methodology
and case
study
The DT concept was
developed to manage
geometrical deviations in
composite components.
Composite
3D, MSC
Marc®
solver and
Matlab
Bao et al.
(2019)
Methodology
DT modeling of product,
process and operation.
Aerospace
structural
parts.
Automation
ML
Zhang et al.
(2022)
Framework
and case
study
Multi-fidelity simulation-
based optimization method
applied to a DT-based
aircraft parts production
workshop with less
computational costs.
Aircraft
parts
production
industry
3D,
Simulation
Lu et al.
(2021)
Framework
and case
study
Production process
reengineered by integrating
VSM & DT.
A musical
instrument
factory
Simulation
Lopes et al.
(2020)
Framework
and case
study
Framework to improve
sample processing times
and utilization rates of
resources.
Pharmaceuti
cal industry
SIMIO,
Simulation
Onaji et al.
(2022)
Framework
and case
study
Integrate product and
process DT in a cyber-
physical production system
(CPPS).
Pharmaceuti
cal Industry
CAD,
Simulation
21
In summary, the DT generates tangible significance, develops flow of revenue, and
aids these in making crucial tactical resolutions. With the assistance of modern
technologies, manufacturing companies can begin developing a DT of the facility.
Throughout its entire life cycle, the DT is useful in a variety of different contexts. It is now
much simpler to provide real-time solutions that are both more accurate and exact to vital
"what-if" inquiries. To stay pace with the evolving manufacturing industry, it is highly
suggested that clothing manufacturing plants have a DT. However, the extant literature
lacks business-oriented examples of garment manufacturing industries.
2.4 Digitalization in Apparel Manufacturing
Although digitalization has been implemented to some extent in other industries,
there has been limited development in the garment and textile sector. Despite having a
global market size equivalent to that of the global car industry, digital transformation in
the garment industry is exceptionally challenging due to its labor-intensive industrial
structure. Also, limited research has been carried out to implement numerous digital
concepts and technologies in the apparel industry, some of those recent research works are
described below.
The IoT or cloud-based IoT with block chain has been proposed for deploying in
information system applications of apparel supply chain and logistics management to
improve the effectiveness and efficiency of business processes while maintaining systems
security and privacy (Pal & Yasar, 2020). 3D scanning, CAD, Big data and information
systems, RFID technology are key factors for the success of mass customization of apparel
production (Liu et al., 2020).
22
A research of Wijewardhana et al. (2021) focused on benefits of the various
technologies of industry 4.0 (Cognitive technologies, Data analytics, Augmented reality,
Virtual reality, 3D prototyping software, Cloud computing, IoT) at different stages of new
apparel product development. Industry 4.0 technologies may be viewed as a good
instrument for identifying the fundamental causes of supply chain issues and proposing
alternative solutions to eradicate those problems (Braglia et al., 2021). They also intend to
highlight the implications of Industry 4.0 on business units, processes, and components in
the Textile and clothing industry. Another study presents a detailed review of the probable
effects of the Fourth Industrial Revolution on this industry using new facts and descriptive
instances. The paper explains how digital transformation may make the apparel sector more
sustainable and customer-driven. But it also highlights conventional businesses' delayed
embrace. Focusing on ongoing, undiscovered occurrences, it provides probable paths,
enabling an effective response (Bertola & Teunissen, 2018). Based on a bibliographic
coupling analysis, this paper identifies six technological trends in the clothing industry.
RFID in the fashion supply chain, virtual technologies for online clothes purchasing,
industry 4.0 in garment production and supply chain, CAD in design and manufacturing,
as well as IT and virtual technologies for garment design (Hoque et al., 2021).
The literature study conducted for this paper revealed that the textile sector in the
worldwide environment is in the early stages of implementing Industry 4.0. Textile and
apparel industries have primarily invested in technological projects that integrate
information technologies, IoT technologies such as sensors, RFID, actuators and QR code;
and software for the planning, programming, control, and management of the production
process in order to increase efficiency, reduce operating costs; and enhance business
23
management (Dal Forno et al., 2021). Another study suggested integrating RFID
technology as a tool of Industry 4.0 with Enterprise Resource Planning (ERP) software for
apparel supply chain to provide better service to the customers (Majeed & Rupasinghe,
2017). A literature review divided the AI applications of the garment sector into four
categories: apparel design, apparel production, apparel retailing, and apparel supply chain
management. The scientific progress achieved in each category was thoroughly evaluated.
The AI approaches utilized to address a variety of challenges were thoroughly evaluated
and assessed (Guo et al., 2011).
The findings of this study suggest that research of Industry 4.0 in the textile and
apparel sector is in its infancy. Researchers have not dug deeply enough into how various
Industry 4.0 efforts may be used in the textile and clothing sectors. The textile and garment
industry has done the most research on the IoT and RFID technology's potential uses.
Research is also being conducted on the potential uses of technologies like cognitive
computers and autonomous robots in the textile and garment sectors (Deepthi & Bansal,
2022). The idea of a smart clothing factory, which is referred to as "Apparel 4.0," is
presented, and both the advantages and disadvantages of "Apparel 4.0" are investigated.
As part of the Fourth Industrial Revolution, "Apparel 4.0" makes use of cutting-edge new
technology. The effects of these new technologies on a clothes and apparel firm's
production system and management procedures are also examined (Gökalp et al., 2019).
Manglani et al. (2019) have shown that the textile industry is far from being digitized. Due
to diverse communication protocols and products, heterogeneity and interoperability are
the greatest technical obstacles they confront while implementing IoT. Noor et al. (2022)
noted that there is a huge amount of potential for the application of AI in the production of
24
clothing items since it has the ability to drastically cut down on the costs of product creation
and result in considerable cost savings.
Below table summarizes the studies done by the researchers in the apparel industry.
25
Table 2. 2 Research contribution in apparel manufacturing digitalization.
Authors
Research
Type
Research Contribution
Guo et al.,
(2011)
Review
This study depicts the digitalization adopted by
the apparel industry.
Majeed and
Rupasinghe,
(2017)
Review
Highlight the necessities of RFID for the apparel
industry.
Bertola and
Teunissen,
(2018)
Review and
methodology
Described the methodology and challenges of
adopting Industry 4.0 in the clothing industry.
Gökalp et al.
(2019)
Review and
methodology
A concept of "Apparel 4.0" was presented
Manglani et al.
(2019)
Review
Identified the barriers of implementing IoT
Pal and Yasar
(2020)
Review
Analyzed the potential of applying IoT in apparel
supply chain and logistics management.
Liu et al. (2020)
Review
Identified as key factors for the success of apparel
mass customization.
Wijewardhana et
al. (2021)
Review
The new product development process can be
improved using Industry 4.0 technologies.
Braglia et al.
(2021)
Review
Prospects of Industry 4.0 to solve supply chain
problems.
Hoque et al.
(2021)
Review
Identified the technological trend in the fashion
industry.
Dal Forno et al.
(2021)
Review
Identified the status of implementations of
Industry 4.0 and so far implemented technologies
in the apparel industry.
Deepthi and
Bansal, (2022)
Review
This study portrayed the digitalization adopted by
the apparel industry and found the potential of
using IoT and RFID technology.
Noor et al.
(2022)
Review
Identified the potentials of applying AI in apparel
manufacturing.
26
A study explored the possibility of utilizing DT for demand forecasting in the fast
fashion business (Henrique dos Santos et al. n.d.). The 3D simulation technology was used
to create a number of virtual human models of varying sizes and sex-specific traits. The
simulation experiment backs up the sanity of the design approach and classifications by
showing that they provide the best match and most logical pressure distribution (Cheng &
Kuzmichev, 2018). Further study is needed to support any potential social interactions
between humans and DT, although a pilot study evaluated the potential appeal of human
DT as consumers of a physical store (Stacchio et al., 2022).
The DT can help apparel manufacturers make important strategic decisions, establish
new income sources, and produce tangible value. With the use of modern technology, the
apparel manufacturing industry may construct a DT which may be used in a wide variety
of ways from initial product ideation through detailed design and finally finished product
production. A more precise solution to the many problems of the plant can be addressed
with this. However, practical examples of DT in the garment industry are lacking in the
current literature. The next chapter will develop a concept for DT for apparel
manufacturing plants.
Extensive literature study about digitalization in the apparel industry revealed that
some studies reviewed the scopes, barriers and necessities of various IoT, AI, RFID, and
very few developed the methodology. It also identified that DT was studied in demand
forecasting and creation of human virtual models. None of the study focused on DT
implementation in apparel manufacturing process. This is why the author carried out this
research to develop methodology of DT in apparel manufacturing.
27
CHAPTER 3: PROPOSED METHODOLOGY
3.1 Introduction
It is evident from the study of the relevant literature that DT has the ability to
provide a solution to the difficulties that currently exist in clothing production. This study
will adhere to International Standardization Organization (ISO)-23247, which establishes
generic standards for the DT framework for manufacturing. The standard contains four
components:
(1) Overview and general principles: It defines terminologies used by the standard.
Physical systems must be modeled depending on the use case's scope and context.
Synchronizing a DT with its OME keeps it current with its OME and optimizing
OMEs.
(2) Reference architecture: It includes a reference model that has four domains and
five entities. Four domains in the reference architecture are Observable
manufacturing domains, Data interconnection domain, Core and User domain. Five
entities are OMEs, Data interconnection Entity, Core Entity, User Entity, and
Cross-System Entity.
(3) Digital representation: Digital representation of OMEs includes both static and
dynamic information.
(4) Information exchange: This outlines the technical specifications for the flow of
information across the many entities that make up the framework.
Following the ISO-23247 standards, the DT will be constructed using the below processes
or phases;
28
1) Collection of required data from the machines and operations or any other
manufacturing resources.
2) Appropriately convert data into information that can be used across the many
control systems of the manufacturing plant.
3) Efficiently incorporate a wide range of data related to the equipment and
machines, process, operators, and environment gathered from a variety of sources
and processed in following a variety of ways.
4) Develop a mechanism to react quickly to real-time situations or facts to make a
better decision.
3.2 DT Concepts for Apparel Manufacturing Plants
According to the available research, a DT for a manufacturing facility must have
these elements; OME in the physical world, digital elements in the virtual world, Data,
integrating both the worlds together, and service.
OME: The manufacturing world is a complicated, varied, and ever-changing
production surroundings, which is made up of equipment and machines, people,
materials, procedures, and environment. Here the manufacturing element refers to
all types of materials, various types of machines, production lines, manufacturing
cells, and any other assets. Materials can be raw materials, work in process, and
finished goods. Other resources such as production line equipment, computing,
storage and software are also considered physical elements (Zheng et al. 2019).
With regard to apparel manufacturing, there is a wide range of physical
elements. These are various types of sewing machines, cutting machines, finishing
29
machines, other manual workstations, various types of cut panels, different types
of raw materials, work in progress of sewing garments, and finished garments.
Digital Element: The digital components of a manufacturing plant can be the
virtual depiction of the physical elements. In addition, various production
procedures, material movement, and machine status can be part of the digital
element. These elements should have the capacity for integration, which is
considered as a digital depiction of the physical components of a factory.
Everything that is a part of the DT, from the process map of the manufacturing
facility to the digital portrayal of the stations, contributes to the formation of the
whole image by working in concert with one another. It may be a simulation where
physical processes are mimicked to represent the functions with restrictions. This
allows for the quantitative and visual analysis of real-time activities, process
parameters, raw materials, Work-in-process (WIP), Finish goods status, production
line layout, production machines and equipment, and other assets of the apparel
manufacturing facilities. A DT could include a number of different computational
or analytical models, ranging from knowledge-based or established science, data-
driven statistical, geometrical or visualized (3D or 2D) modeling, machine learning
and artificial intelligence.
Data: The integration and connectivity of virtual components and physical
elements cannot be accomplished without the use of data or information as the main
component. Perception data in physical space, simulation data in virtual space, and
fusion data between the two may all be considered data (Zheng et al. 2019).
30
Accurate data is required to represent the physical elements of apparel
manufacturing. A DT includes data about its physical elements. Data comprises
simulation or modeling parameters including cutting, sewing and finishing
production information, specifications of product design, processes, engineering
data, manufacturing data such as manufacture equipment and various machines,
material, method, quality data, operations cycle time, machine breakdown time, and
duration to repair, historical and operational state data of the real-world counterpart.
Integration: Connecting the physical and digital elements is the most important
task of DT formation. It is very necessary for the functionality of a DT. These
should be connected in such a way that the digital elements and their physical
counterparts should be receptive in both directions in real time and dynamically.
The connection that takes place between the digital and physical aspects enables
the early detection of any possible issues that may arise. This is the advantage that
DT offers, as well as their primary purpose. A feedback loop that is based on real-
time data and has the potential to change may enormously enhance the DT. It is
even better if the feedback loop is automated. Appropriate integration enables the
DT with the feedback loop. The DT of a manufacturing facility needs some kind of
database or repository in order to function properly, which can be a Structured
Query Language (SQL) or a Non-SQL. If there is access to the database, one can
use any programming language that is appropriate to automatically get the
processed data from the repository. Data collection can be manual if suitable
technologies are not available.
31
Data collection, processing and managing is a critical factor for the success of
digital twin. As a result of the quick growth of detecting, communications, and analytical
technologies, generation, collection, and managing of information can be done
automatically, unlike the old method of processing of information. Technologies would be
used in this stage should have the features of accurate perception, observation, and response
to the physical world in real-time basis for proper functioning of DT. This whole process
can be carried out in four layers that are 1. Sensing layer 2. Network layer 3. Middleware
layer and 4. Application or service layer.
A linkage of the physical world to the digital world is provided by the sensing layer.
It is equipped with a variety of devices for the gathering of information, for instance,
sensors, an RFID reader, a GPS system, an observing camera, and so on. It is through this
layer that DT is able to accomplish real-time monitoring and management of the objects'
characteristics or behavior patterns. The below section will illustrate various types of RFID
technology and sensors that might be utilized in this stage.
Sensors: A sensor converts the physical activity to be measured into an electrical
signal and transforms it so that it may be easily conveyed and further studied. Sensors are
crucial to the process of automating any given application because of their ability to
measure and analyze data in order to identify changes in the state of physical things. This
will create a response that can be measured if there is a shift in any one of the physical
conditions for which it was designed. Below section will illustrate the various types of
available sensors and their functions.
A. Proximity Sensors: A proximity sensor can readily determine the position of any
adjacent object without requiring any direct physical touch with the object. There are
32
several distinct varieties of proximity sensors, including magnetic, capacitive, inductive,
photoelectric, ultrasonic, and so on, all of which are designed specifically for use in a
variety of contexts. These sensors are used for object identification, counting things,
measuring rotation, positioning objects, material identification, movement position,
parking sensors, and more.
B. Position Sensors: The existence of humans or objects in a defined location can
be detected by position sensors by identifying their movement. This type of sensor might
be used for monitoring the position of any objects.
C. Occupancy Sensors: Occupancy sensors, often known as presence sensors,
detect people and things. It monitors temperature, humidity, light, and air remotely.
D. Motion Sensors: This sensor detect kinetic and physical activity and send
photographs and movies to the server. This software takes continuous photos and videos
from motion start to finish.
E. Velocity sensors: It can be linear or angular, which measures continual position
change with values at specified breaks. Angular velocity sensors measure device rotation,
while linear velocity sensors measure straight-line speed.
F. Temperature sensors: This kind of sensors are useful in identifying the physical
changes that have occurred by measuring the heat energy, and this information can then be
utilized to monitor the ecological situation of the neighborhood. After collecting the data,
it needs to be uploaded to the cloud or computer for analysis.
G. Pressure Sensors: Sensors that detect force and transform it into a signal are
referred to as pressure sensors, which can be used in assessing the machine whose
parameters are related or controlled by pressure.
33
H. Chemical Sensors: This type of sensor can detect the chemical make-up of an
environment or any other chemical process. Chemical sensors are commonly used in
environmental monitoring.
I. Humidity Sensors: Temperature and moisture levels in the air are both measured
by a humidity sensor, which then reports the relative humidity of the surrounding air. This
can be used in the moisture and temperature control of fabric in apparel industry.
J. Water Quality Sensors: Water quality is measured by this sensors by monitoring
the ion of water.
K. Infrared Sensors: In order to perceive various features of specific objects,
infrared sensors may emit or detect infrared radiations. They can monitor heat output as
well. These kinds of sensors may be put to use to keep an eye on and operate machinery
and other types of equipment, such as to switch lights on and off automatically.
Additionally, it may be utilized for intelligent security, garbage collection systems, and
intelligent parking.
L. Gyroscope Sensors: By detecting an item's angular velocity, gyroscope sensors
are able to sense any angular movement or tilt of any object.
M. Optical Sensors: This type of sensors are helpful in any applications that need
to detect electromagnetic energy like light.
N. Chemical Sensors: Chemical sensors are sensors that may respond by detecting
group of chemicals, any chemical reaction or chemical substance (Sehrawat and Gill 2019).
RFID: RFID technology can also be used to identify the type of material, machine,
and or any other object with accurate quantity using an electromagnetic system. Scanners,
34
transceivers, and transponders make up any RFID system. RFID readers or interrogators
use the scanning antenna and transceiver. Tags react to RFID transceivers with identifying
information that may be linked to arbitrary data records. Network-connected RFID readers
are portable or fixed. Radio waves activate the tag. The antenna converts the tag's wave
into data after activation. RFID tags have an integrated circuit, antenna, and substrate.
Active RFID tags need batteries to power themselves, whereas passive RFID tags use the
reading antenna's electromagnetic wave to power their antennas. Semi-passive RFID tags
use batteries for electronics and RFID readers for communication. RFID tags store less
than 2,000 KB data along with a serial number. RFID tag detecting range depends on RFID
frequency, tag type, reader type, environment or interface between RFID tag and reader.
Due of their greater power supply, active RFID tags can read farther than passive ones.
RFID systems may be broken down into three broad categories depending on their
operating frequency: high frequency, low frequency, and ultra-high frequency. RFID in the
microwave spectrum is also a viable option. For low-frequency systems, transmission
distance varies typically from a few inches to less than six feet. Typically, the effective
range of a high-frequency RFID technology is several feet. UHF RFID systems typically
have a reading range of more than 25 feet. The reading range for microwave RFID systems
is 30 feet or more (Weis n.d.).
The data that has been acquired at the sensing layer is sent to the top levels of the
system to be processed further and abstracted at the next level, which is the Middleware
layer, by means of the network layer. Various technologies are being used in this stage
35
such as IPv4/IPv6-based Internet, 3G/4G/5G wireless networks, and private networks
(Derakhshan, Orlowska, and Li 2007).
Middleware is a sort of software that bridge the connection between the network
layer and the application layer. The tasks of information collection, filtering, analysis, and
decision-making are only some of the middleware functions that may be performed by
sensor networks. The Application Layer is built on top of the preceding layers to provide
end-users with IoT apps tailored to certain domains (Ma, Wang, and Chu 2013). Data
collection can be manual if suitable technologies are not available.
36
Figure 3. 1 Elements of DT and their integration
37
Physical elements of apparel manufacturing systems are complex, diverse, and
dynamic. And they contain a lot of data that needs to be collected and transferred to the DT
for building an algorithm or a mathematical prototype or a simulation model. The DT needs
to be trained using a model with a historical dataset so that it can provide advice using
optimization, machine learning, fuzzy algorithms, and deep learning. User feedback
contains practical decision suggestions. This decision aid should be simple and alter
manufacturing plant operational parameters. The manufacturing process is repeated until
either the ideal outcome or the predetermined output level is achieved. Then, software
applications can use DT service interfaces to retrieve data and models. DT and their real-
world counterparts have a dynamic, perhaps real-time, bidirectional link. Thus, a DT
provides decision-makers or users with the ability to duplicate and predict the state and
behavior of a real-world counterpart by employing processed data from the past as well as
data collected in real time. It adapts to real-world changes. Figure 3.2 illustrates the
visualization of the concept.
38
Figure 3.2 Illustration of Apparel Manufacturing DT
39
3.3 Methodology of Developing DT for Apparel Manufacturing
DT continues to influence the industrial sector, however, relatively little study has
been conducted to integrate it into the garment industry. A suitable guideline or approach
might be beneficial to the endeavor. DT has been described in numerous ways by
researchers. While some study has concentrated on the concept of the DT as a virtual
duplicate of a physical system, other research has emphasized real-time visualization as a
crucial breakthrough that may be achieved with DT. Some of the studies concentrated on
either the product, the method, or the system. The fidelity of the digital representation can
be expressed as either a whole or partial representation, and the integration can either take
place in real time or offline. A study explained that DT is fit-for-purpose digital
representations based on context and viewpoint for a given use case. A physical system
should have many DT based on its application (Shao & Helu, 2020).
This chapter will describe a generic methodology of DT that can be used in various
cases of apparel manufacturing. In addition, this can be used in for other manufacturing
industries with some customization. The figure 3.3 shows the pictorial views of the
methodology.
40
Figure 3. 3 Development method of DT for apparel manufacturing
41
The need for the creation of a DT via design and development comes from the real
problem in the apparel manufacturing industries. Defining the problem is the first step of a
DT development. The problem statement should contain the identification of the primary
components, as well as the objectives and restrictions. It might involve boosting the
efficiency and productivity of a manufacturing plant or a section of manufacturing,
lowering the throughput time or lead time, or minimizing production downtime or style
changeover time. Constraints and key parameters are required to identify to formulate the
DT goal. Following the goal, a manufacturing element, which is called as OME according
to ISO-23247, is required to select. Acquiring knowledge, both static and dynamic, related
to OME comes next. While dynamic information evolves during the course of the
procedure, static information stays the same. Depending on the OME, static information
will vary. Static information includes no and types of machines to produce a garment,
operation bulletin and standard minute value, cycle time of manufacturing operations,
preventive maintenance schedule, no of cut-panels to make a garment, various finishing
operations, inventory capacity of the warehouse, various work stations and so on. Dynamic
information includes the current status of a machine whether a machine is working or out
of order, at what speed a machine is running, how much outputs are produced at various
workstations or current WIP at various workstations, how much time is taking for material
to flow along the processes or any other information related to OME, which can represent
the current picture of it. After compiling all of the relevant information, a conceptual model
is created. The conceptual model provides an almost exact representation of the actual
42
procedures that are carried out at the production facility. It will allow the sharing of
information on the aspects of the problem in the virtual world.
The specification of the data required occurs concurrently with the conceptual
model design. Which data get into the DT is determined in this stage. Development of Data
collection and device control domain and determination of entity for Data collection and
device control comes next. OME needs to be connected to their DT for synchronization by
collecting data from OMEs using sensors, RFID technologies, manual operations, routers,
and so on. These entities measure critical inputs for the DT. Along with collecting data
from OMEs, they are required to control as well. Therefore, these have to be equipped with
the capability of adjusting their function, behavior, and structure. This task is carried out
by hydraulic, pneumatic, electric, and mechanical actuators. As apparel manufacturing
involves many manual operations, the objective of controlling the OMEs can be carried
out in any other digitized way. Using integration protocols and a data repository, functions
should exchange data. Pub/Sub, OPC-UA, TCP/IP, and ERP transmit data efficiently
(Beňo et al., 2019). The connections may be enabled via the use of network connectivity,
cloud computing, and network security. Wi-Fi, Bluetooth, barcode, QR code, Z-Wave, etc.
are potential consumer networking technologies. Local hardware or the cloud can
aggregate and process data. Now, data is processed. Different data sources must be
compiled into a database. Databases hold historical data. The database also handles real-
time manufacturing plant data. Available data, data requirements, and strategy cohere to
arrange database data. Concerns exist about data quality. Product data quality is key to a
successful DT installation. Product specifications, manufacturing plant characteristics,
machining sequences, etc. are all maintained in a database.
43
After data collection, processing, and conceptual design, the DT domain and entity,
which is the core domain and entity of DT, is built. This domain and entity are responsible
for the overall operation and management of a DT. In order to facilitate provisioning,
monitoring, modeling, and synchronization, it serves as a host for applications and services
such as data analytics, simulation, and optimization. It also includes many functional
entities for operating, managing, applying, servicing, and interchanging for digitally
representing and maintaining OMEs. Modeling the input and output elements of the OMEs,
such as the flow of material, time, yield rate, product output, etc., will result in the creation
of a digital model of the OMEs. The data should come from the OMEs. The established
model requires all entities to be linked in such a way that data may flow freely and instantly
between platforms, and data is viewed, tracked, and evaluated. Use sophisticated analytic
tools and infrastructure to get new perspectives and make informed decisions. It also
interacts with the users of the DT and other DT.
Following the completion of the DT's construction, the user will execute the DT
model. The output of the DT offers the user feedback, allowing the user to make an
educated choice based on the information provided. The conclusion of the suggestion phase
is displayed and brought to the forefront of the discussion. Evaluating the performance and
trying out a variety of various situations are also options. The manufacturing facility will
benefit from the DT's ability to support and capitalize on both short-term and long-term
improvements. OME receives controlling instructions or informed decisions from the user
entity or the DT entity, which are communicated via the device control entity so that they
can follow the improved solutions. Through the use of hypertext transfer protocol and
representational state transfer, the user entity is able to establish a connection with the DT
44
through the internet. Web GL (Graphics Library) and Open GL are two technologies that
enable the sharing of graphical information (Farhadi et al. 2022). Data can be shared via a
cloud or database hosted on the internet. Based on this developed methodology,
stakeholders may tweak the system's infrastructure to better serve their needs.
In the next chapter, a case study is presented to materialize the suggested
methodology. This case study will show how real problems related to apparel
manufacturing can be solved using this methodology.
45
CHAPTER 4: DIGITAL TWIN IMPLEMENTATION IN AN
APPAREL MANUFACTURING
4.1 Introduction
Apparel manufacturing is one of the old and labor-intensive industries. Sewing is
the most significant part of the garment production process. As this component accounts
for the majority of the cost of producing a garment, its efficiency determines the whole
plant's productivity. A concept of DT of a sewing assembly line is developed in this case
study to overcome the existing challenges it has. Numerous reasons affect efficiency
negatively. The performance of sewing machines, the most fundamental equipment used
to manufacture clothing, varies significantly based on the working qualities of their
operators. This results in bottlenecks in the garment production line and ultimately
production delays, consequently diminishing productivity. Proper workload balancing
among the workers of the sewing line is important to maximize the production efficiency.
The technique of assigning tasks to individual operators in such a way that overall line
efficiency is maximized and total number of workers are reduced is known as a line
balancing method. Many researchers have worked to determine the best method to balance
the workload of the sewing line.
4.2 Current Methods and Practice of Assembly Line Balancing
Guo et al. (2011) reviewed various research works of line balancing in apparel
manufacturing. A U-shaped sewing line is proposed to reduce the distance cut parts must
travel. A mathematical model is utilized to compare the PBS with the unit production
46
system (UPS), and flexible assignment of operations in sewing lines is illustrated. The
assembly line is separated into many sewing sections, and a fuzzy operator allocation
system has been devised, however it does not account for the impacts of sewing operators'
productivity on production performance. In the instance of UPS, line balance is optimized
using a genetic algorithm-based methodology. In the case of PBS, another GA-based
algorithm was implemented to fulfill the specified cycle time and decrease the overall idle
time of sewing lines. The model calculated the appropriate assignment of operations to
workstations as well as the fraction of each shared operation's tasks that were performed
on distinct workstations. Another study used RFID based technology to capture data to
control production. Also, a time constant learning curve was adopted to deal with the
change of operative efficiency in sewing production.
A study of developments and trends in assembly line balancing methodology
mentioned that equipment costs, cycle time, the association between job durations and
equipment costs, and the flexibility ratio require significant consideration (Kumar &
Mahto, 2013). Manaye (2019) analyzed numerous methods of sewing line balancing that
include Ranked Positional Weighted Method, Simulation methods, Largest Candidate Rule
(LCR) Method, Probabilistic technique, Kilbridge and Wester Column (KWC) Method,
Hoffman method. He found that Ranked Positional Weight gives better results in the line
efficiency. Another study proposed a system that uses a highly visible red-light warning
system connected to the workstation with the intention of managing the speed of production
by maintaining the appropriate WIP level. Using some of the aspects from industry 4.0,
this research looked into how a garment assembly production system might be transformed
into a smart production line, complete with the ability to track cycle times, machine
47
utilization rates, and employee engagement at individual workstations (Udayangani et al.,
2019). In one article, different sorts of layouts were studied based on the characteristics of
workstations, tasks, limits, and how they are operated. In solving assembly line balancing
problems that arise in real-world scenarios due to manual operation, learning effect, worker
fatigue, lack of motivation, machine or equipment breakdown or poor maintenance, or raw
material defects, he concluded that cutting-edge solution techniques should be utilized
(Chutima, 2020). Another study applied robust optimization and a minmax regret objective
to minimize the largest difference in cycle time between the optimum balancing solution
offered and the ideal balancing solution proposed by the authors (Jin et al., 2022).
The factory considered in this case study is one of the recognized woven shirt
makers in Bangladesh. Most of the factories use skill matrix data to carry out the initial
line balancing either using Microsoft Excel or any other software. The Industrial
Engineering department is responsible for accomplishing this task before a style or order
is fed into the sewing assembly line. The operation bulletin along with the line balancing
instruction is conveyed to the respective production lines, and line supervisors distribute
the workload among the operators. Supervisors and managers use their knowledge and
experience to manage the unexpected situation arising from various reasons. However, it
is observed that supervisors are mostly engaged in solving the problem rather than re-
balance the production line based on the current situation.
From the literature review and industry best practice, it is evident that almost all
the research carried out to remove bottleneck operations are based on the available data;
real time data is not being utilized to solve the problem. Although these methods can assist
the decision maker to some extent, however, today’s dynamic environment may not be
48
suitable to achieve the objective. Recent fashion trends push the manufacturers to produce
a variety of products where each order consists of less quantities compared to before. This
makes the production line occur with frequent changeovers and running products with
shorter life cycles. Low quantity per style causes the manufacturing process' efficiency to
decline because of many style changeovers and the short life cycle of a style in a production
line.
In addition, many uncertainties about the plant area cause it not to achieve the
desired efficiency. There are several causes for this uncertainty. Uncertainty on the
production floor may result from the unavailability of a component, a malfunctioning
machine, taking time to repair or replace the machine or equipment, producing defective
product generation, inconsistency in workers’ performance, change of operators in the
production line, and so on. This circumstance creates a challenging situation for factory or
production managers to balance workload or optimize manufacturing processes or
operations in the sewing line. The author has many years of experience working in the
apparel manufacturing industry. Also, the author interviewed several production engineers
from recognized apparel manufacturers in many countries and found out that majority of
the inefficiencies related to manufacturing plant can be solved with quick and real-time
time decision-making abilities connected to the production plans or manufacturing
problems.
Currently available algorithms, calculations, and way out for anticipating human
behavior in smart manufacturing are not enough for dynamic activities. By using DT,
production managers may enhance their visibility of operations, forecasting, managing
demand, scheduling, planning, controlling production, managing inventory, and
49
purchasing. A DT would aid in managing apparel manufacturing uncertainties to lessen
their detrimental effect on efficiency. The DT is a mix of technologies that enables efficient
production and satisfies individualized production requirements. The DT requires a
manufacturing facility having intelligent production systems, manufacturing resources,
manufactured items, raw materials, and human operators to manage the production
disruptions causing downtime, order planning, production planning, task scheduling,
quality control, and on-time delivery in an automated way. These automated production
system functions aim to reduce lead time, reduce cost by increasing productivity and
production efficiency, and improve product quality.
4.3 Objective of the Case Study
In this case study, the presented methodology of DT is applied, analyzed, and
evaluated in an assembly line of the sewing section of apparel manufacturing. The case
study consists of a complex woven shirt product with many operations in the sewing
section. Plant simulation tools will be enhanced by DT thanks to the incorporation of real-
time data. When sensor or RFID data is included into DT simulations, realism and accuracy
are increased. Therefore, the DT can quickly, cheaply, and effectively rectify the
inefficiency in the manufacturing line caused by the unpredictability of shop floor
activities. A working software concept following the methodology has been created and
put into action to guarantee the validity of the proposed approach with respect to DT
implementation procedures. The main contributions of this case study are:
(1) Utilizing the proposed DT technology to simulate a complicated woven shirt
product in a clothing manufacturing facility.
50
(2) To provide a solution to the problem of workload balancing and quick response to
the problem by using essential data and parameters to generate realistic simulations.
(3) Providing step-by-step implementation guidelines for the proposed methodology.
4.4 Case Study
In this study, we focus on an assembly line of a woven shirt product that requires
various parts and other materials to be joined together by human-operated machines. The
selected apparel manufacturing facility is located in Gazipur district of Bangladesh. The
company produces around 600,000 pcs of woven shirts each month to export to the USA,
Canada, European countries and some Asian countries. The manufacturer works for some
of the reputed apparel brands of the world that includes Calvin Klein, Tommy Hilfiger,
Michael Kors and so on. This company has around 25 sewing assembly lines having 3000
workers (approximately) who are working 8 to 10 hours each day depending on the
demand. Figure 4.1 shows a sewing assembly line of the selected factory.
Acquiring and combining real-world heterogeneous data into digital models, so
constructing digital representations of real-world events with high accuracy, is a major
difficulty in allowing realistic simulations. The purpose of this project is to design an
autonomous system capable of a rapid response to a machine failure and the real-time
balancing of an assembly production line's workload in the face of uncertainty.
51
Figure 4.1 A sewing assembly line (Source: A apparel manufacturing company in
Bangladesh)
52
A linear, discrete-event, time-invariant system is used to represent the physical
world. The production managers and supervisors are the ones who use the DT. All
information is gathered from the manufacturing plant. Here, OME is a sewing assembly
line, from where data has been transmitted to the DT to carry out the simulation. The very
first simulation was carried out based on the historical data from the recorded database.
Then, real-time data has been collected using sensors and RFID; and fed into for
simulation, every after a certain duration, which can be considered real-time. DT then
provides the solutions by assessing the current scenario, and users took initiative based on
the recommendations. An overview of the DT concept for this case study has been
illustrated in figure 4.2.
53
Figure 4. 2 An overview of the DT for the case study
54
4.4.1 Observable Manufacturing Elements (OME)
In this case study, a sewing assembly line and a style from a specific product family
having 61 sewing operations have been selected. These operations were carried out by
human operators using various types of sewing machines and other equipment. These
operations include pressing and forming to make various shapes, making button holes,
stitching various types of buttons, labels, and other parts, and matching and sewing
different types of fabric cut panels to join together to make a complete shirt. The assembly
line consists of 64 operators with 70 workstations to complete 61 operations. Additional
operators and workstations are required to balance the workload following the line-
balancing theory. Data and parameters related to production efficiency have been collected
from the production floor for the 30 days to build the DT. Each individual station is denoted
by the letter "opxx", which comes after the operation number. Here, for the sake of
processing data, are certain crucial assumptions regarding the procedure;
● Operating hours for each day is 8.
● Multiple operations cannot be processed at the same time in one workstation. Some
manual operations can be processed in the same workstation one after another. The
processing of a single operation cannot be interrupted during processing and
restarted at a later time.
● Each workstation is located in its own distinct area, and the material is moved
manually from one operation to the next, and most human operators are at their own
location; few workers go from one workstation to another for workload balancing.
● Transportation time for material and movement of workers is incorporated into the
processing time.
55
● Processing times include breaks at work, unscheduled maintenance, malfunctions,
and troubleshooting, among other factors. Work on Preventive Maintenance is
performed outside normal business hours.
4.4.2 Digital Elements
In order to create the DT of a manufacturing plant, one has to have a solid grasp of
activities, the structure, and operations of the company or the selected factory. A process
map needs to be created for this case study following the proposed framework. The process
map will represent the OME that includes the sequence of the operations, types of
machines, and equipment. An operation bulletin can assist to prepare the map. Table-4.1
shows the operation bulletin along with the data required for building DT, and figure 4. 2
depicts the final product produced from the assembly line.
56
Table 4.1 Operation Bulletin of the Product.
Operati
on No
OPERATION
NAME
Machine
(M/C)
Cycle
Time (CT)
Operation
Target /Hour
Allocated
Manpower
COLLAR
OP01
Bon Sewing at
Panel
S/N
0.45
132
1.00
OP02
Match for
Collar Make
HP
0.25
240
0.50
OP03
Collar Make
S/N(V)
0.50
120
1.00
OP04
Collar Trim &
Turn
TM
0.24
255
0.50
OP05
Collar Forming
FM
0.23
267
0.50
OP06
Collar T/S
S/N
0.44
137
1.00
OP07
Band Rolling
With Tape Joint
S/N
0.44
137
1.00
OP08
Collar Band
Match
HP
0.47
127
1.00
OP09
Collar Band
Joint
S/N(V)
0.51
118
1.00
OP10
Collar Band T/S
S/N
0.45
132
1.00
OP11
Collar Band
Excess Trim
O/L
0.24
251
0.50
OP12
Collar Band
Hole
BH
0.24
251
0.50
OP13
Collar Band
Button Attach
BA
0.24
253
0.50
Front Part
OP14
Upper front part
Sewing
K/S
0.45
134
1.00
OP15
Lower front
placket rolling
S/N
0.47
129
1.00
OP16
Front neck
mark &
scissoring
HP
0.44
138
1.00
OP17
Body hole
BH
0.51
119
1.00
OP18
Mark for button
Attach
HP
0.50
120
1.00
OP19
Body button
Attach
BA
0.47
128
1.00
OP20
Extra Button
attach ( Small )
BA
0.11
533
0.25
OP21
Extra Button
attach (Big )
BA
0.11
533
0.25
57
OP22
Care Label Join
S/N
0.23
267
0.50
Epaulet
OP23
Epaulet Press
IM
0.46
130
1.00
OP24
Epaulet Tack
S/N
0.25
240
0.50
OP25
Epaulet Turning
TM
0.24
248
0.50
OP26
Epaulet T/S
S/N
0.75
80
1.50
OP27
Epaulet Mark &
Scissoring
HP
0.21
282
0.50
OP28
Epaulet joint
S/N
0.53
113
1.50
OP29
Epaulet Hole
BH
0.24
250
0.50
OP30
Epaulet Button
Attach
BA
0.24
250
0.50
Flap(02)
OP31
Flap Make With
Match
S/N(V)
0.47
128
1.00
OP32
Flap Trim &
Turn
TM
0.51
119
1.00
OP33
Flap T/S
S/N
0.41
146
1.00
OP34
Flap Edge O/L
4TO/L
0.42
144
1.00
OP35
Flap Hole
BH
0.21
282
0.50
OP36
Flap Joint
S/N
0.44
135
1.00
OP37
Flap T/S(02)
S/N
0.46
130
1.00
OP38
Mark For
Pocket Button
Attach
HP
0.21
282
0.50
OP39
Pocket Button
Attach
BA
0.23
267
0.50
Pocket (02)
OP40
Pocket Run
Stitch
S/N
0.48
126
1.00
OP41
Pocket Press
IM
0.94
64
2.00
OP42
Pocket Excess
trim
HP
0.72
83
1.50
OP43
Pocket joint
S/N
1.42
42
2.50
Back
Part(Plain)
OP44
Main label Join
S/N
0.45
134
1.00
OP45
Sub label join
S/N
0.73
82
1.50
OP46
Bk Yoke Joint
S/N
0.64
93
1.50
OP47
Back yoke press
& Excess Trim
IM
0.47
129
1.00
SLEEVE(Under
Placket)
OP48
Sleeve Press
IM
0.46
130
1.00
58
OP49
Sleeve Run
Stitch
S/N
0.86
70
2.00
OP50
Sleeve placket
mark & excess
trim
HP
0.32
190
1.00
OP51
Sleeve Tack
S/N
0.47
127
1.00
ASSEMBLY
OP52
Shoulder join
S/N
0.94
64
2.00
OP53
Shoulder
Excess trim
HP
0.44
136
1.00
OP54
Neck false tack
S/N
0.47
129
1.00
OP55
Sleeve join
5TO/L
0.93
64
2.00
OP56
Body side
Excess trim
O/L
0.21
280
0.50
OP57
Collar join &
Collar mark
S/N
0.93
64
2.00
OP58
Collar close
S/N
0.98
61
2.00
OP59
Side Seam
5TO/L
0.94
64
2.00
OP60
Thread Trim
HP
0.61
98
1.50
OP61
Bottom excess
trim & Bottom
hem
S/N
0.98
62
2.00
Note: S/N= Single Needle, HP= Helper, TM= Trimming Machine, FM= Forming Machine,
O/L= Overlock machine, BA= Button Attach
machine, K/S= Kansai Machine,
BH= Button hole, IM= Iron Machine, 4TO/L= 4 thread Overlock, 5TO/L= 5 thread overlock
59
Figure 4.3 Sketch of the final product produced from the assembly line. (Source: A
apparel manufacturing company in Bangladesh)
60
4.4.2.1 Process Map
From the Operation Bulletin, the process map for the assembly line will be created, which
will provide specifications for creating and running the simulation model and data
collection guidance. Although some of the operations were carried out in the same
workstation one after another, the process map showed them separately for ease of
understanding and simulation purposes. The process map depicts a total of 61 operations.
Operations were performed sequentially to produce the following components, followed
by assembly processes. The order of the individual components is as follows: Collar, Front
Part, Epaulet, Flap, Pocket, Back Part, Sleeve, and Assembly. The following are some of
the most important preconditions that must be met before the proposed method may be put
into action;
1. Depict the assembly scenario.
2. Collection and analysis of operational data and key parameters.
3. The DT concept simulates the behavior of OME. Simulation and optimization of
the assembly process based on a set of optimization constraints.
61
COLLAR
OP01
OP02
OP03
OP04
OP05
OP06
OP07
OP08
OP09
OP10
OP11
OP12
OP13
OP14
OP15
OP16
OP17
OP18
OP19
OP20
OP21
OP22
OP23
OP24
OP25
OP26
OP27
OP28
OP29
OP30
OP31
OP32
OP33
OP34
OP35
OP36
OP37
OP38
OP39
OP40
OP41
OP42
OP43
OP44
OP45
OP46
OP47
OP48
OP49
OP50
OP51
Front Part Epaulet Flap (02) Pocket (02) SLEEVE
(Under Plkt)
Back Part
(Plain)
OP52
OP53
OP54
OP55
OP56
OP57
OP58
OP59
OP60
OP61
ASSEMBLY
Operations Name
Figure 4.4 Process map of the woven shirt sewing assembly line.
62
4.4.2.2 Data Collection and Processing
Next, the data points need to be specified, which comes after the creation of the
process map. Proper planning helps to organize the data collection process, making it more
effective than collecting data points at random. Data collection via sensors, RFID, and
manual processes is required for a smooth transformation to be achieved. Additionally,
Establishing the limits and conditions of the data gathering is of the utmost importance.
In this case study, sewing assembly line operations along with cycle time for each
operation, production downtime due to machine breakdown and other issues, response time
to the issues, and total production output for 8 hours have been collected for the observed
time frame. The data collection format is given below. Data was collected for 30 days
during the period of May’ 2022 to June’ 2022, from the selected production line to get a
better picture of the situation. To reduce data collection efforts and to maintain data
accuracy, data were collected only for the assembly operations (last 10 operations), which
is considered a representative scenario for the whole production line. For proper workload
balancing of the whole production line, 16 operators were used for these 10 operations by
the floor management. Data were collected following the format given in Table 4.2, which
is the downtime of Day-1. As the efficiency of the line was not standard at the beginning
or during the learning curve period, the author didn’t collect those data to avoid
inconsistency. These data were collected manually as there was no digital system in place
to carry out this job.
The manufacturing components are in communication with one another virtually
and the system itself constantly assesses the manufacturing activities. In this case, live
machine status and information about produced quantities are being measured
continuously.
63
Table 4.2 Downtime collection of Day-1
Line
No :
A
Production
Unit- 3RD
Floor
Day-1
Productio
n Down
time
(minute)
Reason
Downtim
e start
Respons
e time
(minute)
Repair
Time
(minute
)
OP5
2
Shoulder
join
35
Troubleshootin
g
9:35
15
20
OP5
3
Shoulder
Excess trim
OP5
4
Neck false
tack
40
Machine
breakdown
8:50
20
20
OP5
5
Sleeve join
45
Machine
breakdown
2:30
18
27
OP5
6
Body side
Excess trim
OP5
7
Collar join
& Collar
mark
60
Machine
breakdown
10:40
25
35
OP5
8
Collar close
55
Machine
breakdown
1:30
22
33
OP5
9
Side Seam
OP6
0
Thread
Trim
OP6
1
Bottom
excess trim
& Bottom
hem
30
Troubleshootin
g
4:25
17
13
64
4.4.2.3 Simulation Execution
To carry out the simulation, ProModel (version 10.12.224, www.promodel.com)
software is used. The simulation was done in two situations; one is with the collected data
from the production floor, which represent the current picture. Another one is considering
the DT situation where relevant data is collected using sensors and RFID, and data is
transmitted to the DT to simulate and then receive recommendations from it.
To represent the current picture, which is basically before DT implementation,
collected data needs to be processed to simulate and model the physical assets. In ProModel
software, production downtime can be incorporated into cycle time to simulate the current
picture. In this case, total downtime of a particular machine as a fraction of total working
time will be added to the cycle time. Table 4. 3 shows the format of incorporation of
downtime into the cycle time of an operation.
Since comprehensive data was only collected for the last ten operations, efforts
were taken to ensure that material supply from prior operations did not affect the output of
those operations. Actual output from the production line was compared with the simulated
output to validate the formation of the simulation. Table 4.4 shows cycle time incorporating
downtime of day 1.
65
Table 4.3 Downtime incorporation into Cycle Time (CT)
Operatio
n No
Operation
Name
Cycle
Time
(minute)
Downtime
(minutes)
Working
minutes per
Day
Cycle time
incorporating
downtime
(minute)
OP54
Neck false
tack
0.47
40
480
0.55
.
66
Table 4.4 Cycle time incorporated with Downtime of day-1
Day-1
OP No
Operation Name
Cycle
Time
(minutes)
Downtim
e
(minute)
Cycle time
incorporating
down time
(minute)
Standard
Deviation
OP52
Shoulder join
0.94
35
1.01
0.28
Shoulder join
0.94
0.94
0.28
OP53
Shoulder Excess
trim
0.44
0.44
0.13
OP54
Neck false tack
0.47
40
0.55
0.14
OP55
Sleeve join
0.93
45
1.02
0.28
Sleeve join
0.93
0.93
0.28
OP56
Body side Excess
trim
0.22
0.22
0.07
OP57
Collar join &
Collar mark
0.93
60
1.06
0.28
Collar join &
Collar mark
0.93
0.93
0.28
OP58
Collar close
0.98
55
1.09
0.29
Collar close
0.98
0.98
0.29
OP59
Side Seam
0.94
0.94
0.28
Side Seam
0.94
0.94
0.28
OP60
Thread Trim
0.61
0.61
0.18
Thread Trim
0.61
0.61
0.18
OP61
Bottom excess trim
& Bottom hem
0.98
30
1.04
0.29
Bottom excess trim
& Bottom hem
0.98
0.98
0.29
67
While carrying out the simulation with ProModel, cycle time was considered. Figure 4. 5
shows the simulated view of the OME, where 18 workstations were used for 10 operations
and 16 operators. Using the data provided in Table 4.4, the simulated output was found
863, which is shown in Figure 4.6. Then simulated output was compared with the actual
production output received from the production floor. The comparison was done for the 30
days, which is shown in table 4.5. Although the deviation in some cases were around 2%,
overall deviation is found within the acceptable limit. Numerous variables influence the
divergence, including the fact that the material flow between machines is not consistent
during the movement and that human operators do not operate at a constant pace, whereas
the simulation was based on the mean speed.
68
Figure 4.5 Simulated view of the OME
69
Figure 4.6 Simulated output of Day-1
70
Table 4.5 Comparison between actual output vs the simulated output.
Simulated
output
Actual output of
Production Line-A
Deviation
Percentage
Day-1
860
870
10
1.1%
Day-2
892
900
8
0.9%
Day-3
896
880
-16
-1.8%
Day-4
873
860
-13
-1.5%
Day-5
880
875
-5
-0.6%
Day-6
847
865
18
2.1%
Day-7
877
880
3
0.3%
Day-8
897
890
-7
-0.8%
Day-9
906
900
-6
-0.7%
Day-10
897
880
-17
-1.9%
Day-11
858
870
12
1.4%
Day-12
898
890
-8
-0.9%
Day-13
872
880
8
0.9%
Day-14
854
865
11
1.3%
Day-15
873
880
7
0.8%
Day-16
881
870
-11
-1.3%
Day-17
852
860
8
0.9%
Day-18
869
880
11
1.3%
Day-19
902
890
-12
-1.3%
Day-20
889
875
-14
-1.6%
Day-21
875
880
5
0.6%
Day-22
861
865
4
0.5%
Day-23
878
880
2
0.2%
Day-24
894
890
-4
-0.4%
Day-25
879
875
-4
-0.5%
Day-26
868
880
12
1.4%
Day-27
893
890
-3
-0.3%
Day-28
903
900
-3
-0.3%
Day-29
869
865
-4
-0.5%
Day-30
874
875
1
0.1%
Average
878.9
878.7
-0.2
-0.03%
71
A dynamic simulation model lies at the heart of the DT. When compared to online
simulation, offline simulation often has a small feedback loop, executes the whole timespan
at once, and seldom helps the user prepare for the next cycle as the scenario evolves. In the
off-line simulation, workload balancing of the assembly line is carried out based on the
ideal situation and ideal skills of the operators, and the ideal speed of the machine, which
does not consider the dynamic situation of the assembly line. Production downtime for
various reasons, work-pace variation of the operators, and unplanned breaks taken by
operators make the situation far from the ideal scenario. Thus, initial line balancing done
based on an ideal scenario would not offer an effective solution for dynamic situations.
Adaptive simulation enables an effective solution for real-time scenarios. DT based on
adaptive simulations give decision assistance, real-time data input, and flexibility. For this
case study, there are two types of real-time data that must be collected: machine status (if
it is functioning) and production status (quantities being generated at a given moment).
After obtaining a dynamic dataset, the simulation updates to include the new
information in lieu of the old. In this case study, production information will be gathered
and analyzed every half hour in order to propose a solution for a real-world issue.
Frequency can be increased by reducing the duration from 30 mins to 20 or 15 mins.
However, in this case, the authors feel justified to simulate every half an hour. When an
operation stops working, a sensor detects it and sends information to the users with
recommendations from a list of prepared alternative solutions. At the same time, it helps
to reduce the response time so that operations can be made functional quickly. Production
downtime along with the inconsistent pace of operators creates imbalances in the workload.
RFID collects the production information of each workstation and fed into the simulation
72
to identify the bottleneck operations. According to the pro model, bottleneck operations
are identified as the highest utilized locations (operations). For Day-1 data, the “Neck False
Tack” operation was identified as a bottleneck, which is shown in figure 4.7. The scenario
clearly represented the day-1 practical situation of the production line. Users take steps to
eliminate the bottleneck and balance the workload. The setup helps the assembly line to
improve efficiency by improving the bottleneck operations and improving the quick
response for production downtime. The same data has been used for this case to compare
the outcome with the previous one.
Bottleneck operations can be eliminated in many ways such as by adding excess
machines or equipment or adding additional manpower or assigning tasks to the workers
who have the additional capacity. In the DT implemented scenario, the response time for
any troubleshooting or any machine breakdown would be close to zero as the user would
be able to get a notification when the machine or equipment breaks or when any other
troubles are faced by the production line. Table 4. 6 shows the reduced downtime in
comparison to table 4.4.
73
Figure 4.7 Bottleneck operations of Day 1.
74
Table 4. 6 Cycle Time (CT) of day 1 after DT implementation
Operatio
n No
Operation Name
Cycle
Time
(minutes
)
Downtime
(minutes)
Cycle time
incorporating
down time
(minute)
Standard
Deviatio
n
OP52
Shoulder join
0.94
20
0.98
0.28
Shoulder join
0.94
0.94
0.28
OP53
Shoulder Excess trim
0.44
0.44
0.13
OP54
Neck false tack
0.47
20
0.51
0.14
OP55
Sleeve join
0.93
27
0.99
0.28
Sleeve join
0.93
0.93
0.28
OP56
Body side Excess trim
0.22
0.22
0.07
OP57
Collar join & Collar mark
0.93
35
1.00
0.28
Collar join & Collar mark
0.93
0.93
0.28
OP58
Collar close
0.98
33
1.05
0.29
Collar close
0.98
0.98
0.29
OP59
Side Seam
0.94
0.94
0.28
Side Seam
0.94
0.94
0.28
OP60
Thread Trim
0.61
0.61
0.18
Thread Trim
0.61
0.61
0.18
OP61
Bottom excess trim &
Bottom hem
0.98
13
1.01
0.29
Bottom excess trim &
Bottom hem
0.98
0.98
0.29
75
The data has been processed following the ProModel software’s instruction and
utilized to simulate the DT scenario to see the improved performance. Figure 4.8 shows
that bottleneck operations are reduced, and figure 4.9 shows the increased production
output for Day 1, which is 894 pcs instead of the actual output of 870 pcs.
76
Figure 4.8 Bottleneck operations after DT implementation of Day-1
Figure 4.9 Production output at DT implemented scenario of Day-1
77
Following the similar method, simulation was carried out using the data of another day
(Day-11) before and after DT. Figure 4.10 and 4.11 shows the bottleneck operations
respectively before and after DT implemented scenarios. Figure 4.12 and 4.13 shows the
respective outputs.
Figure 4.10 Bottleneck operations of Day-11.
Figure 4.11 Bottleneck operations of Day-11 after DT implementation.
78
Figure 4.12 Simulated Production output of Day-11
Figure 4.13 Production output after DT implemented scenario of Day-11.
79
After analyzing the bottleneck operations of day-1 and day-11, it was evident that “Neck
false Tack” was the bottleneck operation in day-1 whereas “Bottom Excess trim hem” was
the bottleneck operation for day-11. Similarly, bottleneck operations vary after carrying
out the simulations for each day. In that case, simulation was performed based each day’s
data. If the simulations were performed on hourly basis based on the real time data,
bottleneck operation would found different in each case. This comparison demonstrates
that bottleneck operation changes dynamically and DT identified the bottleneck operation
based on real time scenario.
Simulation was carried out for the rest of the days as well based on the collected
down time, and incorporated into cycle time. Data of downtime and other related
information are given in the appendix. Output of rest of the days of DT implemented
scenario are compared with the output of actual production, which is given in table 4.7.
Outcome shows significant improvement from the actual production.
4.4.3 Integration
Information flows both ways within the DT. First, real-time data from the
manufacturing process is collected, processed, and updated inside the DT. And secondly,
the DT displays outcomes with recommendations, which is then implemented within the
manufacturing process. Most of the information for this case study was gathered by hand
and then processed in a third-party program. After processing, ProModel will run the
simulation to provide an optimized solution. The solution should have a recommendation
list as well, from where the user will decide what action should be taken for the OME. In
80
this case study, users will convey the recommendations to OME and implement them
accordingly.
4.4.4 Model Validation
In order to verify the accuracy of the model, thirty days' actual data taken directly from
the factory floor are compiled. These data are first put to use inside the model's simulation
process as starting data. The simulated output is compared with the actual output to validate
the formation of the model. Then, Data has been processed following the guidelines of the
DT implementation. The processed data has been used for simulation further to see the
outcome of DT implementations. Then the output of the basic simulation is compared with
the DT output. The outcome of the DT shows huge potential. The comparison between
actual production output and DT output is shown in the table 4.7.
81
Table 4.7 Difference between actual production out and after DT implementation
Simulated
output before
DT (pieces)
Actual output
of the
Production
Line-A
(pieces)
Output
After
Digital
Twin
Implementa
tion (pieces)
Output difference
between DT and
actual (pieces)
Day-1
860
870
894
24
Day-2
892
900
932
32
Day-3
896
880
931
51
Day-4
873
860
905
45
Day-5
880
875
917
42
Day-6
847
865
881
16
Day-7
877
880
914
34
Day-8
897
890
930
40
Day-9
906
900
937
37
Day-10
897
880
930
50
Day-11
858
870
899
29
Day-12
898
890
933
43
Day-13
872
880
906
26
Day-14
854
865
894
29
Day-15
873
880
911
31
Day-16
881
870
915
45
Day-17
852
860
893
33
Day-18
869
880
909
29
Day-19
902
890
942
52
Day-20
889
875
929
54
Day-21
875
880
915
35
Day-22
861
865
901
36
Day-23
878
880
918
38
Day-24
894
890
934
44
Day-25
879
875
919
44
Day-26
868
880
908
28
Day-27
893
890
933
43
Day-28
903
900
943
43
Day-29
869
865
909
44
Day-30
874
875
914
39
82
CHAPTER 5: RESULT AND DISCUSSION
The developed DT of the apparel assembly line clearly addressed and solved the
problem of bottleneck operations identifications in real-time and recommended the
solutions. This setup also developed a method to communicate real time information with
the users. This proved the applicability of the proposed methodology of DT for apparel
manufacturing plants. Manufacturing plant has to deal with so many unexpected situations
because of its complex nature. Line imbalance can be created for various reasons at any
point of time. The DT model will respond to the situation in two ways; one approach is
communicating the problem or unexpected situations instantly with the responsible
persons, and the other one is providing an optimized solution. Before the DT
implementation, response to solve the problem was delayed as the communication was
carried out in a traditional way. Solutions can vary depending on the specific situation of
the physical elements. This will allow the management to overcome the challenges of the
apparel manufacturing plants.
The outcome of the developed DT for the assembly sewing line prevailed over the
traditional method of managing the unexpected situations. Unexpected situations create
production downtime and thus loss of efficiency or production output. In the case of DT,
users will be notified instantly as the communication will be carried out digitally unlike the
traditional way of communication. In the DT implemented situation, lines will be balanced
based on the actual data collected from the output of each machine. In this case study,
production output and production downtime for machine breakdown and for other
problems. Also, time to respond to the problem and time to repair the problem was
collected for the 30 days of a sewing assembly line. The first Simulation was carried out
83
using the data of actual cycle time incorporating the production downtime. The simulated
outputs of the 30 days compared with actual outputs of the production line. The difference
is within considerable range. This proves the accuracy of simulation logic considered here.
Next, simulation was performed considering the elimination of response time. This will
result in significant reduction of production downtime. Actual production output of the 30
days is compared with the simulated output. The production output after the DT
implementation in an apparel assembly line significantly outperforms the actual production
output of the production line. Below line graphs 5.1 shows the production output after DT
implementation is considerably more than that of actual production and simulated output
of practical scenarios.
Figure 5.1 Comparison of the production output among different scenarios.
780
800
820
840
860
880
900
920
940
960
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Produced Quantity
Days
Case Comparison
Simulated output Actual output of the Production Line-A Output After Digital Twin Implementation
84
The expected output of DT would be much more than this because in the case of DT
implemented situation, steps will be taken after identifying the bottleneck operations. The
case study and the outcome of it presented here demonstrates the potentiality of the
successful implementation of DT in assembly sewing line. The improved performance
indicates that the investment to implement the DT would be appropriate though the cost-
benefit ratio was not analyzed here.
5.1 Implication of the Study
This research is probably the first research of DT implemented for apparel
manufacturing plants. The methodology and case study will pave the way for many further
research and practical implementations in the apparel industry for solving various current
and future problems. This study aims to contribute to both academic discourse and practical
implementation. This section outlines the theoretical and practical implications and
contributions of the study.
5.1.1 Implications in Theory
This study proposed a comprehensive methodology where various elements of DT
are defined for apparel manufacturing. This methodology outlined what types of
information should be focused, how these can be collected and how heterogeneous data
can be processed to use for analysis. This makes the process of creating DT easy for
businesses, particularly with the assistance of a previously created simulation tool. The
research emphasized that the physical resources of apparel manufacturing can be visualized
digitally and data can be communicated in both ways, physical elements to digital elements
and digital to physical. Three types of information can be transferred. These include the
85
functional capabilities, the technological qualities, and the real-time status. This conceptual
framework will help other researchers to develop DT for other labor intensive
manufacturing plants. This study will pave the way of developing DT for other areas
besides manufacturing activities. These can be managing supply chain activities,
maintaining machine maintenance, controlling the material flow and so on. While
developing the DT, this study considered current practices of apparel manufacturing
industries in certain countries. This study can be applied to apparel industries with minor
modifications.
Circular economy concept is going to be a must do thing for the fashion brands as
the fashion and textile business is one of the most polluting sectors in the world. This is
mostly due to the fact that the volume of output in this industry eclipses that of virtually all
other industries. Collecting accurate data about amount and types of fabric leftover, and
designing garments based on these is vital for circular fashion (Aus et al., 2021). This study
will help the management to achieve the goal of circular economy for fashion industries by
tracking real time information about the materials and providing optimized solutions of
fashionable design with these materials.
Using eco-friendly materials is a popular business technique in the fashion industry,
which can affect a product's cost and environmental impact during manufacture. Fashion
uses 93 billion cubic meters of water annually and 10% of the world's yearly carbon
emissions. Textile treatment and dyeing contribute to roughly 20% of water contamination
in the fashion sector, according to a 2017 research. Blockchain technology is considered
as a potential concept for sustainable practices in the fashion industry. This study can assist
the researcher to prepare a roadmap to implement block chain in the fashion industry as it
86
has the potential of tracking and verifying each and every material required to make the
final product.
This research would encourage many researchers and entrepreneurs to digitalize
the manufacturing processes and operations by adopting other digital concepts (Industry
4.0, IoT, Artificial Intelligence, and Machine learning and so on) as well.
5.1.2 Implications in Practice
The outcome of this research also shows that it has the potential of applying in various
cases or problems of apparel manufacturing facilities. The developed method that was
offered in this research suggested a strategy that could be used to transfer information from
the physical world of an apparel manufacturing facility to its digital counterpart. This
method creates an environment in which data can be collected dynamically, and where it
can be further processed and evaluated in an efficient manner. Therefore, the notion of a
DT indicates the potential to be employed as a problem-solver for a variety of difficulties
that are present in the garment manufacturing industry. In order to provide the industry
with a model for how it should go about creating DT, a course of action is provided and a
resource virtualization approach has been developed. The proposed approach takes its cues
from the results of previous attempts at resource virtualization, with a particular focus on
how easily the proposed solution may be implemented
Apparel manufacturers are facing many other challenges besides efficiency loss
owing to production down time. Short shipment, delay in delivery, high material wastage,
high level of reworks and rejects, lack of standardization of the processes and operations,
long lead time, high level of monitoring and supervision, efficiency loss because of
87
frequent style changeovers are major among the challenges. This study will assist
managers/engineers/management to implement DT to overcome these challenges.
This paper would assist management to take initiatives to digitalize the processes
and operations of other industries as well with the help of software, sensors, RFID,
actuators and other technologies. This will create huge demand for software development,
and manufacturing of related technologies, this in turn will create many job opportunities.
As a result, manufacturers also would be able to produce these products at less cost due to
the high demand from apparel manufacturers.
This study will also help to reduce the living expenses of consumers as apparel
manufacturers would be able to reduce manufacturing cost by improving efficiency and
productivity in most of the functions.
88
CHAPTER 6: CONCLUSION
To rapidly produce customized fashion products at fluctuating scales, the apparel
manufacturing industry must adopt a new digital approach to handling issues and problems.
The DT generates tangible value by assisting production managers/supervisors or industrial
engineers in making crucial strategic decisions. The DT has various uses across the
manufacturing process, from receiving raw materials to delivering completed items, and
managing processes, operations, and resources efficiently. This type of research is
anticipated to help the effective execution of activities, the better exploitation of resources,
the reduction of lead times, and the improvement of due date dependability. This directly
benefits the company's bottom line by reducing manufacturing costs.
The second chapter examines existing research and identifies research gaps. Very
few studies focus on the use of DT to garment manufacturing facilities. It is evident from
the literature research that clothing manufacturing sectors are not adopting the DT idea due
to a lack of clear guidance and real-world examples. Some case studies pertain to different
industrial areas, however, the majority are too hypothetical or oversimplified. It is difficult
to modify and customize these for use in troublesome scenarios in the textile manufacturing
industry. The solution has resilience, but it is quite field-specific. Visualization is lacking
and failed to create an effect on the actual textile manufacturing factory, which is subject
to fluctuating conditions and impending uncertainty. The third chapter offers a DT
technique for apparel manufacturing to harness current advancements in computing and
networking to enable DT in this industry. The suggested technique is general for this
industry, and offers an easy-to-follow guideline.
89
A case study is presented in the fourth chapter to validate the suggested
methodology of apparel manufacturing. Manufacturing of woven shirts is selected for this
case study. The nature of the manufacturing process is very dynamic and diverse, and
involves many human-operated operations. The ultimate goal is to build a DT of the
processes of clothing manufacturing so that this can be used in decisions making at the
production line level. The economic viability of an industry is dependent on the ongoing
development of the swing line. An adaptive simulation-based method, as opposed to a
conventional simulation or an offline simulation, enables proactive action and improved
management of uncertainties throughout the production process. The DT utilizes adaptive
simulation to continually evaluate the real-time scenario and offer feedback depending on
the current state. It aids in identifying the process's bottleneck and offers notification for a
prompt solution to the situation. The offered case study was examined with production
managers and industrial engineers from the world's major garment manufacturers. It is
highly valued and acknowledged as a valuable asset by them. According to them, the
proposed strategy can greatly increase the assembly line's efficiency and will also lead to
opportunities in other areas. Therefore, it may drastically reduce operational expenses,
resulting in substantial bottom-line gains. The technology may be broadly integrated into
the existing infrastructure.
Overall, this work contributes the following:
● An assessment of research papers pertaining to the field of DT focused in the
manufacturing industry and other areas.
● Research gap in the field of DT is identified and a DT methodology for apparel
manufacturing industries is created.
90
● To validate the suggested methodology, a case study utilizing real-world data is
presented in the field of clothing making to validate the proposed methodology.
The results are presented to highlight the possible advantages of a DT.
● The case study may serve as a step-by-step direction for clothing manufacturers to
implement DT. This case study is selected to address one of the important and
crucial challenges facing the management of garment makers.
● ProModel simulation software is used to demonstrate the outcome of the
implementation and compare results of DT implemented scenarios with the actual
production output of the production line.
6.1 Limitations of the Study
Nevertheless, there are a few recommendations for making the situation better. The
technique for implementing this model is case-specific, and the general execution method
may not apply to products with more complexity. More case studies can be performed to
address more realistic problems to make the methodology more generic. There is no
instruction on how to address the missing methodology step or component. The suggested
approach and case study rely largely on human input and manual software operation.
Integration of heterogeneous data with the program and the software's feedback are not
automated. Other software might be beneficial to make the system more automated or self-
sufficient. The data collecting and integration technologies are not described in detail here.
This is included in the future work scope. Some of the limitations are outside the scope of
the study. Integrating decision-makers or users into the DT is out of the scope of the
research.
91
6.2 Future Work
Implementation of DT requires cost and the right expertise. Future studies should
focus on estimating the cost of implementation of DT and cost-benefit ratio of this type of
investment. The reliability of the model can be enhanced by incorporating more data,
variables, and proven algorithms. Neural networks, advanced fuzzy theorems, and other
evolutionary mathematical models can be used to enhance the outcome of adaptive
simulations. Future research also needs to focus on advanced data transmission
technologies and protocols. Fuzzy applications or machine learning can be researched and
applied for identifying the best alternative solution to the real problem of apparel
manufacturing.
92
REFERENCE
Alexopoulos, K., Nikolakis, N., & Chryssolouris, G. (2020). Digital twin-driven
supervised machine learning for the development of artificial intelligence applications
in manufacturing. International Journal of Computer Integrated Manufacturing,
33(5), 429–439. https://doi.org/10.1080/0951192X.2020.1747642
Aus, R., Moora, H., Vihma, M., Unt, R., Kiisa, M., & Kapur, S. (2021). Designing for
circular fashion: integrating upcycling into conventional garment manufacturing
processes. Fashion and Textiles, 8(1). https://doi.org/10.1186/s40691-021-00262-9
Bae, J. H., & May-Plumlee, T. (2005). Customer focused textile and apparel manufacturing
systems: Toward an effective ecommerce model. Journal of Textile and Apparel,
Technology and Management, 4(4), 161–179.
Bao, J., Guo, D., Li, J., & Zhang, J. (2019). The modelling and operations for the digital
twin in the context of manufacturing. Enterprise Information Systems, 13(4), 534–
556. https://doi.org/10.1080/17517575.2018.1526324
Beňo, L., Pribiš, R., & Leskovský, R. (2019). Processing data from OPC UA server by
using Edge and Cloud computing. IFAC-PapersOnLine, 52(27), 240–245.
https://doi.org/10.1016/j.ifacol.2019.12.645
Bertola, P., & Teunissen, J. (2018). Fashion 4.0. Innovating fashion industry through
digital transformation. Research Journal of Textile and Apparel, 22(4), 352–369.
https://doi.org/10.1108/RJTA-03-2018-0023
Boschert, S., & Rosen, R. (2016). Digital twin-the simulation aspect. In Mechatronic
Futures: Challenges and Solutions for Mechatronic Systems and Their Designers (pp.
59–74). Springer International Publishing. https://doi.org/10.1007/978-3-319-32156-
93
1_5
Botkina, D., Hedlind, M., Olsson, B., Henser, J., & Lundholm, T. (2018). Digital Twin of
a Cutting Tool. Procedia CIRP, 72, 215–218.
https://doi.org/10.1016/j.procir.2018.03.178
Braglia, M., Marrazzini, L., Padellini, L., & Rinaldi, R. (2021). Managerial and Industry
4.0 solutions for fashion supply chains. Journal of Fashion Marketing and
Management, 25(1), 184–201. https://doi.org/10.1108/JFMM-12-2019-0285
Braun, D., Biesinger, F., Jazdi, N., & Weyrich, M. (2020). A concept for the automated
layout generation of an existing production line within the digital twin. Procedia
CIRP, 97, 302–307. https://doi.org/10.1016/j.procir.2020.05.242
Cheng, Z., & Kuzmichev, V. E. (2018). Digital twin and men’s underwear design. IOP
Conference Series: Materials Science and Engineering, 459(1).
https://doi.org/10.1088/1757-899X/459/1/012075
Chutima, P. (2020). Research trends and outlooks in assembly line balancing problems.
Engineering Journal, 24(5), 93–134. https://doi.org/10.4186/ej.2020.24.5.93
Cwikła, G., & Foit, K. (2017). Problems of integration of a manufacturing system with the
business area of a company on the example of the Integrated Manufacturing Systems
Laboratory. MATEC Web of Conferences, 94, 1–11.
https://doi.org/10.1051/matecconf/20179406004
Dal Forno, A. J., Bataglini, W. V., Steffens, F., & Ulson de Souza, A. A. (2021). Industry
4.0 in textile and apparel sector: a systematic literature review. Research Journal of
Textile and Apparel. https://doi.org/10.1108/RJTA-08-2021-0106
Deepthi, B., & Bansal, V. (2022). Industry 4.0 in Textile and Apparel Industry: A
94
Systematic Literature Review and Bibliometric Analysis of Global Research Trends.
Vision: The Journal of Business Perspective, 097226292211302.
https://doi.org/10.1177/09722629221130233
Derakhshan, Roozbeh, Maria Orlowska, and Xue Li. 2007. “RFID Data Management:
Challenges and Opportunities.” 2007 IEEE International Conference on RFID, IEEE
RFID 2007 175–82. doi: 10.1109/RFID.2007.346166.
Ding, K., Chan, F. T. S., Zhang, X., Zhou, G., & Zhang, F. (2019). Defining a Digital
Twin-based Cyber-Physical Production System for autonomous manufacturing in
smart shop floors. International Journal of Production Research, 57(20), 6315–6334.
https://doi.org/10.1080/00207543.2019.1566661
Farhadi, A., Lee, S. K. H., Hinchy, E. P., O’Dowd, N. P., & McCarthy, C. T. (2022). The
Development of a Digital Twin Framework for an Industrial Robotic Drilling Process.
Sensors, 22(19). https://doi.org/10.3390/s22197232
Flores, H. S., Paucar, L. F., Castro, P., Marcelo, E., & Alvarez, J. C. (2021). Increased
Efficiency in a Garment Sector by the Integration of Lean Manufacturing Tools. 2021
IEEE International Conference on Industrial Engineering and Engineering
Management, IEEM 2021, 654–658.
https://doi.org/10.1109/IEEM50564.2021.9672821
Fuller, A., Fan, Z., Day, C., & Barlow, C. (2020). Digital Twin: Enabling Technologies,
Challenges and Open Research. IEEE Access, 8, 108952–108971.
https://doi.org/10.1109/ACCESS.2020.2998358
Gökalp, E., Gökalp, M. O., & Eren, P. E. (2019). Industry 4.0 revolution in clothing and
apparel factories: Apparel 4.0. Industry 4.0 from the MIS Perspective, September,
95
169–183. https://doi.org/10.3726/b15120/21
Grieves, M., & Vickers, J. (2016). Digital twin: Mitigating unpredictable, undesirable
emergent behavior in complex systems. Transdisciplinary Perspectives on Complex
Systems: New Findings and Approaches, August, 85–113.
https://doi.org/10.1007/978-3-319-38756-7_4
Guo, Z. X., Wong, W. K., Leung, S. Y. S., & Li, M. (2011). Applications of artificial
intelligence in the apparel industry: A review. Textile Research Journal, 81(18),
1871–1892. https://doi.org/10.1177/0040517511411968
He, B., & Bai, K. J. (2021). Digital twin-based sustainable intelligent manufacturing: a
review. Advances in Manufacturing, 9(1), 1–21. https://doi.org/10.1007/s40436-020-
00302-5
Henrique dos Santos, C., Teodoro Gabriel, G., Victor Soares do Amaral, J., Arnaldo Barra
Montevechi, J., & Antônio de Queiroz, J. (n.d.). Decision-making in a fast fashion
company in the Industry 4.0 era: a Digital Twin proposal to support operational
planning. https://doi.org/10.1007/s00170-021-07543-z/Published
Hoque, M. A., Rasiah, R., Furuoka, F., & Kumar, S. (2021). Technology adoption in the
apparel industry: insight from literature review and research directions. Research
Journal of Textile and Apparel, 25(3), 292–307. https://doi.org/10.1108/RJTA-08-
2020-0090
Ivanov, D., Dolgui, A., & Sokolov, B. (2019). The impact of digital technology and
Industry 4.0 on the ripple effect and supply chain risk analytics. International Journal
of Production Research, 57(3), 829–846.
https://doi.org/10.1080/00207543.2018.1488086
96
Jin, W., He, Z., & Wu, Q. (2022). Robust optimization of resource-constrained assembly
line balancing problems with uncertain operation times. Engineering Computations
(Swansea, Wales), 39(3), 813–836. https://doi.org/10.1108/EC-01-2021-0061
Jung, W. K., Kim, H., Park, Y. C., Lee, J. W., & Suh, E. S. (2022). Real-time data-driven
discrete-event simulation for garment production lines. Production Planning and
Control, 33(5), 480–491. https://doi.org/10.1080/09537287.2020.1830194
Kumar, N., & Mahto, D. (2013). Assembly Line Balancing: A Review of Developments
and Trends in Approach to Industrial Application. Global Journal of Researches in
Engineering Industrial Engineering, 13(2), 29–50.
Latif, H., & Starly, B. (2020). A simulation algorithm of a digital twin for manual assembly
process. 48, 932–939. https://doi.org/10.1016/j.promfg.2020.05.132
Lattanzi, L., Raffaeli, R., Peruzzini, M., & Pellicciari, M. (2021). Digital twin for smart
manufacturing: a review of concepts towards a practical industrial implementation.
International Journal of Computer Integrated Manufacturing, 34(6), 567–597.
https://doi.org/10.1080/0951192X.2021.1911003
Lin, S. H., Moore, M. A., Kincade, D. H., & Avery, C. (2002). Dimensions of apparel
manufacturing strategy and production management. International Journal of
Clothing Science and Technology, 14(1), 46–60.
https://doi.org/10.1108/09556220210420336
Liu, J., Zhou, H., Liu, X., Tian, G., Wu, M., Cao, L., & Wang, W. (2019). Dynamic
Evaluation Method of Machining Process Planning Based on Digital Twin. IEEE
Access, 7, 19312–19323. https://doi.org/10.1109/ACCESS.2019.2893309
Liu, N., Chow, P. S., & Zhao, H. (2020). Challenges and critical successful factors for
97
apparel mass customization operations: recent development and case study. In Annals
of Operations Research (Vol. 291, Issues 1–2). Springer US.
https://doi.org/10.1007/s10479-019-03149-7
Liu, Q., Zhang, H., Leng, J., & Chen, X. (2019). Digital twin-driven rapid individualised
designing of automated flow-shop manufacturing system. International Journal of
Production Research, 57(12), 3903–3919.
https://doi.org/10.1080/00207543.2018.1471243
Lopes, M. R., Costigliola, A., Pinto, R., Vieira, S., & Sousa, J. M. C. (2020).
Pharmaceutical quality control laboratory digital twin–A novel governance model for
resource planning and scheduling. International Journal of Production Research,
58(21), 6553–6567. https://doi.org/10.1080/00207543.2019.1683250
Lu, Y., Frechette, S., Lu, Y., & Frechette, S. (2016). Current standards landscape for smart
manufacturing systems current standards landscape for smart manufacturing systems.
In Manufacturing.Report (Issue February).
Lu, Y., Liu, Z., & Min, Q. (2021). A digital twin-enabled value stream mapping approach
for production process reengineering in SMEs. International Journal of Computer
Integrated Manufacturing, 34(7–8), 764–782.
https://doi.org/10.1080/0951192X.2021.1872099
Luo, W., Hu, T., Zhang, C., & Wei, Y. (2019). Digital twin for CNC machine tool:
modeling and using strategy. Journal of Ambient Intelligence and Humanized
Computing, 10(3), 1129–1140. https://doi.org/10.1007/s12652-018-0946-5
Majeed, M. A. A., & Rupasinghe, T. D. (2017). Internet of things (IoT) embedded future
supply chains for industry 4.0: An assessment from an ERP-based fashion apparel and
98
footwear industry. International Journal of Supply Chain Management, 6(1), 25–40.
Ma, Meng, Ping Wang, and Chao Hsien Chu. 2013. “Data Management for Internet of
Things: Challenges, Approaches and Opportunities.” Proceedings - 2013 IEEE
International Conference on Green Computing and Communications and IEEE
Internet of Things and IEEE Cyber, Physical and Social Computing, GreenCom-
IThings-CPSCom 2013 1144–51. doi: 10.1109/GreenCom-iThings-
CPSCom.2013.199.
Manaye, M. (2019). Line Balancing Techniques for Productivity Improvement
Improvement. International Journal of Mechanical and Industrial Technology,
7(June), 89–104. https://www.researchgate.net/publication/333310098
Manglani, H., Hodge, G. L., & Oxenham, W. (2019). Application of the Internet of Things
in the textile industry. Textile Progress, 51(3), 225–297.
https://doi.org/10.1080/00405167.2020.1763701
Nikolakis, N., Alexopoulos, K., Xanthakis, E., & Chryssolouris, G. (2019). The digital
twin implementation for linking the virtual representation of human-based production
tasks to their physical counterpart in the factory-floor. International Journal of
Computer Integrated Manufacturing, 32(1), 1–12.
https://doi.org/10.1080/0951192X.2018.1529430
Noor, A., Saeed, M. A., Ullah, T., Uddin, Z., & Ullah Khan, R. M. W. (2022). A review of
artificial intelligence applications in apparel industry. In Journal of the Textile
Institute (Vol. 113, Issue 3, pp. 505–514). Taylor and Francis Ltd.
https://doi.org/10.1080/00405000.2021.1880088
O’Sullivan, J., O’Sullivan, D., & Bruton, K. (2020). A case-study in the introduction of a
99
digital twin in a large-scale smart manufacturing facility. Procedia Manufacturing,
51(2019), 1523–1530. https://doi.org/10.1016/j.promfg.2020.10.212
Onaji, I., Tiwari, D., Soulatiantork, P., Song, B., & Tiwari, A. (2022). Digital twin in
manufacturing: conceptual framework and case studies. International Journal of
Computer Integrated Manufacturing, 35(8), 831–858.
https://doi.org/10.1080/0951192X.2022.2027014
Pal, K., & Yasar, A. U. H. (2020). Internet of Things and Blockchain Technology in
Apparel Manufacturing Supply Chain Data Management. Procedia Computer
Science, 170, 450–457. https://doi.org/10.1016/j.procs.2020.03.088
Park, K. T., Jeon, S. W., & Noh, S. Do. (2022). Digital twin application with horizontal
coordination for reinforcement-learning-based production control in a re-entrant job
shop. International Journal of Production Research, 60(7), 2151–2167.
https://doi.org/10.1080/00207543.2021.1884309
Park, K. T., Nam, Y. W., Lee, H. S., Im, S. J., Noh, S. Do, Son, J. Y., & Kim, H. (2019).
Design and implementation of a digital twin application for a connected micro smart
factory. International Journal of Computer Integrated Manufacturing, 32(6), 596–
614. https://doi.org/10.1080/0951192X.2019.1599439
Park, K. T., Son, Y. H., & Noh, S. Do. (2020). The architectural framework of a cyber
physical logistics system for digital-twin-based supply chain control. International
Journal of Production Research, 0(0), 1–22.
https://doi.org/10.1080/00207543.2020.1788738
Polini, W., & Corrado, A. (2020). Digital twin of composite assembly manufacturing
process. International Journal of Production Research, 58(17), 5238–5252.
100
https://doi.org/10.1080/00207543.2020.1714091
Qi, Q., & Tao, F. (2018). Digital Twin and Big Data Towards Smart Manufacturing and
Industry 4.0: 360 Degree Comparison. IEEE Access, 6, 3585–3593.
https://doi.org/10.1109/ACCESS.2018.2793265
Rajput, D., Kakde, M., Chandurkar, P., & Raichurkar, P. P. (2018). Enhancing efficiency
and productivity of garment industry by using different techniques. International
Journal on Textile Engineering and Processes, 4(1), 5–8. http://engineering-
shirpur.nmims.edu/docs/enhancing-efficiency-and-productivity-of-garment-
industry-by-using-different-techniques.pdf
Redelinghuys, A. J. H., Basson, A. H., & Kruger, K. (2020). A six-layer architecture for
the digital twin: a manufacturing case study implementation. Journal of Intelligent
Manufacturing, 31(6), 1383–1402. https://doi.org/10.1007/s10845-019-01516-6
Shao, G., & Helu, M. (2020). Framework for a digital twin in manufacturing: Scope and
requirements. Manufacturing Letters, 24, 105–107.
https://doi.org/10.1016/j.mfglet.2020.04.004
Stacchio, L., Perlino, M., Vagnoni, U., Sasso, F., Scorolli, C., & Marfia, G. (2022). Who
will Trust my Digital Twin? Maybe a Clerk in a Brick and Mortar Fashion Shop.
Proceedings - 2022 IEEE Conference on Virtual Reality and 3D User Interfaces
Abstracts and Workshops, VRW 2022, 814–815.
https://doi.org/10.1109/VRW55335.2022.00258
Sehrawat, Deepti, and Nasib Singh Gill. 2019. “Smart Sensors: Analysis of Different Types
of IoT Sensors.” Proceedings of the International Conference on Trends in
Electronics and Informatics, ICOEI 2019 (Icoei):523–28. doi:
101
10.1109/ICOEI.2019.8862778.
Talkhestani, B. A., Braun, D., Schloegl, W., & Weyrich, M. (2020). Qualitative and
quantitative evaluation of reconfiguring an automation system using digital twin.
Procedia CIRP, 93, 268–273. https://doi.org/10.1016/j.procir.2020.03.014
Tao, F. (2017). PM10 - Digital Twin Shop-Floor: A New Shop-Floor Paradigm Towards
Smart Manufacturing. Robotics and Computer-Integrated Manufacturing, 61, 10.
Tao, F., Cheng, J., Qi, Q., Zhang, M., Zhang, H., & Sui, F. (2018). Digital twin-driven
product design, manufacturing and service with big data. International Journal of
Advanced Manufacturing Technology, 94(9–12), 3563–3576.
https://doi.org/10.1007/s00170-017-0233-1
Udayangani, J., Karunanayaka, I., & Abeysooriya, R. (2019). Industry 4.0 Elements and
Analytics for Garment Assembly Production Lines. MERCon 2019 - Proceedings, 5th
International Multidisciplinary Moratuwa Engineering Research Conference, 745–
750. https://doi.org/10.1109/MERCon.2019.8818882
Vatankhah Barenji, A., Liu, X., Guo, H., & Li, Z. (2021). A digital twin-driven approach
towards smart manufacturing: reduced energy consumption for a robotic cell.
International Journal of Computer Integrated Manufacturing, 34(7–8), 844–859.
https://doi.org/10.1080/0951192X.2020.1775297
Wang, X. V., & Wang, L. (2019). Digital twin-based WEEE recycling, recovery and
remanufacturing in the background of Industry 4.0. International Journal of
Production Research, 57(12), 3892–3902.
https://doi.org/10.1080/00207543.2018.1497819
Wijewardhana, G. E. H., Weerabahu, S. K., Nanayakkara, J. L. D., & Samaranayake, P.
102
(2021). New product development process in apparel industry using Industry 4.0
technologies. International Journal of Productivity and Performance Management,
70(8), 2352–2373. https://doi.org/10.1108/IJPPM-02-2020-0058
Zhang, X., & Zhu, W. (2019). Application framework of digital twin-driven product smart
manufacturing system: A case study of aeroengine blade manufacturing. International
Journal of Advanced Robotic Systems, 16(5), 1–16.
https://doi.org/10.1177/1729881419880663
Zhang, Z., Guan, Z., Gong, Y., Luo, D., & Yue, L. (2022). Improved multi-fidelity
simulation-based optimisation: application in a digital twin shop floor. International
Journal of Production Research, 60(3), 1016–1035.
https://doi.org/10.1080/00207543.2020.1849846
Zheng, X., Lu, J., & Kiritsis, D. (2021). The emergence of cognitive digital twin: vision,
challenges and opportunities. International Journal of Production Research.
https://doi.org/10.1080/00207543.2021.2014591
Zheng, Y., Yang, S., & Cheng, H. (2019). An application framework of digital twin and its
case study. Journal of Ambient Intelligence and Humanized Computing, 10(3), 1141–
1153. https://doi.org/10.1007/s12652-018-0911-3
103
APPENDIX
Table A1: Downtime of rest of the observed days. Part-1
Day-2
OP
No
Operation Name
Cycle
Time
(minu
te)
Response
time
(minute)
Repair
Time
(minute
)
CT
incorporati
ng down
time
(minute)
CT_DT
incorporati
ng down
time
(minute)
OP5
2
Shoulder join
0.94
0.94
0.94
Shoulder join
0.94
18
22
1.02
0.99
OP5
3
Shoulder Excess
trim
0.44
15
26
0.53
0.49
OP5
4
Neck false tack
0.47
0.47
0.47
OP5
5
Sleeve join
0.93
23
31
1.04
0.99
Sleeve join
0.93
0.93
0.93
OP5
6
Body side Excess
trim
0.22
0.22
0.22
OP5
7
Collar join & Collar
mark
0.93
15
38
1.04
1.01
Collar join & Collar
mark
0.93
0.93
0.93
OP5
8
Collar close
0.98
20
25
1.07
1.03
Collar close
0.98
0.98
0.98
OP5
9
Side Seam
0.94
0.94
0.94
Side Seam
0.94
0.94
0.94
OP6
0
Thread Trim
0.61
0.61
0.61
Thread Trim
0.61
0.61
0.61
OP6
1
Bottom excess trim
& Bottom hem
0.98
16
20
1.06
1.02
Bottom excess trim
& Bottom hem
0.98
0.98
0.98
104
Table A1: Downtime of rest of the observed days. Part-2
Day-3
Day-4
Day-5
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
0.94
0.94
15
17
1.01
0.98
0.94
0.94
19
22
1.03
0.99
0.94
0.94
0.94
0.94
0.44
0.44
18
16
0.51
0.47
0.44
0.44
18
24
0.56
0.52
0.47
0.47
17
20
0.55
0.51
0.93
0.93
0.93
0.93
18
24
1.02
0.98
16
37
1.04
1.01
0.93
0.93
23
37
1.06
1.01
17
27
0.31
0.28
0.22
0.22
0.22
0.22
0.93
0.93
22
35
1.05
1.00
0.93
0.93
15
20
1.00
0.97
0.93
0.93
18
19
1.01
0.97
0.98
0.98
0.98
0.98
20
17
1.06
1.02
0.98
0.98
22
26
1.08
1.03
0.98
0.98
23
29
1.05
1.00
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.61
0.61
20
33
0.72
0.68
0.61
0.61
0.61
0.61
0.61
0.61
22
36
0.73
0.69
0.98
0.98
0.98
0.98
0.98
0.98
0.98
0.98
18
38
1.10
1.06
0.98
0.98
105
Table A1: Downtime of rest of the observed days. Part-3
Day-6
Day-7
Day-8
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
0.94
0.94
0.94
0.94
18
29
1.04
1.00
20
22
1.03
0.99
0.94
0.94
0.94
0.94
18
17
0.51
0.48
0.44
0.44
16
28
0.53
0.50
0.47
0.47
15
22
0.55
0.52
0.47
0.47
18
27
1.02
0.99
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
17
23
1.01
0.98
0.22
0.22
15
18
0.29
0.26
0.22
0.22
15
37
1.04
1.01
0.93
0.93
0.93
0.93
0.93
0.93
18
20
1.01
0.97
0.93
0.93
0.98
0.98
17
22
1.06
1.03
0.98
0.98
15
28
1.07
1.04
0.98
0.98
18
25
1.07
1.03
0.94
0.94
17
27
1.03
1.00
0.94
0.94
0.94
0.94
0.94
0.94
19
31
1.04
1.00
25
32
0.73
0.68
0.61
0.61
0.61
0.61
0.61
0.61
0.61
0.61
0.61
0.61
0.98
0.98
0.98
0.98
17
23
1.06
1.03
0.98
0.98
22
38
1.11
1.06
0.98
0.98
106
Table A1: Downtime of rest of the observed days. Part-3
Day-9
Day-10
Day-11
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
0.94
0.94
0.94
0.94
15
20
1.01
0.98
18
22
1.02
0.99
0.94
0.94
0.94
0.94
0.44
0.44
18
15
0.51
0.47
0.44
0.44
0.47
0.47
20
21
0.56
0.51
0.47
0.47
19
27
1.03
0.99
0.93
0.93
18
27
1.02
0.99
0.93
0.93
0.93
0.93
20
27
1.03
0.99
22
23
0.31
0.27
15
29
0.31
0.28
0.22
0.22
17
23
1.01
0.98
0.93
0.93
0.93
0.93
0.93
0.93
18
22
1.01
0.98
0.93
0.93
0.98
0.98
15
25
1.06
1.03
0.98
0.98
0.98
0.98
0.98
0.98
0.98
0.98
16
22
1.02
0.99
0.94
0.94
22
27
1.04
1.00
0.94
0.94
0.94
0.94
0.94
0.94
0.61
0.61
22
37
0.73
0.69
0.61
0.61
23
31
0.72
0.67
0.61
0.61
0.61
0.61
0.98
0.98
0.98
0.98
16
26
1.07
1.03
0.98
0.98
0.98
0.98
15
33
1.08
1.05
107
Table A1: Downtime of rest of the observed days. Part-4
Day-12
Day-13
Day-14
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
0.94
0.94
0.94
0.94
15
22
1.02
0.99
20
22
1.03
0.99
0.94
0.94
0.94
0.94
18
40
0.56
0.52
0.44
0.44
18
19
0.52
0.48
0.47
0.47
25
33
0.59
0.54
0.47
0.47
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
25
27
1.04
0.99
0.22
0.22
16
26
0.31
0.27
0.22
0.22
0.93
0.93
20
32
1.04
1.00
0.93
0.93
16
27
1.02
0.99
0.93
0.93
0.93
0.93
0.98
0.98
0.98
0.98
0.98
0.98
15
32
1.08
1.05
18
32
1.08
1.05
0.98
0.98
0.94
0.94
0.94
0.94
18
22
1.02
0.99
18
28
1.04
1.00
0.94
0.94
0.94
0.94
0.61
0.61
0.61
0.61
17
29
0.71
0.67
22
28
0.71
0.67
0.61
0.61
0.61
0.61
0.98
0.98
23
33
1.10
1.05
0.98
0.98
0.98
0.98
0.98
0.98
15
31
1.08
1.04
108
Table A1: Downtime of rest of the observed days. Part-5
Day-15
Day-16
Day-17
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
0.94
0.94
15
20
1.01
0.98
0.94
0.94
20
23
1.03
0.99
0.94
0.94
0.94
0.94
0.44
0.44
0.44
0.44
18
15
0.51
0.47
0.47
0.47
0.47
0.47
17
21
0.55
0.51
0.93
0.93
15
27
1.02
0.99
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
15
27
0.31
0.28
17
18
0.29
0.26
0.22
0.22
0.93
0.93
0.93
0.93
15
34
1.03
1.00
18
20
1.01
0.97
0.93
0.93
0.93
0.93
22
27
1.08
1.04
0.98
0.98
19
40
1.10
1.06
0.98
0.98
23
28
1.09
1.04
0.98
0.98
0.94
0.94
0.94
0.94
21
27
1.04
1.00
16
38
1.05
1.02
0.94
0.94
0.94
0.94
0.61
0.61
22
23
0.70
0.66
0.61
0.61
0.61
0.61
18
37
0.72
0.69
0.61
0.61
17
19
1.06
1.02
0.98
0.98
0.98
0.98
0.98
0.98
0.98
0.98
19
34
1.09
1.05
109
Table A1: Downtime of rest of the observed days. Part-6
Day-18
Day-19
Day-20
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
0.94
0.94
0.94
0.94
19
22
1.03
0.99
16
22
1.02
0.99
0.94
0.94
0.94
0.94
0.44
0.44
18
15
0.51
0.47
0.44
0.44
0.47
0.47
15
20
0.54
0.51
0.47
0.47
18
28
1.03
0.99
0.93
0.93
18
27
1.02
0.99
21
36
1.05
1.01
0.93
0.93
0.93
0.93
0.22
0.22
16
30
0.32
0.28
19
29
0.32
0.28
0.93
0.93
22
37
1.05
1.01
0.93
0.93
15
22
1.01
0.98
0.93
0.93
0.93
0.93
0.98
0.98
0.98
0.98
16
17
1.05
1.02
0.98
0.98
22
26
1.08
1.03
0.98
0.98
0.94
0.94
0.94
0.94
22
27
1.04
1.00
17
24
1.03
0.99
0.94
0.94
0.94
0.94
0.61
0.61
0.61
0.61
0.61
0.61
0.61
0.61
0.61
0.61
16
38
0.72
0.69
19
33
1.09
1.05
0.98
0.98
0.98
0.98
0.98
0.98
17
36
1.09
1.06
0.98
0.98
110
Table A1: Downtime of rest of the observed days. Part-7
Day-21
Day-22
Day-23
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
0.94
0.94
0.94
0.94
17
25
1.03
0.99
20
22
1.03
0.99
0.94
0.94
18
27
1.03
1.00
0.44
0.44
18
15
0.51
0.47
0.44
0.44
0.47
0.47
15
22
0.55
0.52
0.47
0.47
0.93
0.93
0.93
0.93
0.93
0.93
25
37
1.06
1.01
0.93
0.93
0.93
0.93
0.22
0.22
0.22
0.22
16
28
0.31
0.28
0.93
0.93
0.93
0.93
0.93
0.93
19
26
1.02
0.98
18
22
1.01
0.98
0.93
0.93
0.98
0.98
0.98
0.98
0.98
0.98
0.98
0.98
0.98
0.98
15
29
1.07
1.04
17
37
1.05
1.02
0.94
0.94
0.94
0.94
0.94
0.94
22
33
1.05
1.01
0.94
0.94
15
30
0.70
0.67
0.61
0.61
21
29
0.71
0.67
0.61
0.61
15
23
0.69
0.66
0.61
0.61
0.98
0.98
18
31
1.08
1.04
18
22
1.06
1.03
17
39
1.10
1.06
0.98
0.98
0.98
0.98
111
Table A1: Downtime of rest of the observed days. Part-7
Day-24
Day-25
Day-26
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
0.94
0.94
0.94
0.94
25
26
1.05
0.99
0.94
0.94
0.94
0.94
0.94
0.94
18
23
0.53
0.49
0.44
0.44
18
23
0.53
0.49
16
21
0.55
0.51
0.47
0.47
0.47
0.47
0.93
0.93
18
27
1.02
0.99
0.93
0.93
23
27
1.03
0.99
0.93
0.93
0.93
0.93
0.22
0.22
17
19
0.30
0.26
0.22
0.22
0.93
0.93
23
33
1.05
1.00
0.93
0.93
0.93
0.93
0.93
0.93
17
33
1.03
1.00
18
37
1.09
1.06
15
27
1.07
1.04
0.98
0.98
0.98
0.98
0.98
0.98
0.98
0.98
0.94
0.94
0.94
0.94
17
29
1.04
1.00
19
32
1.05
1.01
0.94
0.94
0.94
0.94
0.61
0.61
20
31
0.72
0.67
15
26
0.70
0.66
16
28
0.70
0.67
0.61
0.61
0.61
0.61
0.98
0.98
0.98
0.98
16
37
1.09
1.06
0.98
0.98
15
29
1.07
1.04
0.98
0.98
112
Table A1: Downtime of rest of the observed days. Part-8
Day-27
Day-28
Day-29
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
Res
pon
se
tim
e
(mi
nut
e)
Re
pai
r
Ti
me
(mi
nut
e)
CT
incor
porati
ng
down
time
(minu
te)
CT_
DT
incor
porati
ng
down
time
(minu
te)
0.94
0.94
0.94
0.94
17
24
1.03
0.99
19
23
1.03
0.99
0.94
0.94
0.94
0.94
0.44
0.44
16
19
0.51
0.48
0.44
0.44
18
20
0.55
0.51
0.47
0.47
15
21
0.55
0.51
0.93
0.93
17
26
1.02
0.98
18
27
1.02
0.99
22
31
1.04
0.99
0.93
0.93
0.93
0.93
0.22
0.22
15
22
0.30
0.27
0.22
0.22
0.93
0.93
0.93
0.93
20
38
1.05
1.01
0.93
0.93
0.93
0.93
0.93
0.93
15
17
1.05
1.02
0.98
0.98
0.98
0.98
0.98
0.98
23
28
1.09
1.04
0.98
0.98
0.94
0.94
16
39
1.05
1.02
0.94
0.94
16
29
1.03
1.00
0.94
0.94
17
31
1.04
1.00
0.61
0.61
0.61
0.61
0.61
0.61
0.61
0.61
0.61
0.61
19
35
0.72
0.68
0.98
0.98
21
33
1.09
1.05
0.98
0.98
18
38
1.10
1.06
0.98
0.98
0.98
0.98
113
Table A1: Downtime of rest of the observed days. Part-9
Day-30
Respons
e time
(minute)
Repair
Time
(minute
)
CT
incorporatin
g down time
(minute)
CT_DT
incorporatin
g down time
(minute)
0.94
0.94
20
22
1.03
0.99
18
15
0.51
0.47
0.47
0.47
17
27
1.02
0.99
0.93
0.93
0.22
0.22
0.93
0.93
18
20
1.01
0.97
19
29
1.08
1.04
0.98
0.98
0.94
0.94
0.94
0.94
15
31
0.71
0.67
0.61
0.61
0.98
0.98
0.98
0.98
