Proceedings of the 2017 Winter Simulation Conference

W. K. V. Chan, A. D'Ambrogio, G. Zacharewicz, N. Mustafee, G. Wainer, and E. Page, eds.

INTEGRATING SIMULATION-BASED OPTIMIZATION, LEAN, AND THE CONCEPTS OF

INDUSTRY 4.0

Enrique Ruiz Zúñiga

Matias Urenda Moris
Anna Syberfeldt

Production and Automation Engineering Division

School of Engineering Science, University of Skövde
Högskolevägen, Box 408
Skövde, 541 28, SWEDEN

ABSTRACT

Nowadays, due to the need of innovation and adaptation for the mass production of customized goods,
many industries are struggling to compete with the manufacturing sector emerging in different countries
around the world. The understanding and implementation of different improvement techniques is
necessary in order to take part in the so-called fourth industrial revolution, Industry 4.0. This paper
investigates how two well-known improvement approaches, namely lean and simulation-based
optimization, can be combined with the concepts of Industry 4.0 to improve efficiency and avoid moving
production to other countries. Going through an industrial case study, the paper discusses how such a
combination could be carried out and how the different strengths of the three approaches can be utilized
together. The case study focuses on how the efficiency of a production site can be increased and how
Industry 4.0 can support the improvement of the internal logistics on the shop floor.

1 INTRODUCTION
Simulation is a powerful and widely recognized technique for approaching the current challenges in the
manufacturing industry. Industrialized countries are often struggling to keep the growing manufacturing
sector within their borders. New manufacturing factories are being built from the very beginning to adapt
to the tendency going from mass production to mass customization while increasing profitability
(Comstock, Johansen, and Winroth 2004). Manufacturing will then be highly flexible at the volume and
customization level of production, firmly integrated among costumers, companies and suppliers, and
above all sustainable (Shrouf, Ordieres, and Miragliotta 2014). Hence, mainly due to the incorporation of
emerging countries with lower-cost labor and major tax-benefits for the international industry, those
countries with a well-established industrial sector have seen how major factories have moved abroad.
However, in many cases, the costs of transport, logistics and education are significantly increased and the
expertise of factories with a great deal of industrial tradition is lost (Fang, Gunterberg, and Larsson 2010).
For these reasons, there is a huge effort placed by corporations in different technologies to improve the
production in existing factories in order to avoid the alternatives of shutting down or relocating.
Some of the technologies being the base for improvement approaches are simulation-based
optimization (SBO), lean production and the innovative concept of Industry 4.0. Lean is an improvement
method founded in Japan to improve the manufacturing industry after the Second World War. Industry
4.0 is a promising paradigm based on the emerging technologies of Internet of Things (IoT) and Cyber-
Physical Systems, defined to increase the efficiency in manufacturing (Shrouf et al. 2014). These
concepts together with SBO have a huge potential to improve production in middle-size and large
manufacturers. The implementation of these Industry 4.0 concepts can significantly increase or improve

978-1-5386-3428-8/17/$31.00 ©2017 IEEE 3828

 Ruiz-Zúñiga, Moris, and Syberfeldt

production quality, efficiency, flexibility, and security (Shrouf et al. 2014). In this paper, a literature
review of Industry 4.0 and SBO combined with lean is presented. A case study with SBO and lean to
analyze the feasibility of merging two production lines to increase efficiency is presented. Furthermore,
how Industry 4.0 can complement this optimized scenario is discussed.
The structure of this paper is presented as follows: Following the introduction, a literature review of
SBO and lean in manufacturing is presented in Section 2 and a review of Industry 4.0 and IoT is
presented in Section 3. Section 4 introduces the above-mentioned case study with SBO and lean to merge
two production lines and Section 5 presents an analysis to feed these production lines and the potential
benefits of Industry 4.0 for this implementation. Finally the conclusions are presented in Section 6.

2 SIMULATION-BASED MULTI-OBJECTIVE OPTIMIZATION AND LEAN IN

MANUFACTURING
During the 21st century, simulation has started to be a key technology for supporting and improving many
different kinds of systems. Simulation presents a huge potential for product and manufacturing process
development and improvement (Tempelmeier 2003). Building detailed simulation models of the different
production and logistics systems is the base for a good system understanding and analysis. Nowadays,
there are different techniques for process improvement. All of them have different applications for a great
variety of more specific purposes such as different process improvement and design or feasibility studies.
Some examples of process improvement approaches are linear programming, Markov Chain Analysis,
Discrete-Event Simulation (DES), System Dynamics, Monte Carlo Simulation or Value Stream Mapping
as well as some other lean approaches.
DES is often necessary in order to model and represent the complex and stochastic flows in
production lines and logistics systems (Tempelmeier 2003). With the use of these DES models, system
analysis and comparisons of different parts of the shop floor can be done easily with relatively minor
investment. DES provides the results of specific what-if experiment scenarios. However, in order to
analyze several scenarios, a large amount of modelling time is usually required, and although an
improved scenario can be found, an optimum solution cannot be guaranteed. Since simulation is not an
optimization tool by itself, a step that combines simulation and optimization is needed (Pidd 2012). SBO
can be especially suitable in those cases where what-if scenarios can become a time consuming and
tedious task due to the increasing amount of possible solutions.
Depending on the type of problem to analyze, there are different optimization methods that can be
combined with simulation of which several are presented by Figueira and Almada-Lobo (Figueira and
Almada-Lobo 2014). SBO facilitates the search for trade-offs between several conflicting objectives (Deb
2001). Moreover, if there are multiple objectives to be analyzed at the same time, SBO is the best
approach (Ng, Svensson, and Urenda 2008). These characteristics make this approach suitable for
improving complex manufacturing systems with several conflicting objectives.
On the other hand, during the last decades, lean has been commonly adopted by many manufacturing
industries around the world (Melton 2005). Although there is consensus on the definition, “it is a
philosophy that when implemented reduces the time from customer order to delivery by eliminating
sources of waste in the production flow” (Bhasin and Burcher 2006). The lean concept evolved from the
Toyota Production System after the Second World War in Japan. Based on some research studies on the
huge success of Japanese automobile manufacturers, Womack, Jones, and Roos (1990) originally used the
word “lean” to describe the Japanese manufacturing philosophy with Toyota as the leading company
(Womack, Jones, and Roos 1990).
During the 1980s and 1990s, automobile producers around the world started to realize the power of
this philosophy, benchmarking the production and product development systems of the Japanese car
manufacturer based on the vision of its founder, Taiichi Ohno (Liker 2004). Continuous improvement
started to be the focus of this genuine learning enterprise (Liker 2004). Since then, lean production has
become widely adopted in manufacturing companies. The main principles that led to this revolution of the

3829
 Ruiz-Zúñiga, Moris, and Syberfeldt

Toyota Production System were its long-term philosophy, that the right process will produce the right
results, adding value to the organization by developing your people and partners, and continuously
solving root-problems that drive organizational learning (Liker 2004).
Several authors have reported the use of SBO with lean for the improvement of manufacturing
systems (Uriarte, Moris, Ng, and Oscarsson 2015; Yang, Kuo, Su, and Hou 2015). However, a
methodology based on SBO and lean is not commonly presented in manufacturing (Uriarte et al. 2015).
Furthermore, many articles about how the concepts of Industry 4.0 can support lean and SBO within
manufacturing are absent in the literature.

3 THE INTERNET OF THINGS, FACTORY OF THINGS AND INDUSTRY 4.0

Principally due to manufacturing systems around the world representing a significant growing share of the
global trade, and due to the increasing demand of individualized products and natural resources,
manufacturing is becoming more challenging than ever (Herrmann et al. 2014). The IoT is a technology
based on physical devices embedded with electronics, usually with a wireless internet connection. It is
fueling the fourth industrial revolution as well as some other technological revolutions such us Space 2.0
and Health 2.0 (Sacchi et al. 2015; Van De Belt et al. 2010). The IoT is considered to be the key
technology to support the implementation of the concepts Smart Factory, Factory of Things (FoT),
Factory of the Future, Smart Manufacturing etc., as the extension of the IoT concept to manufacturing
systems (Gilchrist 2016; Zuehlke 2010).
Industry 4.0 is considered to be the next step of the evolution of industry following the integration of
information technology and automation systems in manufacturing (Dumitrescu 2017). This next step aims
at integrating the new technology of IoT to build Cyber-Physical Systems to allow most of the different
actuators present on an industrial shop floor (machines, robots, processors, computers, workers…) to be
connected to their surrounding environment, databases and to the outside world in order to interoperate
and cooperate to achieve individual as well as jointly aggregated goals. This can help to achieve the
increased flexibility levels required in production to adapt the capacity to a nowadays more variable and
customized demand (Askar et al. 2007).
The concept of FoT can cover from the product design stage to the production planning, final
assembly, and shipment of products (Radziwon et al. 2014; Weckesser 2016). The IoT and FoT are
intended to help fill the existing gap between the physical world of industrial systems and its
representation in information systems (Haller, Karnouskos, and Schroth 2009). However, in large
manufacturers’ shop floors, this technology is still far from being completely implemented. One major
consideration of this Industry 4.0 concept, besides the extensive digitalization and monitoring effort
performed, is that still large shop floors are far from being considered as a smart factory, the FoT or the
Factory of the Future (Vermesan and Friess 2016). Generally, there are no machines, robots, or
production cells communicating with each other to achieve an optimized flow that depends on the
demand and individual performance. Instead, in the most successful cases, the machines are usually
connected to a central system where the data is collected, analyzed and presented as performance
information.
The goal is to have a compound system where different production cells, machines, robots, and
devices are interconnected, usually with a wireless connection, to improve and make more efficient
production systems (Shrouf et al. 2014). This common wireless connectivity of many devices at the same
or different locations or even factories would help establish a highly reliable communication
infrastructure where every entity is able to know the state of every machine, device, worker, and product
at every time. Most of the information will be shared at the different stages of creating and manufacturing
a product. In this way, the device will inform itself about its new position, will adapt to the devices
upstream and downstream, and will adapt to its required tasks and production pace. This will allow for
sending or receiving data that can help monitor, control, design, program, maintain, and coordinate,
among other tasks, those devices in cooperation with all the devices connected or related to that system.

3830
 Ruiz-Zúñiga, Moris, and Syberfeldt

Regarding the information management through databases, the main idea is to have most of the
information required by the devices stored in data clouds where remote and distributed processors can
access, work with, and modify these common and shared information systems and databases (Gubbi et al.
2013). This technology can be significantly useful when the amount and complexity of the connected
devices is not easy to manage in traditional systems. The foundation for achieving Industry 4.0 is starting
to be present in many companies with digitalization, monitoring, and integration of production processes.
This base should help the future launch of Industry 4.0 and the implementation of real FoT, Smart
Factories or the Factories of the Future. Meanwhile, many issues still have to be addressed, such as
compatibility of different devices, the concept of “plug and work,” more flexibility of machines and
devices, reliable wireless communication infrastructures and, finally, security and privacy aspects.
In Section 4, an industrial case study with SBO and lean to improve a production system is presented.
In this case study, the important potential to implement the concepts and ideas of IoT and FoT to improve
the production is presented in Section 5, Industry 4.0: A more tangible implementation within internal
logistics.

4 INDUSTRIAL CASE STUDY

In order to study the possibilities of combining SBO, lean, and Industry 4.0, the use of SBO and lean to
analyze the feasibility of merging two production lines in the main shop floor of an industrial partner is
discussed. It has been identified that the combination of SBO, lean and Industry 4.0 concepts potentially
can be useful for achieving and extending system improvement regarding the production lines, internal
logistics, and shop-floor layout. This section is sub-divided into an introduction, the simulation modelling
part, the lean production approach and the results.

4.1 Introduction and Problem Description

The objective in this case study is to develop different concept layouts to obtain a more efficient
production, considering the merging possibilities of two production lines. The industrial partner has an
approximated production of 100,000 units per year with around 300 versions in their main shop floor.
There are around 1,200 employees in the factory, 800 of whom are directly related to the manufacturing
process. There are eight production lines in the main shop floor and the main objectives are to increase
the efficiency of the production lines, to reduce the amount of forklift-truck traffic, and to analyze the
possibilities of freeing space in the shop floor in order to expand production.
The production lines share the available space on the shop floor with different machining areas,
production of different components, and painting and packing stations. Furthermore, large decentralized
storage capacity is dispersed along the shop floor to ensure there is no lack of materials or components at
any time. This distribution creates a dense traffic situation with the risks that this generates. Most of the
accidents in the factory are related to forklift-truck driving and human factors.
The two production lines in question to be analyzed, lines A and B, have been considered the starting
point of a major ambitious and long-term redistribution plan in the factory to meet the production
envisioned for the coming years. The actual state of the shop floor, with plenty of manual processes
combined with new automated cells and endless improvement possibilities, combined with this vision of
the company, made this industrial partner the perfect candidate for the development of this research.

4.2 Simulation Models

The simulation methodology used is based on the simulation steps presented by Banks (Banks et al.
2005). Due to the complexity of the systems to analyze, DES was chosen. After defining the objectives
and the description of the problem, data collection and model tasks followed. The problem formulation
was based on the shop-floor layout space limiting the expansion and development of new long-term plans
as envisioned by the company. The objectives of the project are to develop several proofs for concept

3831
 Ruiz-Zúñiga, Moris, and Syberfeldt

models of the possible alternative solutions to improve two production lines, aiming at the production of
an increased amount of versions and at the merging alternative.
Once the problem and objectives were clearly defined, it was possible to start model
conceptualization and data collection. These steps are the “key steps” to a good model in order to obtain
accurate results. The required data can be collected by using historical records, work measurement
procedures or estimates from subject-matter experts (Freivalds and Niebel 2009). When not available,
however, historical data can be enough if based on the records of similar, previously performed studies
that can serve as an estimate of the required data (Freivalds and Niebel 2009).
In this project, a great amount of available data for the new technologically adapted production cells
existed; however, there was an important lack of data on the more primitive machines and manual
processes. Time studies were required and time standards were applied in different processes of the
production lines to obtain accurate data from the real system. With the help of operators and extra
resources, some sub-projects of data collection were organized to obtain the requested processing times of
all the processes in the production lines. A clear definition of the boundaries (start and finish moments) of
every process had to be made to perform accurate measurements.
Common problems in this data collection phase are the large amount of different product versions and
the amount of non-reliable data. On the one hand, there were different versions with different process
times, sometimes with significant differences. Due to this reason, an analysis of the versions with similar
characteristics was performed in order to group them into bigger families to simplify the data collection
process. Nine version families were defined and the processing times for each of them were collected. On
the other hand, non-reliable data is generated automatically by the system when the data of a product or
process is not introduced correctly by the staff or device, or when some activities are interrupted
Therefore, some boundaries, previously specified by experts and managers, were established in the
different processes and classifications of products.
In parallel, the construction of the conceptual model was performed. The conceptual model of a
system is an accurate representation of itself. It can also be considered as a detail process mapping of the
production lines in this case. It is useful for understanding the system, when establishing the processes to
model for the purpose of the project, and to define the data necessary for the model construction. In this
case, all the processes represented in the conceptual model were revised by going through the system as a
product. In order to explain why and which processes that were rejected from the conceptual model, a list
of assumptions was created. These assumptions contain the information of all the specific cases that were
rejected since they were not considered important for the aim of this project. Once the model
conceptualization and the data collection are finished, it is possible to start with the model translation.
In order to translate the conceptual model into the simulation model, the software tool FACTS
Analyzer was selected mainly due to its DES and optimization capabilities combined with a user-friendly
interface and without additional major programming efforts (Ng, Svensson, and Urenda 2008). Nine
different kinds of product versions in the merged solution of lines A and B were represented in three
optimized scenarios: One scenario with a fishbone concept, mounting some of the components outside the
main flow of materials, one scenario with separated mounting cells, and a third scenario with a common
flow of materials. The three scenarios are optimized to minimize the buffer capacity and amount of
operators while maximizing the throughput. While the first scenario has a main production flow with
some parallel stations outside the main flow to meet the demand of different versions, the second scenario
has an independent production flow and processes for some specific product versions (20% of the total
amount of products). The mentioned 20% of special versions are produced in a separated area and then
added to the main flow of products. The third scenario has a common flow for all the versions, thus, the
working stations have to be more flexible to be able to work with the different product versions.
The original simulation models had to be verified and validated in order to consider that the
simulation represents the system as it is. Verification is a determination of whether the computer
implementation of the conceptual model is correct (Banks 1998). In this case, to perform the verification,

3832
 Ruiz-Zúñiga, Moris, and Syberfeldt

every different family of products and staff of the model was monitored during a few simulations to
ensure they performed their processes and tasks as they should. The number of resources, product
versions, and schedules was also revised. The next step was to validate the model.
A replication analysis and the definition of the warm-up period were implemented to avoid the
variability of the output of the model and to know how accurate the obtained results were. Minor
deviation of the results was noticed running twenty replications on nine days with a warm-up period of
four days of simulation. At this point, the model was verified and validated when the comparison of
results of the real system and simulation model was accurate enough. The validation was mainly based on
the parameters of lead time and throughput of the different main version families of the production lines.

4.3 Lean Production

After analyzing the results of the original system and the three main scenarios, refined simulation models
were built considering some of the lean production concepts that appeared promising for the improvement
of the production lines. These concepts must have a clear flow and effectivity in production, emphasize
the value creation processes, include waste identification and reduction, and have the just-in-time and pull
system approaches. Implementing these concepts based on lean principles to improve the system was the
focus of this application. Figure 1 presents a diagram representing these main principles for continuous
improvement in manufacturing systems.

Figure 1: Toyota Production System (TPS) principles (Liker 2004).

This figure represents the basic pillars the Toyota Production System: standardization, balancing
(Heijunka), Kaizen workshops for continuous improvement, just-in-time production and the Jidoka
concept to find the root causes of problems. All together combined with the expertise of the different
production team members point at the operational excellence of the Toyota Production system.
The three scenarios of the merging alternatives of the production lines were improved by considering
the mentioned lean concepts and principles. Several meetings with the operators, supervisors and
managers were arranged. Kaizen Workshops, and Genchi Genbutsu (get the shop-floor perspective of the
problems and improvements) were continuously performed during the development of the project with
the different production team members.
With an iterative improvement process combining lean and simulation, the results of the different
alternatives considered for merging both production lines were found. The best alternatives were
optimized for different amounts of operators and different processing time variation for the versions of
products in the new production line. Some of the results are presented in the following sub-section.

3833
 Ruiz-Zúñiga, Moris, and Syberfeldt

4.4 Results
The results of this SBO and lean project are a range of optimal alternatives to merge both production
lines. The optimization study measures the throughput against the amount of workforce required and the
variation in processing time of the different merging alternatives. Figure 2 shows the throughput (TP) for
each of the three mentioned alternative scenarios, presented in the first column, combined with the
standardization work in processing time of the different variables (low, medium and high variation in
processing time), presented in the second column. Additionally, the extra workforce variation, is shown in
the third column. These three parameters are presented against the throughput expressed in the Y axis. As
mentioned earlier, the three scenarios considered to maximize the output of the system are: Scenario 1
with a fishbone concept by mounting some of the components outside the main production line, Scenario
2 with separated mounting cells for the different flows of products, and Scenario 3 with a common flow
of products with shared mounting cells.

Figure 2: Throughput results for the three scenarios, processing time variation, and number of operators.

The different processing time variation, Low, Medium, and High represent the variability of the time
needed to perform most of the tasks for the different product versions. Low represents the lowest
variation in processing time, achieved with extensive work with standardization of the processes for the
different product versions; Medium represents an improved reduced-product variation by standardizing
some of the tasks in the production line; and High represents the originally considered processing times
for the different product versions. The throughput of the two original production lines was reached by the
throughput of the new merged designed scenarios with half of the original workforce. A relocation of
these operators to perform other tasks on the shop floor was considered.
Going through the results, different optimized scenarios to perform the merging implementation are
compared and valuable knowledge for managers and stakeholders is presented. It is clear how by
expanding the workforce in this new production line the throughput increases, but also how by adding a
second extra operator it results in a more significant impact than if adding a third one. A significant
impact on throughput by changing the layout of the production lines can also be estimated. In the same
way, it is possible to estimate how the throughput results of working with the processing times of the

3834
 Ruiz-Zúñiga, Moris, and Syberfeldt

different versions affect the system; however, it showed to be less significant than expected by production
managers. The following section presents how the concepts of Industry 4.0 can contribute to the general
improvement of the production and the implementation of the new production line.

5 INDUSTRY 4.0: A MORE TANGIBLE IMPLEMENTATION WITHIN INTERNAL

LOGISTICS
Industry 4.0 and the concepts of IoT and FoT can be implemented in a wide range of manufacturing
industries (it has to be considered that a certain level of automation is required). In this project, in order to
support the new production lines improved with SBO and lean, an analysis of the future state of the shop
floor implementing this technology is presented.
The main shop-floor layout taken into consideration can be divided into machining and assembly,
supported by a complex internal logistics system with several storage areas acting sometimes as buffers
disseminated along the shop-floor layout. There is storage for incoming goods from internal and external
suppliers as well as buffers before and after the different machining centers and production lines. Most of
the required internal logistics are performed by forklift trucks using pallets as containers. Generally there
is a lack of communication between the production lines and the machining areas. Additionally,
information regarding the amount of available places in the different buffers and storage areas is usually
not available. This situation generates a huge amount of traffic in specific parts of the shop floor where
important flows of people and other transports converge. As already mentioned, accidents involving
human errors amplified by the use of forklift trucks are not isolated issues. The following sub-sections are
dedicated to the main areas for improvement analyzed in this research considering Industry 4.0: buffers
and storage, material flows and internal logistics.

5.1 Buffers and Storage

In order to improve the overall system and to find solutions that follow the future-vision plans of the
factory (such us increased production, space restructuration, new production lines for new products etc.),
some studies show how the redistribution of the store capacity on the shop floor can be rearranged and
minimized. Some studies point to the advantage of having a Material Preparation Area (MPA) or central
kitting area located on the shop floor in order to receive and supply material for the machining area and
the production lines. The integration of Industry 4.0 in this implementation has to consider the monitoring
and control of the storage levels, input and output flows of materials, transports, production planning and
demand. Every item going through these storages and buffers should be uniquely identified and this
information shared in real-time with the interconnected devices.
Both possibilities of having the storage areas needed to feed the lines located in a dispersed manner or
having a central storage area have been analyzed. Several meetings with the managing team of the
factory, production managers, and operators were performed when analyzing the potential Industry 4.0
benefits. Lean approaches combined with the knowledge obtained from different simulation models were
analyzed and discussed. Finally, it was demonstrated that a mixed-model solution could fit the necessities
and requirements of internal and external suppliers on the shop floor; a central MPA could free more
space, especially in the production-line area. Additionally, it would reduce drastically the traffic around
the production lines where usually the transit of people is concentrated. However, to fully benefit from the
implementation of Industry 4.0 technology, the internal logistics have to be entirely redesigned to cover
the pick-up and delivery of material to these storage and buffer areas. This is presented in the sub-section
5.3 Internal logistics. Following the analysis of the different flows on the shop floor is summarized in 5.2
Material flows.

3835
 Ruiz-Zúñiga, Moris, and Syberfeldt

5.2 Material Flows

Having as a point of departure a solution with a MPA complemented with disseminated small buffers, as
mentioned before, a priority was to monitor the stock levels and the input and output flows of materials.
The input flows of this MPA coming from internal as well as external suppliers and the foundry of the
factory have to be analyzed, classified, and stored. In some cases, the material should be classified and
reunified to be prepared in kits into standard plastic boxes or pallets. Every product in these flows should
be uniquely identified by RFID or barcode readers and additionally, in order to reduce failures,
augmented reality could be used for complex or tedious tasks of pick-up and classification (Syberfeldt et
al. 2016). The output of this MPA also had to be identified and a just–in–time automated transport system
seemed to be the most suitable solution to cover the transport of boxes and pallets to the different
machining cells and eventually directly to the production lines.
These boxes or pallets are delivered directly to the production lines or to different buffers within close
proximity. The demand for the production lines and available space at the buffers located at the beginning
of every line should be considered in real-time. In order to analyze the different devices that are part of
this process, IoT could be embedded into most of the electronic systems involved in this flow of materials
such as machines, robots, conveyors, RFID or barcode scanners, automated warehouses, and transports
(Gubbi et al. 2013). How these material flows are transported to the different locations to support the new
production lines and the production planning is presented in the following sub-section.

5.3 Internal Logistics

In order to deliver material to the machining areas, MPA, and from there to the production lines, different
transport alternatives have been considered. A mathematical optimization study to compare the
requirements for transporting all the material was performed. In this study, the use of forklift trucks, tow
trains, and AGCs (Automated Guided Carts) were considered. Optimization with mathematical models
showed the amount of AGCs or tow trains needed to substitute the existing forklift trucks. As
optimization parameters, the capacity and amount of tow trains or AGCs, and the minimum safety storage
levels of the production lines were considered as variables. As a result, considering the output of the
optimization, the long-term cost and the reduction of accident risk, the research team led to the AGC
solution, a decision supported by managers and stakeholders. Having this as a starting point, an analysis
was made on how Industry 4.0 with AGCs can support the new designed production lines. A detailed
explanation of the characteristics of the processes and the equipment needed to supply material to the
machining area as well as to the merged production line is presented here.
In the machining and assembly areas, the different processors should have a computer or device
connected to an internal Ethernet or secured internet network. This allows the remote data collection and
cloud storage as well as interaction with the station thus analyzing among others: production planning,
produced parts, input and output flows, setup times, failures, shortages, stock levels, and product
information (Gubbi et al. 2013). The transport should have a computer or equivalent mobile device
connected to the network via Wi-Fi, Bluetooth, WIBREE, ZIGBEE or equivalent wireless technology
which could be used in real-time (Gubbi et al. 2013). Transports should be able to communicate with
automatic doors, machines, robots, operators, and other transports to handle traffic situations in a more
efficient manner (for example avoiding collisions, temporary obstacles in the path, peak times of
operators starting/finishing a shift, going for lunch etc.). In order to be able to adapt to unplanned
dynamic changes and obstacles, the transports should have a simulation aided, knowledge-based routing
algorithm or equivalent technology (Klaas et al. 2011).
The main advantages of AGCs are the high flexibility and efficiency as well as the possibility of
automatic loading and unloading of the material to conveyors. The sensors and motors of these conveyor
systems should be controlled by a PLC (Programmable Logic Controller) with embedded IoT technology
for the communication with the different stations and transports (Gilchrist 2016). Information about the

3836
 Ruiz-Zúñiga, Moris, and Syberfeldt

transports should also be gathered (failures, state, location, items loaded, identified obstacles, speed,
battery load etc.). This information could be accessed in real-time by different computers or mobile
devices with permission to adapt the production pace, establish task priorities, traceability, decisions,
performance monitoring, and routes and tasks definition (Gubbi et al. 2013).
Having all this data as a base of digitalization, production rate and inventory levels can be
significantly improved, adapting to the demand in real-time (Zuehlke 2010). Additionally, with all this
historical data stored, organized, and accessible all the time, the steps of system analysis and data
collection for lean and simulation projects can save a huge amount of time and resources. In this case, the
improved production lines with lean and SBO require the redesign of the internal logistics. As presented,
this redesign, supported by Industry 4.0 concepts, can contribute significantly to an overall production
improvement of the industrial partner. This could be extrapolated to similar manufacturing companies. In
addition, most of the technology specified here is already present in manufacturing processes. However, a
huge effort in standardization, data security and privacy still has to be performed by most of the
manufacturing companies. In the case of the industrial partner of this research project, after several
meetings with the managing team and project leaders, the results of this research project showed the
benefits of this Industrial 4.0 paradigm combined with lean and simulation. Furthermore, the extension of
this project to other production lines of the factory is being considered and the allocation of an Industry
4.0 team and resources to support system improvement with lean and simulation is now a reality.

6 CONCLUSIONS
This paper demonstrates how SBO combined with lean has a huge potential for the system and process
improvement in manufacturing. While SBO can be used to analyze complex dynamic systems with high
variability and a large amount of possible solutions, lean tools can make up a base for continuous
improvement in individual processes (Uriarte et al. 2015). However, an important factor is that there is
often a lack of data in order to start working with lean and simulation approaches, especially in factories
with a long history and tradition. Hence, the integration on the shop floor of Industry 4.0 concepts can be
a solution to this common problem of lack of digital data (Vogel-Heuser and Hess 2016). The first steps
to this Industry 4.0 paradigm, the FoT or the smart factory of the future, include monitoring and
digitalization. Going through this project, it can be appreciated these pillars can to a great extent facilitate
and integrate the use of lean and simulation for system improvement.
Furthermore, as demonstrated in this project, when a major change is required within manufacturing
and the improvement of the system considers production, internal logistics and shop-floor layout, the
combination of SBO and lean with the paradigm of Industry 4.0 can be extremely valuable for managers
and stakeholders. Especially when considering a long term improvement perspective in the production
facilities, the combination of the three approaches can lead to an overall system improvement; beginning
with the base of lean and Industry 4.0, for the local as well as more strategic improvement, both of them
can be complemented with SBO to analyze, materialize, and evaluate the possible improvement of the
system in consideration. Most of the technology required for the Industry 4.0 era is already present in
production. In the same way, it is common that manufacturing companies have lean and simulation teams
and projects; however is less common that they work in a coordinated manner. A solid methodology for
lean and SBO in manufacturing can be very useful and a strong coordination with the emerging paradigm
of Industry 4.0 should be considered in order to increase the efficiency and concurrence of the existing
manufacturing facilities.

ACKNOWLEDGMENTS

The authors want to thank the Swedish Knowledge Foundation (KK Stiftelsen) for research funding, the
industrial partner Xylem Water Solutions for giving us the opportunity to develop this project, and the
IPSI research school of the University of Skövde.

3837
 Ruiz-Zúñiga, Moris, and Syberfeldt

REFERENCES

Askar, G., T. Sillekens, L. Suhl, and J. Zimmermann. 2007. “Flexibility Planning in Automotive Plants”.
In Management logistischer Netzwerke: Entscheidungsunterstützung, Informationssysteme und OR-
Tools, edited by H.-O. Günther, D. C. Mattfeld, and L. Suhl, 235-255. Heidelberg: Physica-Verlag,
Germany.
Banks, J. 1998. Handbook of Simulation. Atlanta, Georgia, USA: John Wiley e sons, Inc.
Banks, J., J. S. Carlson II, B. L. Nelson, and D. M. Nicol. (2005). Discrete-Event System Simulation.
Champaign, Illinois, USA: Pearson Prentice.
Bhasin, S., and P. Burcher. 2006. “Lean Viewed as a Philosophy”. Journal of Manufacturing Technology
Management 17(1):56-72.
Comstock, M., K. Johansen, and M. Winroth. 2004. “From Mass Production to Mass Customization:
Enabling Perspectives from the Swedish Mobile Telephone Industry”. Production Planning &
Control 15(4):362-372.
Dumitrescu, R., K. Amo, and M. Rabe. 2017. Industry 4.0 - Evolution, not Revolution. It's Owl.
http://www.its-owl.com/industry-40/evolution-not-revolution.
Fang, T., C. Gunterberg, and E. Larsson. 2010. “Sourcing in an Increasingly Expensive China: Four
Swedish Cases”. Journal of Business Ethics 97(1):119-138.
Figueira, G., and B. Almada-Lobo. 2014. “Hybrid Simulation–Optimization Methods: A Taxonomy and
Discussion”. Simulation Modelling Practice and Theory 46:118-134.
Freivalds, A., and B. W. Niebel. 2009. Niebel's Methods, Standards, and Work Design. New York, USA:
Mc Graw Hill.
Gilchrist, A. 2016. Smart Factories: Industry 4.0: The Industrial Internet of Things. New York, USA:
Apress.
Gubbi, J., R. Buyya, S. Marusic, and M. Palaniswami. 2013. Internet of Things (IoT): “A vision,
architectural elements, and future directions”. Future Generation Comp Systems 29(7):1645-1660.
Haller, S., S. Karnouskos, and C. Schroth. 2009. “The Internet of Things in an Enterprise Context”. In
Future Internet, edited by J. Domingue, D. Fensel, and P. Traverso, 14-28. Berlin, Heidelberg:
Springer Berlin Heidelberg.
Herrmann, C., C. Schmidt, D. Kurle, S. Blume, and S. Thiede. 2014. “Sustainability in Manufacturing
and Factories of the Future”. International Journal of Precision Engineering and Manufacturing-
Green Technology 1(4):283-292.
Klaas, A., C. Laroque, W. Dangelmaier, and M. Fischer. 2011. “Simulation Aided, Knowledge Based
Routing for AGVs in a Distribution Warehouse”. In Proceedings of the 2011 Winter Simulation
Conference, edited by S. Jain, R. R. Creasey, J. Himmelspach, K. P. White, and M. Fu, 1673-1684.
Piscataway, New Jersey: Institute of Electrical and Electronics Engineers, Inc.
Liker, J. K. 2004. The Toyota Way. New York, USA: Mc Graw Hill.
Melton, T. 2005. “The Benefits of Lean Manufacturing”. Chemical Engineering Research and Design
83(6):662-673.
Ng, A., J. Svensson, and M. Urenda. 2008. “Introducing Simulation-based Optimization for Production
Systems. Design to Industry: the FACTS Game”. In Proceedings of the 18th International
Conference on Flexible Automation and Intelligent Manufacturing, 1359-1372. University of Skövde,
Sweden.
Pidd, M. 2012. “Mixing other Methods with Simulation is no Big Deal”. In Proceedings of the 2012
Winter Simulation Conference, edited by C. Laroque, J. Himmelspach, R. Pasupathy, O. Rose, and
A.M. Uhrmacher, 751-757. Piscataway, New Jersey: Institute of Electrical and Electronics Engineers,
Inc.
Radziwon, A., A. Bilberg, M. Bogers, and E. S. Madsen. 2014. “The Smart Factory: Exploring Adaptive
and Flexible Manufacturing Solutions”. Procedia Engineering 69:1184-1190.

3838
 Ruiz-Zúñiga, Moris, and Syberfeldt

Sacchi, C., K. Bhasin, N. Kadowaki, and F. Vong. 2015. “Toward the "Space 2.0" Era”. IEEE
Communications Magazine 53(3):16-17.
Shrouf, F., J. Ordieres, and G. Miragliotta. 2014. “Smart Factories in Industry 4.0: A review of the
Concept and of Energy Management Approached in Production based on the Internet of Things
Paradigm”. In Proceedings of the 2014 International Conference on Industrial Engineering and
Engineering Management.
Syberfeldt, A., O. Danielsson, M. Holm, and L. Wang. 2016. “Dynamic Operator Instructions Based on
Augmented Reality and Rule-based Expert Systems”. Procedia CIRP 41:346-351.
Tempelmeier, H. 2003. “Practical Considerations in the Optimization of Flow Production Systems”.
International Journal of Production Research 41:149-170.
Uriarte, A. G., M. Urenda, A. H. Ng, and J. Oscarsson. 2015. “Lean, Simulation and Optimization: A
Win-Win Combination”. In Proceedings of the 2015 Winter Simulation Conference, edited by L.
Yilmaz, W. K. V. Chan, I. Moon, T. M. K. Roeder, C. Macal, and M. D. Rossetti, 2227-2238.
Piscataway, New Jersey: Institute of Electrical and Electronics Engineers, Inc.
Van De Belt, T. H., L. J. L. P. G. Engelen, S. A. A. Berben, and L. Schoonhoven. 2010. “Definition of
Health 2.0 and Medicine 2.0: a Systematic Review”. Journal of Med. Internet Research 12(2):1-14.
Weckesser, P. 2016. “Driving the Digital Enterprise with the Digital Enterprise Software Suite”.
http://www.industry.siemens.com/topics/global/en/digital-enterprise-suite/Pages/Default.aspx.
Vermesan, O., and P. Friess. 2016. Digitising the Industry-Internet of Things Connecting the Physical,
Digital and Virtual Worlds. Gistrup, Denmark: River Publishers.
Vogel-Heuser, B., and D. Hess. 2016. “Industry 4.0–Prerequisites and Visions”. IEEE Transactions on
Automation Science and Engineering 13(2):411-413.
Womack, J. P., D. T. Jones, and D. Roos. 1990. Machine that Changed the World. New York, USA: RA
Scribner.
Yang, T., Y. Kuo, C. T. Su, and C. L. Hou. 2015. “Lean Production System Design for Fishing Net
Manufacturing Using Lean Principles and Simulation Optimization”. Journal of Manufacturing
Systems 34:66-73.
Zuehlke, D. 2010. “SmartFactory—Towards a Factory-of-Things”. Annual Reviews in Control 34(1):
129-138.

AUTHOR BIOGRAPHIES

ENRIQUE RUIZ ZÚÑIGA is a PhD candidate at the University of Skövde, Sweden. He holds a B.Eng.
degree in Technical Industrial Engineering from the University of Malaga, Spain, and both a B.Sc. in
Automation Engineering and a M. Sc. Degree in Industrial Informatics from the University of Skövde.
His research interests include system improvement, lean, discrete-event simulation and simulation-based
optimization in healthcare and manufacturing systems. His e-mail address is [email protected].

MATIAS URENDA MORIS is an Assistant Professor at the University of Skövde, Sweden. He holds a
B.Sc. degree in Automation Engineering, a M.Sc. degree in Manufacturing Management and a Ph.D.
degree in Healthcare Engineering from the University of Skövde, Loughborough University, UK, and De
Montfort University, UK, respectively. His main research area is Discrete Event Simulation for
manufacturing and healthcare systems with emphasis on system modeling and analysis. His e-mail
address is [email protected].

ANNA SYBERFELDT is a senior researcher at the University of Skövde, Sweden. She holds a PhD in
Computer Science from the De Montfort University, UK and a Master’s degree in Computer Science
from the University of Skövde, Sweden. Her research interests include soft computing techniques and
simulation-based optimization. Her email address is [email protected].

3839