Int. J.

Production Economics 245 (2022) 108413

Contents lists available at ScienceDirect

International Journal of Production Economics

journal homepage: www.elsevier.com/locate/ijpe

Digital transformation in project-based manufacturing: Developing the

ISA-95 model for vertical integration
Lütfi Apilioğulları
Department of Industrial Engineering, Fenerbahce University, Atatürk Mahallesi, Ataşehir Bulvarı, Metropol, 34758, Ataşehir, İstanbul, Turkey

A R T I C L E I N F O A B S T R A C T

Keywords: The Smart Factory is the intelligent management of all business operations that aims to produce innovative
ISA-95 products in the most efficient and faster way through the cyber-physical system (CPS). CPS integrates the cyber
CPS (IT) and the physical world (OT) through the hierarchical layers of ISA-95 model and forms the backbone of
Smart manufacturing
Industry 4.0. The CPS plays a crucial role in transforming data from interconnected systems, and it is a pre­
Digital transformation
Job shop manufacturing
requisite for the entire system to be able to interact. The ISA-95 model consists of components that can be used to
determine how the interaction between domains vertically within the facility. However, the current ISA-95
Model does not include all the components that need to be integrated in vertical integration of all domains.
The product domain is one of them. All IT and OT components need to be represented in the ISA-95 layers to
achieve an ideal CPS structure. In this study, all components needed to achieve the ideal CPS structure have been
identified in an expanded ISA-95 Model for project-based manufacturers. An extended ISA-95 model is proposed
so that each manufacturing industry can use it in the digital transformation process by adding a product domain
and connectivity layer to the Model. The Model was tested and confirmed, by conducting a case study which was
based on implementing the vertical integration on it. This model helps businesses to develop more effective
digital transformation strategies by defining where they should focus on or from where they should start the
digital transformation processes.

1. Introduction the business domain that located on layer-4 (ERP: Formed by sales,
finance, logistics), manufacturing domains that located on layer-3 (MES:
Businesses that adopt Industry 4.0 technologies to their processes Created by manufacturing, quality, and maintenance systems) and
increase their productivity and become more competitive (Tao et al., physical components that located layer 0-1-2 can be realized through
2019; Elia et al., 2019; Rymaszewska et al., 2017) by using the power of layers (Rossit et al., 2019).
knowledge (Yusuf and Gunasekaran, 2004) obtained from the vertical However, NIST states that production operations are formed by
integration of domains, called CPS (Arm et al., 2018; Tao et al., 2018). In combining business, manufacturing, and product domains (Lu et al.,
Industry 4.0, this communication backbone is CPS (Zhong et al., 2017), 2016; Li, et al., 2016). So, another domain that is not represented in
which integrates the cyber (IT: Information Technology) and the phys­ ISA-95, called the product domain (responsible for designing products),
ical world (OT: Operations Technology) through the layers (Aaldering is associated with these two domains in production operations.
and Song, 2020) that work discretely before. Everything interacts with Manufacturing and business domains manage their processes through
each other on CPS (Alcácer et al., 2019), while there was limited the product domain’s data, such as technical drawings, product docu­
communication between functions, applications, and technologies in the ments, and BoM, especially in project-based manufacturing. In other
industrial age. At the core of digital transformation lies CPS. words, the product domain is an integral part of the manufacturing and
In many studies in the literature, the structure and components of business domains in the production environment. However, the current
CPS have been tried to be explained using the ISA-95 layers shown in ISA-95 model is insufficient to explain the relationship of the product
Fig. 1. ISA-95 is a standard that defines the relationship between en­ domain with other domains because it does not contain the product
terprise and control systems hierarchically. It consists of models and domain. This result causes vertical integration to not be fully achieved
terminology that can be used to determine how the interaction between and requires the ISA-95 model to be extended to include all IT

E-mail address: Lutfi.apiliogullari@fbu.edu.tr.

https://doi.org/10.1016/j.ijpe.2022.108413
Received 7 May 2021; Received in revised form 5 January 2022; Accepted 6 January 2022
Available online 17 January 2022
0925-5273/© 2022 Elsevier B.V. All rights reserved.
L. Apilioğulları International Journal of Production Economics 245 (2022) 108413

the reason for this situation. So, how project-based businesses can start
the digital transformation process based on the ISA-95 model is one of
the issues to be considered at this stage because the current ISA-95
model seems insufficient in project-based enterprises.
Using current technology is now inevitable (Khan et al., 2019). The
need for studies explaining how and where to start the digital trans­
formation for project-based manufacturers has increased. There is not
any study in the literature that examine the digital transformation by
considering three domains together in the scope of ISA-95 framework.
In this study, an expanded ISA-95 model that every sector, especially
project-based enterprises can use, was proposed as illustrated in Fig. 2.
By adding the product domain and sub-connectivity layer to the model,
all the building blocks used in CPS integration are represented in the
extended ISA-95 model. In this way, the conceptual inadequacy of the
current ISA-95 model has been eliminated. Three separate case studies
on the expanded ISA-95 model confirmed the model’s validity and
motivated project-based businesses to enter the digital transformation
process.

2. Conceptual analysis of industrial digital ecosystem

Fig. 1. The ISA-95 model.

Digital transformation-oriented issues continue to attract both in­

components. Because, in studies based on the current model, the CPS dustry and academia’s attention (Frank et al., 2019). This situation
structure cannot be fully modeled due to this deficiency. This situation raises interesting in both academia and industry in digital trans­
also makes it difficult for the project-based manufacturing industry, formation subjects. As a result of new research and applications every
which is already behind the process industry in terms of digitalization, to day, different concepts and theories are put forward under dominoes’
enter digital transformation. influence as its shown in Fig. 3. Ideas such as ‘smart manufacturing
Zippel (2018) states that with ISA-95, most process industry facilities (SM), Smart Factory (SF), Smart Supply Chain (SSC), and digital
achieve a nearly fully automated manufacturing process where assets Ecosystem (DES)’ are some of them (Tao et al., 2019; Elia et al., 2020).
are connected to a central control system, and historians/manufacturing These new concepts expand the literature, resulting in concept confusion
execution systems (MES) are often deployed. He argued that many on the industry side (Frank et al., 2019; Moghaddam et al., 2018). The
companies outside the process industry do not understand the impor­ effect of this confusion is often encountered in job-shop manufacturers
tance of ISA-95 and that ISA-95 model should be used as a reference (Apiliogullari, 2019) that usually use ERP to manage accounting pro­
model to begin a digital transformation, without mentioning the IT side cesses, do not yet have MES in production, and still manage business
of CPS. Similarly, Jiang (2017) defined his CPS modeling on the ISA-95 processes through excel and e-mail (see Fig. 4).
model through business, production segments, and physical compo­ Data is the fundamental concept of Industry 4.0 (Porter and Hep­
nents. But he did not specify in his study the product domain that should pelmann, 2014). Everything we do in the scope of Industry 4.0 depends
be on the IT integration side. In another study Rossit at al., (2018) in­ on data produced, transmitted, stored, and analyzed by smart Systems
terprets the CPS concept differently, arguing that MES and its lower (Zhang et al., 2019). To draw meaningful conclusions from data, smart
layers constitute CPS, without considering ERP. systems must communicate seamlessly on the same backbone (Napo­
A broad consensus exists about the OT side of CPS in literature. The leone et al., 2020; Klaus-Dieter Thoben et al., 2019). CPS enables to
physical components that make up the bottom three layers of ISA-95 exchange of data between different systems in all domains (Napoleone
form the OT. However, there are not very comprehensive studies on et al., 2020) by integration OT and IT.
how IT integration in the CPS structure will be. Although some studies
have examined IT integration within the framework of PDM and PLM, 2.1. Operations Technology (OT)
there are limited studies based on the ISA-95 model. One of the reasons
for this can be shown that digital transformation is mainly seen as OT The integration of lower part of ISA-95, called OT, consists of the
integration or that not all business software of IT side is represented in physical components such as sensor, PLC, SCADA. All manufacturing
the ISA-95 model. operations are handled over smart things with different communication
The current ISA-95 model can be valid for continuous production protocols in the OT segment. The connectivity issue of the things to the
such as glass, iron, or process industries such as chemicals and phar­ backbone, interoperability, storage, and process the data collected from
maceuticals for OT integration. In such companies, the central theme is the OT segment are the main focus points in smart manufacturing. Smart
the constant use of production equipment, ensuring product and process things must have a connection property to communicate. In other words,
reliability (Zippel, 2018). The cost of stopping/running heavy equip­ IoT compatibility is a prerequisite for things to conduct two-way
ment or poor quality in giant, expensive equipment is relatively high. communication (Porter and Heppelmann, 2014) within the backbone
The fearful dream is that giant-sized, expensive equipment stops or network obtained by OT integration.
quality error in production. Maintenance and process control activities
are vital in these enterprises, where unplanned downtime and death are 2.1.1. The connectivity layer
almost equivalent. For this reason, the OT side is more developed than Interoperability is the ability of smart systems or software to ex­
project-based enterprises, and it is not so challenging to obtain OT change and make use of information. At the time of Industry 3.0, man­
integration by existing automation technology. However, there is no ufacturers could achieve a limited connection between Things. The main
intensive use of technology on the manufacturing side in job-shop reason was the interoperability problems of things with different
manufacturing. Already, the level of digitalization in these industries protocols.
is well behind the process industry. Vast and old equipment park that The industrial connectivity layer, which does not exist in the ISA-95
does not interact with each other, lack of MES system, excess model, decimates connectivity problems of things with different
labor-intensive processes such as welding, and assembly can be cited as communication protocols (Aaldering and Song, 2020; Napoleone et al.,

2
L. Apilioğulları International Journal of Production Economics 245 (2022) 108413

Fig. 2. Proposed model for ISA-95 (expanded ISA-95).

Fig. 3. The domino effect.

The connectivity layer allows data to flow from many places with
different formats. These data can be large-sized static data that are not
sensitive to time (data at rest) produced from IT systems (Mourtzis et al.,
2016) or small-sized dynamic data that need to be read and analyzed in
milliseconds (data in motion) built-in OT systems (Ren et al., 2019).
SQL-based relational databases work in row columns and with the
principle of event-based data storage. These databases effectively store
and querying data that does not change very often over time and have a
relationship (Qi et al., 2019). On the other hand, fast-generated dynamic
data has meaning in real-time. There is not always a relational structure
between these data. Continuous and fast production of dynamic data can
cause rapid swelling of databases and performance losses (Mourtzis
et al., 2016). Many researchers who support this idea recommend
no-SQL databases rather than SQL-based databases to store dynamic
data (Ren et, al, 2019). Using a no-SQL database, called Historian, data
analytics are more effective by making the data query process faster.
Different compression techniques achieve that capacity gain at the data
storage stage (Mourtzis et al., 2016).
Data gains value as it is processed (Tao et al., 2019). Data collected in
Fig. 4. OT structure and components.
OT integration converts to meaningful results through big data analytics
algorithms. In this way, businesses can be more agile by making the right
2020). All data collected from various smart things with different decisions quickly (Abburu et al., 2020). Data analytics is conducted in
communication protocols can be converted to OPC|UA standard (Lydon, three steps: (1) Extracting pollution within the dataset, parsing repeti­
2020) with small steps through the connectivity layer without any tive data, eliminating meaningless data, reducing the size of the data,
interface software (Ignition, 2021). The connectivity layer is responsible and consolidating big data to be processed. (2) Understanding the
for collecting, transmitting, storing, and analyzing data generated on the relationship between data, modeling the process by defining the pattern
communication backbone. It contains some tools to perform these structure. (3) Ability to derive meaningful insights from data (Tao et al.,
functions. These are connection ports (OPC|UA, MQTT, REST), data­ 2019; Belhadi et al., 2019).
base, and algorithms. The connectivity layer, also called the IIoT plat­
form (Ray, 2016), plays a crucial role in OT integration (Tamas and 2.1.2. The smart manufacturing
Murar, 2019), and can be local or on top of the cloud. This study’s Manufacturing is converting raw material into semi or finished goods
sub-connectivity layer has been added to the ISA-95 model to make the (Li et al., 2018) using the material, machine, and people within the shop
model clearer. floor (Apiliogullari, 2018). Any workshop can produce any part without

3
L. Apilioğulları International Journal of Production Economics 245 (2022) 108413

business domain supports. Today, many businesses, such as the process system’s most essential tasks are formulating and storing the product
industry, start digital transformation by integrating the lower part of domain data structure in acceptable form for integration purposes
ISA-95 (Zippel, 2018). OT integrating increase their efficiencies by (Kul’ga et al., 2018).
allowing them to manage manufacturing operations such as planning, Integration between PDM and ERP systems allows direct sharing of
maintenance, and quality control issues in an intelligent way (Klaus-­ product and business domain data through an automated process (Pra­
Dieter Thoben et al., 2019). In this context, smart manufacturing’s shanth and Venkataram, 2017). By synchronizing these two domains,
definition can be made as follows. Intelligent management of whole engineering teams can access business-level data from the ERP system
manufacturing operations, including maintenance and quality func­ and purchasing team can get the bill of material (BoM). Similarly,
tions, produces innovative products using smart connected things within integration between PDM and MES allows sharing of documents pro­
the shop floor such as material, equipment, machine, and people duced in the product domain directly with the manufacturing domain. In
through the communication backbone achieved by OT integration as it’s that way, manufacturing teams can access product-related data such as
shown in Fig. 4. technical drawings, NC programs from the PDM system (Enríquez et al.,
2019), and the engineering team to get real-time production status.
2.2. Information Technology Moreover, integration between ERP and MES finally allows sharing
work orders, routing, and production status between domains (Lu et al.,
As stated earlier, production system is formed by combining busi­ 2015). By integrating these two domains, planning teams can access
ness, manufacturing, and product domains according to NIST definition production status in real-time and plan for the future state.
(Lu et al., 2016; Li, et al., 2016). The business domain is on top of all PLM is seen as an evolution of PDM (Enríquez et al., 2015) that adds
domains. It is primarily responsible for managing orders, material the business and manufacturing domains to the product domain to deals
management, logistics, and accounting issues. ERP is the maestro of the with all the interactions between domains involved with the Product at
business domain (Mourtzis et al., 2016). The product domain handles all every stage concept to product retirement (Feng, 2015). PLM is not a
tasks related to the new product. Design, test, validation, and production single program. Many industrial companies are developing platforms
engineering issues are the main topics of this domain. CAx, and product that will not require integration between domains. ERP companies such
data management (PDM) are the main design team programs (Kul’ga as SAP, Oracle, IFS can offer ERP/MES segments together. Similarly,
and Gil’fanov, 2008). The manufacturing domain is responsible for companies such as SolidWorks and Autodesk try to eliminate integration
planning, executing, and controlling all shop floor activities. MES soft­ by offering single platforms treating data, process, and project man­
ware precisely serves this purpose (Khan and Turowski, 2016). To agement in the product domain (PDM Manage, Fusion). Some com­
achieve the desired effect from Industry 4.0, all domains in the system panies are more wishful in developing PLM platforms (Siemens, PTC),
must interact (Napoleone et al., 2020). covering all three segments (business, Product, manufacturing). These
IT integration establishes the communication network that enables studies can provide flexibility by eliminating integration problems.
interaction between three domains. However, this separated system is PLM improves the speed and quality of communication and allows
heterogeneous when they are developed. So, the communication for the engineers to minimize communication losses that take an average of
system is an important issue for IT integration. No layer will eliminate sixty percent of their time. In this way, meaningful information can be
the IT side’s interoperability problem, such as the OT’s connectivity reached quickly, and decisions can be made correctly and timely. For
layer. Interface programs provide required integration. example, recording the actions against a problem during the design
Integration is a connection established between systems to share phase to PLM can help solve a manufacturing process problem. Simi­
designated information. It is possible to transfer information between larly, entering information about a customer’s complaint into PLM can
systems in one or two ways by identifying the areas where information support designing a more functional product in product development
will be shared. Synchronization is the retention of the same data on both processes. The information needed in this structure obtains from pre­
sides. When changes occur on any side, the other side updates its viously recorded data without reproduction. All data in PLM is struc­
database (Feng et al., 2015). tured and categorized to be indexed, searched, and easily manipulated
using SQL transactions (Bruun et al., 2015).
2.3. IT integration with PLM Businesses focus on capturing, organizing, and reusing information
instead of searching for and reproducing in the new era (Prashanth and
Business management is like ship management in stormy weather. Venkataram, 2017). PLM is now settling into the IT integration part at
Decision makers must evaluate large amounts of data for quick and the core of CPS. PLM opens the way to being a learning and agile
accurate decisions (Napoleone et al., 2020). However, the complexity organization.
generated by uncertainty makes the decision-making process difficult In this context, IT integration’s definition can be made as follows. IT
(Chen Xin-chun et al., 2021). This situation causes business and per­ integration is a set of systems that allow data produced on the
formance losses of enterprises (Christopher, M, 2000) and weakens their communication network obtained by integrating business, Product, and
competitiveness (Khan and Turowski, 2016). The speed-up decision-­ manufacturing segments to be captured and stored in an organized
making process has always been a significant issue for businesses manner to be reused for different purposes throughout the product life
(Agarwal and Shankar, 2013). Many studies have highlighted agility’s cycle. The innovative PLM model obtained by combining IT Technolo­
importance (Christopher, 2000; Yusuf and Gunasekaran, 2004), gies is given in Fig. 5.
combining flexibility and speed to improve decision-making process
efficiency (Moron and Haan, 2010). 2.4. OT & IT integration
The need for agility increases the importance of Information Man­
agement Systems (IMS). In this context, new methods and systems that CPS integrates the cyber (IT) and the physical world (OT) through
will enable integrating all three domains (Product, Manufacturing, the connectivity layer and forms the backbone of Industry 4.0 (Arm
Business) have been developing. Product data management (PDM) and et al., 2018; Tao et al., 2018). CPS plays a crucial role in transforming
Product lifecycle management (PLM) lead these systems. data from interconnected systems and is a prerequisite for the entire
PDM is a system mainly used in the product domain for managing system to interact (Lee et al., 2016). At the core of digital transformation
design data and engineering processes in one central location (Kul’ga lies CPS, as outlined in Fig. 6.
and Gil’fanov, 2008). All the data on products created in the product In this context, a smart factory’s definition can be made as follows.
domain are stored and managed in the PDM system. With a single source Smart Factory is the Intelligent management of all business operations,
for product data, engineers save time and avoid mistakes. Thus, the PDM including all domains, to produce innovative products in the most

4
L. Apilioğulları International Journal of Production Economics 245 (2022) 108413

digital ecosystem, data is received from anywhere (Pereira and Romero,

2017), the received data is associated with each other, and more accu­
rate decisions can be taken quickly Cimini et al., (2017), Sailesh et al.,
2020. Big data-driven supply chain management (BDD-SCM), where
integration and adaptation are the main principles, is becoming the
norm for manufacturers (Elia et al., 2020). Many researchers, such as
Gunasekaran (2017), Tao et al. (2018), Örjan Larsson and Wallin, 2020,
argue that big data-driven supply chain management positively impacts
business operations and the customer relationship process.
In this context, a smart supply chain’s definition can be made as
follows. Smart Supply Chain is the Intelligent management of end-to-end
supply chain operations by integrating customer and supplier into fac­
tory operations to produce innovative products in the most efficient and
fast way through the horizontal integration obtained by cloud connec­
tion, as shown Fig. 7.

2.6. Proposed ISA-95 model

The expanded ISA-95 model presented in Fig. 8 consists of six layers

and one sub-layer.
OT segment indicates the Physical components found in the bottom
three layers (Level 0-1-2). The connectivity layer, which built OT inte­
gration and allowed OT/IT integration, has been added to the model as a
sub-layer. IT segment indicates the Cyber components found at the
model’s top layers (Level 3–4). The ISA-95 model expanded by inserting
the product domain on the same layer as the manufacturing domain. The
Cloud is positioned at the top of the model (Level 6), allowing horizontal
integration.
Fig. 5. IT structure and components.
OT and IT network integration vertically (CPS) are essential for
Smart Factory. CPS and Cloud integration is an essential requirement for
Smart Supply Chain. Expanding the CPS network horizontally over
Cloud allows CPS to be part of the Digital ecosystem.
The differences of the model from the current ISA-95 are as follows in
Fig. 8.

3. Case studies

The validity of the proposed model was tested by case study. Three
manufacturers that have not yet started digitalization has been selected.
Digitization studies were carried out in two stages as shown Fig. 9, in a
project-based manufacturing enterprise.
The current situation analysis was carried out through observations,
meetings, discussions, and literature research with business directors.
After this phase, which lasted four months, the actions to be taken were
determined. The application was carried out over 12 months using the
PDCA (Deming cycle) method based on planning, implementation,
Fig. 6. CPS structure (OT & it integration). control of results, and standardization of earnings.

efficient and fast way through the CPS backbone obtained by OT & IT 3.1. The profile and previous state of selected factory
integration.
Manufacturer Ma produces pressure vessels of 30–150 m3 for LPG/
LNG sector. Customer requests vary each time. For this reason, 60
2.5. Digital eco-system percent of orders are products covered by innovative products. Manu­
facturer Mb produces tailor made industrial furnaces of different size for
IoT, CPS, and cloud technologies’ development allow communica­ its customers. Mc produces hydraulic power units (HPU) for machine
tion between intelligent systems and data centers to be more reliable industry for different shape and size. The common properties of selected
(Lee et al., 2016). BDA technologies placed on reliable communication job-shop manufacturers are listed in Table 1.
are reshaping the supply chain (Tao et al., 2018; Moghaddam et al.,
2018). • There is no intensive use of technology on the production side. Vast
CPS allows communicating with other intelligent systems horizon­ and old equipment park that does not interact with each other, lack
tally through the Cloud (Örjan Larsson and Wallin, 2020). It can connect of MES system, excess labor-intensive processes such as welding, and
to supplier systems for concurrent engineering activities, customer sys­ assembly can be cited as the reason for this situation.
tems to monitor consumer behavior and product performance, or • Work orders and product documents are given to the shop floor using
external systems that report traffic and weather through the cloud. manual methods. The production process takes months, and real-
Systems’ connection ability to other systems expands the network by time information cannot be obtained from the processes. This situ­
forming a digital ecosystem (Schumacher et al., 2016). Thanks to the ation makes it challenging to track production.

5
L. Apilioğulları International Journal of Production Economics 245 (2022) 108413

Fig. 7. CPS and digital ecosystem.

Fig. 8. The current and proposed expanded ISA-95 model.

• On the IT side, processes are managed through different software,

among which there is no integration. Functions such as engineering,
production, sales, purchasing work in the form of silos. Product data
is stored in shared folders on the design engineers’ local computer or
the server. Cross-departmental information sharing is provided
through intermediate solutions such as e-mail or decommissioning
files from a shared folder.
• Problems in production and delays in delivery occur due to losses
such as working through false documents, late arrival of information,
searching for or reproducing information. This situation, along with
inefficiency, causes some orders to be unable to be received.
Fig. 9. The road map of study.

6
L. Apilioğulları International Journal of Production Economics 245 (2022) 108413

Table 1 The transfer of product data (CAD data) to PDM was considered a
Selected Manufacturer’s assessment. priority for IT integration. Instead of getting any PDM program, the
Segment Layer Domain Ma Mb Mb (Ind. Solid works PDM Manage (SW_PDM) program was selected due some
(Vessel) (HPU) Furnace) reasons as follows. First, SW_PDM contains many functions needed in it
Cloud Layer Cloud Domain No Cloud and it includes PDM, CRM, advanced engineering design programs,
5 project management, and workflow management modules that enable
IT Layer Business Domain ERP system exist but its mainly used the automatic conditional workflows progression. Second, they used
4 for accounting and purchasing. No SW_PDM-based CAD software in their mechanical design processes. This
integration to any system.
Layer Product Domain Design process are managed diffrent
was a significant advantage. Because it would not require the integration
3 SW, among which there is no of mechanical design data. On the other hand, the Solid Works (SW)
integration. platform also had hydraulic and electrical design software. Replacing
Manaufacturing WO and product documents are given electrical and hydraulic design programs at selected enterprises with
Domain to the shop floor by manuel
similar software available within SW would eliminate the integration
operations. There ise no MES.
OT Layer Scada Level There is no intensive use of problem necessary for these areas. So, in Ma and Mc manufacturers,
2 Technology. electrical and hydraulic design software was replaced by their coun­
Layer PLC Level terparts on the SW platform. In Mb, only the hydraulic design program
1 was changed. Because the electrical design program was not requested
Layer Sensor Level
0
to be changed due to the insufficient competence of the SW electrical
design program and the fact that the design engineers were used to the
existing program. Therefore, only the integration of Mb’s electrical
3.2. Action taken design program into SWPDM was performed.

The process work was carried out by dividing into four main phases: (3) PDM Manage integration fills an essential gap in engineering
(1) Determining the priority and area to focus on, (2) Consolidating all processes and data management. However, PDM manage does
programs and replacing them with alternatives on Solid PDM Manage not affect Material Management (purchasing and inventory
(PDM Manage is a program), (3) Integrating the PDM Manage system management). An integration between Manage and ERP is
into ERP, (4) Integrating the PDM Manage system into production needed to automatically send the BoM to ERP or access infor­
processes. mation such as inventory status/material cost during the design
phase. An interface program has been designed for SW_PDM ERP
(1) Availability, efficiency, and quality factors are essential for integration in the SW platform. Using this interface, simultaneous
project-based manufacturers. On the other hand, rapid product transfer of BoM data to ERP and access to inventory data via ERP
development/fast production and agility are the priority in this was achieved.
sector (Cimini et, al. 2017; Nagy et al., 2018) because order (4) Product files are the lifeblood of manufacturing organizations.
winning depends on speed (Putnik and Putnik, 2019; Apilio­ So, managing these files effectively and controlling, who can
gullari, 2019). If the project/manufacturing time takes months, modify them, how they are shared, and how data is moved should
reducing the time by 30–40 percent can bring quite significant be a company’s number one priority. PDM lets the company
advantages in competition. Fast delivery can affect both winning conduct all these tasks along the product life cycle through se­
orders in the market (Kuruppali, 2018) and pushing up its price. curity access controlled in a distributed environment. It allows
So much so that customers may agree to pay higher costs for early the version management, classification, searches, analysis, and
delivery in some cases. reports.

The entire process within all selected manufacturers is managed PDM focus on the product domain. Management of engineering data
according to the data of the engineering department. Timely sharing of is the main subject of PDM. On the other hand, PDM Manage has
the data generated here with the relevant departments directly affects additional properties such as process workflow management. Workflow
business performance. For example, the correct and timely sharing of the represents a business process’s automation; information or tasks are
BoM with the purchasing department shortens the procurement period, passed from one participant to another for action, according to a set of
or the correct and timely transfer of changes in product characteristics to procedural rules. In that way, PDM manage can be seen as an evolution
the production department affects both performance and quality in­ of PDM.
dicators. The main reason for the quality and performance losses in all In this study, product and production domains were integrated
three enterprises were problems with the transfer of information. Since through the SW_PDM Manage workflow feature, allowing production
the BoM files did not arrive on time in Ma and Mb, delays occurred due cells to access business plans and related product documents via PDM
to material waiting. In Mc, the difficulties experienced in accessing the Manage. In that way, machining, welding, assembly manufacturing cells
production documents caused quality errors in production. So, the heart became part of the PDM Manage system.
of the business is the product data produced in product domain. For that The model applied within the study’s scope is as shown in Fig. 10.
reason, we decided to start primarily with IT/Product domain side to
achieve the CPS structure. 4. Result and discussion

(2) Software used for three different purposes during the product Business performance depends on the speed and quality of data
development periods of selected enterprises: electrical design, (Yusuf and Gunasekaran, 2004). Project-based manufacturers where
hydraulic design, and mechanical design. To carry out production product diversity is high must establish processes with an effective in­
and supply processes, all the product data produced in this soft­ formation management system to ensure agility (Christopher, 2000;
ware is needed. All three enterprises used the Solid Works pro­ Yusuf and Gunasekaran, 2004; Guzzi et al., 2018). Many researchers
gram for mechanical design, while different programs were used have supported high-cost losses in firms that do not have an effective
for electrical and hydraulic designs. The lack of a PDM structure information management System (Kul’ga et al., 2018; Tao et al., 20. The
also leads to the fact that electrical, hydraulic, and mechanical main reason for the loss of time caused by factors such as working
design data are stored in one place in a consolidated manner. through an incorrect or outdated document, late arrival of information,

7
L. Apilioğulları International Journal of Production Economics 245 (2022) 108413

Fig. 10. Applied model to ‘manufacturer A’.

are at an advanced level. However, the integration of CPS has been

started from the IT/product domain side in this study.
The followings can cite the reason for this as:

⁃ Business performance is dependent on product development per­

formance in project-based manufacturers. So, improvement in the
product segment affects all domains’ performance positively.
⁃ Agility is a priority for the project-based business. The adequate
knowledge management system required for agility is achieved by IT
integration.
⁃ Excess of low-tech equipment park and labor-intensive processes in
the production segment.

As stated above this approach can differ in continuous production or

process industries. According to OT competence, where to start CPS
integration can be considered a separate study topic.
Result 3. The use of the PDM Manage platform has provided several
advantages. These can be explained as follows.

⁃ PDM Manage eliminated complexity and reduced number of soft­

ware. Consolidation of engineering software and single suppliers
eliminated the need for integration.
⁃ Selecting a single suppliers program reduced the training cost
required for design engineers. Everyone knows the same design
program has contributed to the standard and flexibility.

Fig. 11. Result of actions.

Result 4. One of the main problems in project-based enterprises is the
disconnect between production and planning functions (Fang et al.,
search – waiting - reproduction of information is that the standard 2019). Lack of integration between functions and the managing process
approach to information management is not applied (Enríquez et al., by manual methods causes complexity (Lange and Werner, 2019). In the
2015). manufacturing industry, work orders are produced in ERP and trans­
Result 1. The gains obtained from the study are shown in Fig. 11. The ferred to production via MES (Ramya et al., 2019). However, de­
total duration of engineering design and production times has been ficiencies in ERP and lack of MES prevented this situation in the working
reduced by 20–25 percent in operational terms in all selected manu­ enterprise.
facturers (see Fig. 12). In this study, PDM Manage used an unusual approach for commu­
nication networks as it’s shown in Fig. 12. Work steps in PDM Manage/
Result 2. There are many ways to start digital Transformation. In Project Management modules were used as a work order, and the PDM
many studies in literature, digitalization is started by OT side. The main Manage viewer license was given to manufacturing cells. PDM Manage
reason for this is that industrial automation technologies on the OT side viewer license allowed manufacturing cells to access work orders and

8
L. Apilioğulları International Journal of Production Economics 245 (2022) 108413

Fig. 12. Vertical integration over PDM manage.

documents without MES and ERP, and the PDM Manage module to domain, which was confirmed in this study. In addition, other
receive production status in real-time. PDM Manage established the production segments can determine their digital transformation
primary communication network between product and manufacturing strategies through this model by determining their priority areas.
domains. PDM ERP integration allowed transferring BoM files generated Because all three domains that form the basis of the production
product domain to business domain automatically. business have been defined in the model.
Even if this integration allows information sharing between product (3) Technology brings new alternative solutions. Technological devel­
and manufacturing domains, MES’s functionality cannot achieve. opment reduces the need for integration, allowing alternative
Because the information shared is limited. Work order, documents, and solutions. This study confirmed that PDM Manage could achieve
production status are shared mutually. However, MES performs more limited vertical integration within the enterprise without MES.
than this, such as planning, monitoring, and execution functions. Adding Besides, such flexibility brought about by technology provides
a small MES module to PDM Manage, basic operations such as sending positive motivation for enterprises to enter the digital trans­
work orders, receiving status information from the manufacturing formation process. However, in this study, a small number of
domain, identifying employees can improve limited functionality. The design software was considered in the selected companies, and
project-based manufacturers with intensive labor processes does not almost all programs had their equivalent on the SW platform,
need full functional MES. In this way, product and production segments even if they were partial. However, different software can be used
can operate on a single platform without integration. This proposal may for different purposes in the design processes of some enterprises.
be one of the issues that companies developing engineering software In this case, even if a suitable PDM program is selected, it is
should consider. sometimes not possible to integrate each software into the PDM
since support cannot be obtained from the company that pro­
Result 5. In this study, the selected manufacturers did not prefer the
duces the software. In this case, full integration cannot be ach­
MES System. The main reason is that for MES to receive data from PDM,
ieved, especially on the product segment side. However, a similar
it is necessary to link PDM to MES. Because removing documents from
problem can be solved through the connectivity layer in pro­
the PDM system is not a desirable situation for information security. In
duction processes. Communication between PLC/SCADA systems
this case, MES needs a PDM Manage viewer license to view documents
that generate data in different protocols can now be done over
through the link. However, the manufacturer selected limited MES
OPC/UA standard without the need for integration. In this way,
functionality obtained by PDM Manage instead of getting full MES
all the components in the OT layer can become interconnected
functions by purchasing MES due to more economical.
with each other. Even if some companies try to host all the
components within the scope of PLM for this purpose, this situ­
5. Conclusion ation may impose restrictions on the user’s choice of different
alternatives or not being sufficient because of not being able to
This study’s results and their contribution to Literature can be stated specialize in all areas. The fact that the needs are different and
as follows. specific does not allow all software to be obtained from a single
PLM supplier.
(1) Expanded ISA-95 model: In this study, an innovative ISA-95 model
that can be applied to every sector has been developed by adding In this context, it will be especially beneficial to conduct future ac­
the connectivity layer and product segment, which are not rep­ ademic research on the development of new standards or platforms that
resented in the classic ISA-95 model, explaining the basic build­ will not require integration for the connection of data generated by
ing blocks that make up the digital eco-system and their design software to PDM or PLM systems.
relationships. This model helps businesses to develop more
effective digital transformation strategies by defining where they References
should focus on or where to start digital transformation
processes. Aaldering, L., Song, C., 2020. Of leaders and laggards - Towards digitalization of the
(2) One size does not fit all: Project-based manufacturers manage the process industries. Technovation. https://doi.org/10.1016/j.
technovation.2020.102211.
entire process according to the engineering department’s data. In
this context, CPS integration should start from the IT/Product

9
L. Apilioğulları International Journal of Production Economics 245 (2022) 108413

Abburu, S., Berre, A., Jacoby, M., Roman, D., Stojanovic, L., Stojanovic, N., 2020. Lu, Y., Riddick, F., Ivezic, N., Sep 2016. The paradigm shift in smart manufacturing
Cognitwin – hybrid and cognitive digital twins for the process industry. In: 2020 system Architecture. In: IFIP International Conference on Advances in Production
IEEE International Conference on Engineering, Technology, and Innovation (ICE/ Management Systems (APMS). Iguassu Falls, Brazil, pp. 767–776.
ITMC). Lydon, B., 2020. Top tech: 75 years of automation milestones: from transistors to the
Agarwal, A., Shankar, R., 2013. Modelling the metrics of lean, agile and Leagile supply brink of Industry 4.0. Internat. Soc. Automation 67 (5), p12–21, 10pp.
chain: an ANP-based approach. Eur. J. Oper. Res. 173, 211–225. Moghaddam, M., Cadavid, M., Kenley, Deshmukh, 2018. Reference architectures for
Alcácer, V., Cruz-Machado, V., 2019. Scanning the Industry 4.0: A Literature Review on smart Manufacturing: a critical review. J. Manuf. Syst. 49, 215–225.
Technologies for Manufacturing Systems, Engineering Science and Technology. Int. Moron, D.K., Haan, J., 2010. Improving supply chain performance to satisfy final
J. https://doi.org/10.1016/j.jestch.2019.01.006. customers: leagile experience of a polish distributor. Int. J. Prod. Econ. 133,
Apiliogullari, L., 2018. An assessment model study for lean and agile (leagile) index by 127–134.
using fuzzy AHP. In: 12th NCMCONFERENCES. Ankara, 11-12 Sep 2018. Mourtzis, D., Vlachou, E., Milas, N., 2016. Industrial big data as a result of IoT adoption
Apiliogullari, L., 2019. Digital Transformation/Smart Factories. System Publishing. in manufacturing. Procedia CIRP 55, 290–295.
İstanbul). Nagy, J., Oláh, K., Erdei, E., Máté, D., Popp, J., 2018. The role and impact of industry 4.0
Arm, C., Zezulka, F., Bradac, Z., Kaczmarczyk, V., Benesi, T., Schroeder, T., 2018. and the internet of things on the business strategy of the value chain—the case of
Implementing industry 4.0 in discrete manufacturing: options and drawbacks. IFAC- Hungary. Sustainability 10, 3491. https://doi.org/10.3390/su10103491.
PapersOnLine 51–6, 473–478. Napoleone, A., Macchi, M., Mozzetti, A., 2020. A review on the characteristics of cyber-
Belhadi, A., Zkik, K., Cherrafi, A., Yusof, S., El fezazi, S., 2019. Understanding big data physical systems for the future smart factories. J. Manuf. Syst. 54, 305–335.
analytics for manufacturing processes: insights from literature review and multiple Örjan Larsson, Ö., Wallin, P., October 2020. Digital transformation in the Swedish
case studies. Comput. Ind. Eng. 137, 106099. process industries: trends, challenges, actions. J. Bus. Chem. Voil 17 (3). https://doi.
Bruun, H., Mortensen, N., Harlou, U., Proschowsky, M., 2015. PLM system support for org/10.17879/60119503185.
modular product development. Comput. Ind. 67, 97–111. Pereira, A., Romero, F., 2017. A review of the meanings and the implications of the
Chen Xin-chun, C., Peng, D., Nai-qing, Y., Meng-xue, B., 2021. Study on discrete industry 4.0. Procedia Manuf. 13, 1206–1214.
manufacturing quality control technology based on big data and pattern recognition. Porter, M., Heppelmann, J., 2014. How Smart Connected Products Are Transforming
Math. Probl Eng. https://doi.org/10.1155/2021/8847094. Article ID 8847094, 10 Competition. HBR.
pages. Prashanth, B.N., Venkataram, R., 2017. Development of modular integration framework
Christopher, M., 2000. The Agile supply chain: competing in volatile markets. Internat. between PLM and ERP systems. Mater. Today Proc. 4, 2269–2278.
Market. Manag. 29 (1), 37–44. Putnik, G., Putnik, Z., 2019. Defining sequential engineering, simultaneous engineering,
Cimini, C., Pinto, R., Cavalieri, S., 2017. The business transformation towards smart concurrent engineering and collaborative engineering: on similarities and
Manufacturing: a literature review about reference models and research agenda. differences. Procedia CIRP 84, 68–75.
IFAC-PapersOnLine 50–1, 14952–14957. Qi, Q., Tao, F., Hu, T., Anwer, N., Liu, A., Wei, Y., Wang, L., Nee, A., 2019. Enabling
Elia, G., Polimeno, G., Solazzo, G., Passiante, G., 2019. A multi-dimension framework for technologies and tools for digital twin. J. Manuf. Syst. https://doi.org/10.1016/j.
value creation through Big Data. Ind. Market. Manag. https://doi.org/10.1016/j. jmsy.2019.10.001.
indmarman.2020.03.015. Ramya, G., Chandrasekaran, M., Shankar, E., 2019. Case study analysis of job shop
Enríquez, J.G., Sánchez-Begines, J.M., Domínguez-Mayo, F.J., García-García, J.A., scheduling and its integration with material requirement planning. Mater. Today
Escalona, M.J., 2019. An approach to characterize and evaluate the quality of Proc. 16 (Part 2), 1034–1042.
product lifecycle management software systems. Comput. Stand. Interfac. 61, 77–88. Ray, P., 2016. A survey of IoT cloud platforms. Future Comput. Inform. J. 1, 35e46.
Fang, Y., Peng, C., Lou, P., Zhou, Z., Hu, J., Yan, J., DECEMBER 2019. Digital-Twin-based Ren, S., Zhang, Y., Liu, Y., Sakao, T., Huisingh, D., Almeida, C., 2019. A comprehensive
job shop scheduling toward smart manufacturing. IEEE Trans. Ind. Inf. 15 (NO. 12). review of big data analytics throughout product lifecycle to support sustainable
Feng, Y., Integration Model Based on the Integration of “CAD/CAPP/PLM/ERP” smart Manufacturing: a framework, challenges and future research directions.
Framework Research, 2015. Integration model based on the integration of “cad/ J. Clean. Prod. 210, 1343e1365.
capp/plm/erp” framework research. Chem. Eng. Trans. 46, 1111–1116. https://doi. Rossit, D.A., Tohmé, F., Frutos, M., 2019. Production planning and scheduling in cyber-
org/10.3303/CET1546186. physical production systems: a review. Int. J. Comput. Integrated Manuf. 32 (4–5),
Frank, A., Dalenogare, L., Ayala, N., 2019. Industry 4.0 technologies: implementation 385–395. https://doi.org/10.1080/0951192X.2019.1605199.
patterns in manufacturing companies. Int. J. Prod. Econ. 210, 15–26. Rymaszewska, A., Helo, P., Gunasekaran, A., 2017. IoT powered servitization of
Ignition, 2021. https://inductiveautomation.com/solutions/iiot. Feb 2021. Manufacturing – an exploratory case study. Int. J. Prod. Econ. 192, 92–105.
Khan, A., Turowski, K. A survey of current challenges in manufacturing industry and Sailesh Abburu, S., Berre, A., Jacoby, M., Roman, M., Stojanovic, L., Stojanovic, N., 2020.
preparation for industry 4.0. Proceedings of the first international scientific COGNITWIN – Hybrid and Cognitive Digital Twins for the Process Industry. IEEE
conference “intelligent information technologies for industry” (IITI’16), Adv. International Conference on Engineering.
Intelligent. Syst. Comput. 450, DOI 10.1007/978-3-319-33609-1_2. Schumacher, A., Erol, S., Sihn, W., 2016. A maturity model for assessing Industry 4.0
Klaus-Dieter Thoben, D., Wiesner, S., Wuest, T., 2019. Industrie 4.0 and smart readiness and maturity of manufacturing enterprises. Procedia CIRP 52, 161–166.
manufacturing – a review of research issues and application examples. Int. J. Autom. Tamas, L., Murar, M., 2019. Smart CPS: vertical integration overview and user story with
Technol. a cobot. Int. J. Comput. Integrated Manuf. 32 (NOS.4–5), 504–521. https://doi.org/
Kul’ga, K., Gil’fanov, R., 2008. Integration of CAD/CAM/PDM/MES and ERP systems. 10.1080/0951192X.2018.1535196.
Russ. Eng. Res. 28 (2), 169–172. Tao, F., Qi, Q., Liu, A., Kusiak, A., 2018. Data-driven smart manufacturing. J. Manuf.
Kuruppali, Z., 2018. Measuring leanness and agility of job shops: a rating scale based on Syst. 48, 157–169.
expert consensus. J. Bus. Manag. Sci. 6 (No. 3), 112–117. Tao, F., Qi, Q., Wang, L., Nee, A., 2019. Digital twins and cyber–physical systems toward
Lange, J., Werner, F., 2019. A permutation-based heuristic method for the blocking job smart manufacturing and industry 4.0: correlation and comparison. Engineering 5,
shop scheduling problem. IFAC-PapersOnLine 52 (Issue 13), 1403–1408. 653–661.
Lee, J., Bagheri, B., Jin, C., 2016. Introduction to cyber-Manufacturing. Manuf. Let. 8, Yusuf, Y., Gunasekaran, A., 2004. Agile supply chain capabilities: determinants of
11–15. competitive objectives. Int. J. Prod. Econ. 159, 379–392.
Li, Q., Tang, Q., Chan, I., Wei, H., Pu, Y., Jiang, Li, Y., Zhou, Y., 2018. Smart Zhang, C., Chen, D., Tao, F., Liu, A., 2019. Data driven smart customization. Procedia
manufacturing standardization: architectures, reference models and standards CIRP 81, 564–569.
framework. Comput. Ind. 101, 91–106. Zhong, R., Xu, X., Klotz, E., Newman, S., 2017. Intelligent Manufacturing in the Context
Lu, Y., Morris, K.C., Frechette, S., 2015. Current standards landscape for smart of Industry 4.0: A Review, Engineering. https://doi.org/10.1016/J.ENG.2017.05.01
manufacturing systems. https://nvlpubs.nist.gov/nistpubs/ir/2016/NIST.IR.8107. 5.
pdf b access 20 of Feb 2021. Zippel, S., 2018. Process Industry 4.0 Transforming the Process Industry with Industry
4.0. INTECH NOVEMBER/DECEMBER. WWW.ISA.ORG.

10