Modeling and simulation for optimization of a mattress production line based on flexsim software at dunlopillo (vietnam) LTD

stay time and input/output quantity of all objects in the Cover Section – Actual simulation model.... stay time and input/output quantity of all objects in the Spring Section – Actual si

MINISTRY OF EDUCATION AND TRAINING

HO CHI MINH CITY UNIVERSITY OF TECHNOLOGY AND EDUCATION

FACULTY FOR HIGH QUALITY TRAINING

INDUSTRIAL MANAGEMENT

MODELING AND SIMULATION FOR OPTIMIZATION OF A MATTRESS PRODUCTION LINE BASED ON FLEXSIM SOFTWARE AT

MINISTRY OF EDUCATION AND TRAINING

HO CHI MINH CITY UNIVERSITY OF TECHNOLOGY AND EDUCATION

FACULTY FOR HIGH QUALITY TRAINING

INSTRUCTOR’S COMMENTS

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

Ho Chi Minh City, day … month … year …

Instructor

MEMBER OF THE THESIS COMMITTEE’S COMMENTS

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

Ho Chi Minh City, day … month … year …

Member of the Thesis committee

ACKNOWLEDGEMENT

The process of completing the graduation thesis is the most important stage in every student's life It is the hardest time for me because I did present the biggest challenge that I’ve never met with full of efforts Graduation thesis gives me an opportunity to review all knowledge and skills in four years of university before starting a career

First of all, I would like to express my sincerity, thanks to teachers of the Faculty for High Quality Training and the Faculty of Economy who always have enthusiastically taught and equipped students with all necessary knowledge and experience

Secondly, I would like to express my gratitude to my instructor Mr Nguyen Danh Ha who gave me all useful guidance and enthusiastic support Th ey are very valuable suggestions not only in the process of making this thesis but also as a stepping stone for

me in the process of studying and setting up a career in the future

Finally, thank you to my family and my friends who are always by my side and encourage me to overcome all my hard times

Because of the time limit, the difficulty in approaching such a new researching direction i.e Simulation and lack of experience, the thesis cannot avoid mistakes and shortages Therefore, I am extremely appreciated to receive comments and recommendation from you to help me to complete this graduation thesis as well as possible

Thank you so much,

Best regards

Ho Chi Minh City, 18 August 2021

Student

DO THUY LINH

LIST OF ACRONYMS

LIST OF TABLES

Table 2 1 The advantages and disadvantages of four basic layout types 13

Table 2 2 Characteristics of a manufacturing system model 19

Table 2 3 The advantages & disadvantages of Simulation 21

Table 3 1 The working schedule 24

Table 3 2 The average daily yield of mattress production system in third quarter, 2020 25

Table 3 3 The Average Production Productivity Rate per day 26

Table 3.4 The demand forecast for 2023 27

Table 3 5 List of machines and processing tables in the mattress production line 28

Table 3 6 Table of processing time of machines and processing tables in the mattress production line 29

Table 3 7 The output of mattress production line - the actual model 42

Table 3 8 Detail state of all objects in the Cover Section – Actual simulation model 44 Table 3 9 Detail Avg stay time and input/output quantity of all objects in the Cover Section – Actual simulation model 45

Table 3 10 Detail state of all objects in the Spring Section – Actual simulation model 47

Table 3 11 Detail Avg stay time and input/output quantity of all objects in the Spring Section – Actual simulation model 49

Table 3 12 Detail state of all objects in the Finishing Section – Actual simulation model 50

Table 3 13 Detail Avg stay time and input/output quantity of all objects in the Finishing Section – Actual simulation model 51

Table 4 1 The output of mattress production line - Scenario1 simulation model 54

Table 4 2 Detail state of all objects in the Cover Section – Scenario1 simulation model 56

Table 4 3 Detail Avg stay time and input/output quantity of all objects in the Cover Section – Scenario1 simulation model 57

Table 4 4 Detail state of all objects in the Spring Section – Sncenario1 simulation model 59 Table 4 5 Detail Avg stay time and input/output quantity of all objects in the Spring

Table 4 6 Detail state of all objects in the Finishing Section – Scenario1 simulation model 61Table 4 7 Detail Avg stay time and input/output quantity of all objects in the

Finishing Section – Scenario1 simulation model 62Table 4 8 The output of mattress production line - Scenario2 simulation model 64Table 4 9 Detail state of all objects in the Cover Section – Scenario2 simulation model 66Table 4 10 Detail Avg stay time and input/output quantity of all objects in the Cover Section – Scenario2 simulation model 67Table 4 11 Detail Avg stay time and input/output quantity of all objects in the Spring Section – Scenario2 simulation model 68Table 4 12 Detail state of all objects in the Finishing Section – Scenario2 simulation model 70Table 4 13 Detail Avg stay time and input/output quantity of all objects in the

Finishing Section – Scenario2 simulation model 71

LIST OF FIGURES

Figure 0 1 Process of the research deployment 3

Figure 1 1 Image of Dunlopillo Vietnam Ltd 5

Figure 1 2 Medium End (ME): DUN AUDREY N) MATT 200X180X25CM VN 6

Figure 1 3 High End (HE): DUN FIRMREST LUXE MATT 205X193X25CM BS 6

Figure 1 4 Very High End (V-HE): DUN CORINA PRE MATT 200X160X35CM (N) 7

Figure 1 5 DUN LATEX WORLD PURE MATT 200X160X20CM 7

Figure 1 6 Factory Organization Chart 8

Figure 2 1 Volume and variety characteristics for each process type 10

Figure 2 2 The volume – variety characteristics influence the manufacturing layout 12 Figure 2 3 The relationship between manufacturing process types and basic layout types 14

Figure 2 4 A logo image of Flexsim software 20

Figure 3 1 Mattress production process 24

Figure 3 2 2D Layout drawing of Spring Section 30

Figure 3 3 2D Layout drawing of Cover Section 31

Figure 3 4 2D Layout drawing of Finishing Section 31

Figure 3 5 3D Logic model of Glue – box station 33

Figure 3 6 Setup of the Glue-box table 4’s properties for 3D Logic model 33

Figure 3 7 Setup of the PU Sheet queue’s properties for 3D Logic model 34

Figure 3 8 Process flow for activities at Glue-box station 35

Figure 3 9 Setup parameters for Process Flow Properties 35

Figure 3 10 3D model of process flow for Glue – box station 36

Figure 3 11 Setup parameters for Source Quilting Top 37

Figure 3 12 Setup parameters for Quilting Top machine 38

Figure 3 13 The complete existing 3D model of mattress production line 38

Figure 3 14 All of process flows for the logic model 40

Figure 3 15 State of all objects in the Cover Section Pie chart – Actual simulation

model 44

Figure 3 16 Average stay time of Semi – Mattress on Queue in the Cover Section Bar chart – Actual simulation model 46

Figure 3 17 State of all objects in the Spring Section Pie chart – Actual simulation model 47

Figure 3 18 Average stay time of Semi – Mattress on Queue in the Spring Section Bar chart – Actual simulation model 48

Figure 3 19 State of all objects in the Finishing Section Pie chart – Actual simulation model 50

Figure 3 20 Average stay time of Semi – Mattress on Queue in the Finishing Section Bar chart – Actual simulation model 51

Figure 4 1 The 3D model design of the Scenario1 54

Figure 4 2 Pie chart of Cover Section – Scenario1 simulation model 55

Figure 4 3 Bar chart of Cover Section – Scenario1 simulation model 56

Figure 4 4 Pie chart of Spring Section– Scenario1 simulation model 58

Figure 4 5 Bar chart of Spring Section– Scenario1 simulation model 59

Figure 4 6 Pie chart of Finishing Section– Scenario1 simulation model 61

Figure 4 7 Bar chart of Finishing Section– Scenario1 simulation model 62

Figure 4 8 The 3D model design of the Scenario2 64

Figure 4 9 Pie chart of Cover Section– Scenario2 simulation model 65

Figure 4 10 Bar chart of Cover Section – Scenario2 simulation model 66

Figure 4 11 Bar chart of Spring Section – Scenario2 simulation model 68

Figure 4 12 Pie chart of Finishing Section– Scenario2 simulation model 69

TABLE OF CONTENTS

INSTRUCTOR’S COMMENTS i

MEMBER OF THE THESIS COMMITTEE’S COMMENTS ii

ACKNOWLEDGEMENT iii

LIST OF ACRONYMS iv

LIST OF TABLES v

LIST OF FIGURES vii

INTRODUCTION 1

1 Reasons for choosing the topic 1

2 Objectives of the study 2

3 Scope and object of the study 2

3.1 Object of the study 2

3.2 Scope of the study 2

4 Methodology of the study 2

5 Structure of the thesis 3

CHAPTER 1: INTRODUCTION OF DUNLOPILLO VIETNAM LTD 5

1.1 History and development of the company 5

1.2 Main products 6

1.3 Manufacturing organization chart 8

CHAPTER 2: LITERATURE REVIEW 9

2.1 Manufacturing process types, basic layout types and the relationship between them 2.1.1 Types of manufacturing processes 9

2.1.2 The basic layout types 11

2.1.3 The relationship between manufacturing process types and basic layout types 13

2.2 Bottleneck 14

2.2.1 Definition 14

2.2.2 Types of bottleneck 14

2.2.3 Bottleneck detection methods 15

2.2.4 Type of data for bottleneck analysis 16

2.3 Modeling and Simulation 17

2.3.1 Modeling and Simulation definition 17

2.3.2 The purpose of Modeling and Simulation 18

2.3.3 Modeling of manufacturing system 18

2.3.4 Simulation tool 19

2.3.4.1 Introduction of Flexsim software 20

2.3.4.2 Advantages and disadvantages of Simulation 21

CHAPTER 3: CURRENT PROBLEMS AND THE IMPLEMENTATION OF MODELING & SIMULATION OF THE ACTUAL MATTRESS PRODUCTION LINE 22

3.1 Mattress production process description 22

3.2 Current problem description 25

3.2.1 Current production perspective 25

3.2.3 Future demand forecast 26

3.3 Data collection and analysis 27

3.4 Modeling the simulation with Flexsim software 29

3.4.1 Some assumptions of simulation model 29

3.4.2 Steps of modeling 30

3.5 Simulation results and analysis 40

3.5.1 Simulation running 40

3.5.2 Results and analysis 41

3.5.2.1 Type of output data for analysis 42

3.5.2.2 Output data analysis 43

CHAPTER 4: PROPOSAL OF NEW DESIGNS FOR CAPACITY AND EFFICIENCY IMPROVEMENT OF THE MATTRESS PRODUCTION LINE 53

4.1 Scenario1 for efficiency improvement of the current mattress production line 53 4.1.1 Design of the Scenario1 53

4.1.2 Result and analysis of the effectiveness of the Scenario1 54

4.2 Scenario2 for capacity improvement of the mattress production line 63

4.2.1 Design of the Scenario2 63

4.2.2 Result and analysis of the effectiveness of the Scenario2 64

CONCLUSION 72

REFERENCES 74

INTRODUCTION

1 Reasons for choosing the topic

According to an article by the Industry and Trade newspaper published on October 22,

2019, a study on the mattress market by Mekong Capital indicated that the Vietnamese mattress market has a size of about 1 billion USD with an estimated need of 4 million new mattresses However, there are still 60% of families that have not used them, so there are still a lot of chances for growth The business world predicts that in the future, the mattress market will expand, because the mattress will be an essential item in every Vietnamese family

Currently, Dunlopillo Vietnam focuses on production for export, so it has not really payed attention on the domestic market However, realizing the potential of the Vietnamese market, in the next two years when the covid19 pandemic will be controlled, the company is going to planning for developing and expanding market share in Vietnam

Additionally, in Indonesia, Dunlopillo is one of two mattress brands which have been being dominant in the mattress market Population growth and migration of consumers from rural to urban areas are creating more mattress demand both residential and non – residential segment (e.g hotel, resort, …) Therefore, based on the forecast of Sales department, the company also needs to deal with this increase

According the company’s expansion plan, to cope with this problem, Modeling and

Simulation has been a profitable method to analyze, assess the current status of the

system and modify for improvement It is able to precisely anticipate the performance

of complicated systems, makes it ideally suitable for system planning (Harrell et al.2012) Application of simulation prevents the company from blindly expanding the scale of production to improve capacity

In this study, Flexsim 2019 which is one kind of simulation software was applied, is a state-of-the-art tool for user to simulate and optimize the production line

2 Objectives of the study

This study aims to:

 Make the mattress production line work more efficiently by:

Detect which machines, stations are working inefficiently

Determine and eliminate the bottleneck in the production line, analyze reasons caused it

 Find out optimal solutions for enhancing production capacity to handle the future forecast demand:

Whether the current fixed-resources are enough to meet the demand, or not Which expanding resources plan is suitable and meet the objectives set

3 Scope and object of the study

3.1 Object of the study

The object of this thesis is the mattress production line at Dunlopillo Vietnam Ltd

3.2 Scope of the study

Scope of mattress type: the company has hundreds of types of mattresses classified by

model and size, so it is extremely difficult to simulate in the software Therefore, three main mattress models represent three product segments, structure and size according to forecast demand chosen:

+ Contract (2mx1m8) – Bonnel Spring

+ Duchess (2mx1m8) – IPS Spring

+ Marilyn (2mx1m8) – CPS Spring

Scope of time: in the next 2 years

Scope of area: the plant of Dunlopillo Vietnam Ltd

Scope of data: collected in 2020

4 Methodology of the study

It combines several methods: Observational and documentary methods to collect data for analysis of the actual production line The existing model is analyzed and imported

into Flexsim 2019 software After running this simulation model, problems in the production line will be figured out by analyzing output data Next, new models will be designed to solve the problems until satisfy the targets and objectives In this study, decision-making bases on simulation result This simulation method plays an important role in providing information to determine every solution To make the methodology more clarify, it is summarized in this Figure below

Figure 0 1 Process of the study deployment

Source: Author’s compilation

5 Structure of the thesis

In the process of learning this thesis, to make it easier to follow and understand, I would like to briefly present all main content below:

Chapter 1: Introduction of Dunlopillo Vietnam Ltd

This chapter introduces information about Dunlopillo Vietnam Ltd such as the process

of formation, development, field of activity, products

Chương 2: Literature review

All theories laying foundation for the study are shown in this part

Chapter 3: Current problems and the implementation of modeling & simulation of the mattress production line

This chapter presents about the production process overview, current perspective, all steps from designing to simulating the actual model in flexsim software and analyzing the results of the simulation

Chương 4: Proposal of new designs for capacity and efficiency improvement of the actual mattress production line

After finding out problems, in turn, improvement Scenarios are designed, simulated and analyzed to find the solutions for all the problems

CHAPTER 1: INTRODUCTION OF DUNLOPILLO

VIETNAM LTD

1.1 History and development of the company

 Name: Dunlopillo Vietnam Limited

 Product: spring and latex Mattress

 Charter capital: 6.508.275 USD

 Address: Vietnam – Singapore Industrial Pare, 35 Huu Nghi Boulevard, Thuan

An Town, Binh Duong Province, Vietnam

establishment (1929) in the UK and more than 24 years of establishing a factory in Vietnam (1997), along with great experience and potential, Dunlopillo become a familiar, reliable mattress brand for families, hotels, resort chains in both Vietnamese market and in the world Products of Dunlopillo Vietnam are mainly exported to foreign markets (Asia – Pacific), such as Singapore, Malaysia, China and other small markets like Korea, Cambodia, Brunei, Myanmar, Hong Kong, Saudi Arabia, etc

1.2 Main products

At the factory in Vietnam, the company mainly produces spring mattresses, and assembles latex mattress All Dunlopillo products range from medium to very high-end segments to meet the needs of all domestic and foreign customers, from personal consumption to hotel service business Here are some illustrations for each mattress line and segment:

Figure 1 4 Very High End (V-HE): DUN CORINA PRE MATT 200X160X35CM (N)

Source: Dublopillo.com.vn

 Latex mattress:

Figure 1 5 DUN LATEX WORLD PURE MATT 200X160X20CM

Source: Dublopillo.com.vn

1.3 Manufacturing organization chart

Figure 1 6 Factory Organization Chart

Source: Document from HR Department

CHAPTER 2: LITERATURE REVIEW

In this study, simulation by Flexsim software is applied for analyzing the manufacturing process Therefore, this chapter 2 will in turn point out the related theoretical bases The first one is the theory of factory layout design Simulation will build the entire production process, layout of machines and workers Therefore, determining the layout design plays an extremely important role In addition, to understand the layout design,

it is necessary to know what factors are related to and affect the layout of the factory such as manufacturing process, production method

Next, this study will focus on analysis to identify bottlenecks that reduce production efficiency, so this chapter will learn about what is the bottleneck? How does it effect on production efficiency? what are methods and measure factors used to find out these problems

And finally, the chapter 2 will introduce the Modeling & Simulation method, the Flexsim simulation software, the use and effectiveness of this method for the study as well

2.1 Manufacturing process types, basic layout types and the relationship between them

2.1.1 Types of manufacturing processes

According to Slack et al (2010), there are five types of manufacturing processes which include Project processes, Jobbing processes, Batch processes, Mass processes and Continuous processes Different operations, even those in the same operation, may adopt different types It depends on attributes of the operation such as the variety and volume

of the product which need to be produced The Figure 2.1 shows the relationship

between the types of manufacturing processes and 2 dimensions of volume and variety

Figure 2 1 Volume and variety characteristics for each process type

Source: Operations management book

All the content about the manufacturing process types is displayed below following to the Operations management book of Slack et al (2010):

 Project processes

Project processes usually handle highly customized products The time needed to complete the product or service is pretty long Therefore, low volume and high variety are characteristics of this project processes Because of high customization, the activities

in production can be ill-defined and uncertain They often change to meet the customer demand

 Jobbing processes

Jobbing processes also deal with high variety and low volumes, just behind project processes In jobbing processes, products have to use the same resources with the different modification for each product requirements It different from project process where each product has its own resources This type of process produce more and usually smaller items than project process and they also have common in low repetition

 Batch processes

Batch processes can be quite similar to jobbing processes, but it cannot produce variously as jobbing processes Each time batch processes more than one product, so size of the batch could be just two, three or more, in which case the batch process would

differ little from jobbing Conversely, if the batches are large, and especially if the products are familiar to the operation, batch processes can be fairly repetitive Because

of this, the batch type of process can be found over a wide range of volume and variety levels

 Mass processes

Mass processes deal with production in high volume and pretty narrow variety It is suitable for the production of the large number of products which are the same or similar, the different variants of its product do not affect the basic process of production Several variants of a product could be produced on a mass process such as an assembly line, but the process itself is unaffected The equipment used at each stage of the process can be designed to handle several different types of components loaded into the assembly equipment

 Continuous processes

Continuous processes operate at highest volume and lowest variety with longer periods

of time compared to mass processes In this processes, products are in an endless flow Continuous processes are often associated with relatively inflexible, capital-intensive technologies with highly predictable flow

2.1.2 The basic layout types

Also according to Slack et al (2010), There are four basic layout types: Fixed-position layout, Functional layout, Cell layout and Product layout All practical layout did derive from them The types of layouts are also the same to types of processes, they have to be suitable to characteristics of the products: the company has low or high variety of products, and low or high volume So that The layout types are also relevant to these two dimensions

Figure 2 2 The volume – variety characteristics influence the manufacturing layout

Source: Operations management book

Base on the Operations management book of Slack et al (2010), the content and concept

of the layout types are presented below:

 Fixed-position layout

Fixed-position layout let the product stationary at fixed place while people, equipment, machinery, material move to it as necessary This layout uses for product which is impossible to move conveniently such as: products are too large (Ship, airplane, …), too delicate e.g surgery, or it could object to be move e.g high class service in hotel or restaurant, everything is on-site service, so customer does not have to move to take anything

 Functional layout

Functional layout which is suitable to the needs and convenience of the functions performed by the transforming resources within the processes In this layout, processes are organized by the type of work (function), similar machine, equipment and processes are grouped together, so the utilization of transforming resources is improved Different products will have different needs and requirements; therefore, they take different routes

to complete

 Cell layout

A cell layout is one where the transformed resources entering the operation are selected (or pre-select themselves) to move to one part of the operation (or cell) in which all the transforming resources, to meet their immediate processing needs, are located After being processed in the cell, the transformed resources may go on to another cell

pre-In effect, cell layout is an attempt to bring some order to the complexity of flow which characterizes functional layout

 Product layout

In product layout, machines, equipment and people are arranged in the way following

to the process of the product or group of related products

The decision as to which layout type can be adopted will be influenced by an

understanding of their relative advantages and disadvantages Table 2.1 shows:

Table 2 1 The advantages and disadvantages of four basic layout types

Source: Operations management book

2.1.3 The relationship between manufacturing process types and basic layout types

Process types present the approaches to the organization of processes and activities Layout has intimate relationship with process types Each layout type will be suited for

some process types, but it not rigid, there are often some overlaps between process types and the layouts We can mix flexibly these types of layouts and processes to meet the

actual needs As Figure 2.3:

Figure 2 3 The relationship between manufacturing process types and basic layout

is sick or traveling No one else is qualified or capable to take on that employee's projects, causing a backlog of work until they return

 Long-term bottleneck: when problems occur continuously in the production system and cause serious consequences For example: the inefficient of a working machine leads to the increase of quantity and waiting time of WIP on queue

2.2.3 Bottleneck detection methods

Ericson (2017) stated that there are four main methods to defined the bottleneck in a system: analytically, discrete event simulation, using the data itself and manually looking at buffer levels:

 The analytical method

This method is based on mathematical calculations giving the bottleneck (Yu and Matta, 2014), by measure some data such as starving or blocking percentages The results of this method can be intuitive, satisfying However, it is very difficult to deploy because

of the sophisticated of the production system (Leporis and Králová, 2010)

 The simulation method

According to Ericson (2017), Simulation models are also popular method which have been used since 1980s When using this method, a model of the system is created into simulation software and the model can run for custom periods of time The main advantage of this method is that it can export pretty detailed data about the production from the software such as average stay time on queues, average queue sizes, utilization

of machines and equipment, statistics of failures and maintenance activities as well Additionally, that it is easy to test new changes, experiment and see how they affect the system However, simulation takes a lot of time to create and perform If there are any wrong assumptions can lead to a different result

 Data itself method

Ericson (2017) stated that this method uses the data which is collected from the manufacturing system Applying data itself method helps user to save time, cost and easily to conduct because of available information of manufacturing system Knowing the bottleneck in real-time allow the industrial engineer to optimize it accordingly

 Manually looking at buffer levels method

One more practical approach is “The bottleneck walk” which was pointed out by Roser

et al (2015) In this method, a person walks around the process flow and look at buffer levels in real-time to determine the bottleneck This is the easiest method to defined the bottleneck in the system compared to data driven, analytical or simulation It helps the engineers or operators to use and detect the problems simply

2.2.4 Type of data for bottleneck analysis

Katana (2019) stated that there are 4 tactics can follow to detect the bottlenecks in the system

 Accumulation

In the process steps, collecting unfinished products on queues and then defined which queue accumulates the most semi-products, so next step will be the bottleneck of the system

 Throughput

Collecting data about the increase of throughput of each workstations because bottlenecks affect the throughput time, so if a machine is prone to bottleneck the throughput will increase, and it’ll affect the overall production rate

 Full Capacity

Another way to analyses bottleneck is using data of machine utilization because each of them always has a limit If there is no bottleneck in the production line, all machines will operate at a similar capacity If it appears, one machine in the process will have to work at a higher capacity to deal with a long queue

 Wait times

Collect data of the average waiting time (AWT) of semi-product on queue is another way to identify the bottleneck in the production line because the station will be the bottleneck if its prior queue has the longest AWT stated by Law and Kelton (1991)

2.3 Modeling and Simulation

2.3.1 Modeling and Simulation definition

Panvan (2020) stated:

Modeling and simulation (M&S) is the use of a physical and logical representation of

a given system to generate data and help determine decisions or make predictions about the system M&S is widely used in the social, physical sciences, engineering, manufacturing and product development, among many other

areas

 Modeling

Maria (1997) said that “Modeling is the process of producing a model; a model is a representation of the construction and working of some system of interest” It is useful for industries including manufacturing system because this model can predict the impact

of changes on the system To get high effectiveness, the model need to be designed approximately to the existing system Simultaneously, it shouldn’t be too complicated

to make sure the model can be applied easily and clearly

This method helps researcher can modify and run experiment easily without too many risks because all of scenarios for the new model of system are deployed and performed

in a simulation software Dooley (2002) pointed out the strengths of Simulation

compared to others that “If other methods answer the questions What happened, and

how, and why? simulation helps answer the question What if?”

2.3.2 The purpose of Modeling and Simulation

There are seven applications of M&S in social science pointed out by Axelrod (1997) including: entertainment, theory discovery, training, education, prediction, proof, and performance It can be applied for many areas: design and analysis of production control systems, production planning, determine productivity in production systems, production processes, material handling systems, logistics operations, manage and determine the policy of stocking and purchasing materials of the inventory system

However, this study is about simulation for mattress manufacturing system, so I only focus on 3 of applications which are prediction, performance and proof

 Prediction: base on the design of existing model, the output data can be easily collected From that, it helps to build new model for improvement and forecast how the future behavior of the new one, so that researchers can change, modify the model until meet their expectations and targets

 Performance: with an appropriately calibrated and validated model, simulation can be used to perform real tasks for an organization, such as diagnosis or decision-making

 Proof: as an ability of prediction mentioned above, simulation can provide visual data for experiments, so it can be used to prove which solution is good and feasible to persuade superiors to deploy

2.3.3 Modeling of manufacturing system

Modeling of manufacturing system requires understanding of the types of manufacturing processes including characteristics, issues, objectives which relate to each type of process There are five types of manufacturing processes as mentioned in the former section 2.2.1 including:

In manufacturing system modeling, there are some of characteristics which need to be

noticed to build up a model For more detail, they are presented in the Table2.2:

Table 2 2 Characteristics of a manufacturing system model

Source: Use of simulation in manufacturing and logistics systems planning by Heilala

(1999)

2.3.4 Simulation tool

There are a lot of commercial simulation tools available These tools can be classified into three groups, according to Heilala (1999)

General-purpose simulation language: The type of simulation tool requires the people

who use this tool to be a proficient programmer, expertise at simulation as well

Simulation front-ends: The simulation front-ends are essentially interface programs

between the user and the simulation language being used

Simulators: The simulators utilize constructs and terminology common to the

manufacturing community, and offer graphical presentation and animation This type of simulation tool is simple and easier for user without adept in program

Currently, some of simulators software can easily find in the market such as: AutoMod, WITNESS, Arena, Promodel, Extend+MFG and Flexsim, and so on

2.3.4.1 Introduction of Flexsim software

As mentioned above, there are a lot of Simulation software in the market, but in this thesis, Flexsim version 2019 is used for this study Flexsim is a discrete-event simulation software package based on Windows system developed by Flexsim Software Products, Inc This powerful software allows the user to build a 3D computer simulation model of a real-life system and run experiments on it It provides natural framework for simulation modeling; conceptual guidance; automated gathering of statistics; graphic symbolism for communication; animation; and increasingly, flexibility to change the model More specifically, with Flexsim, the user can collect output data through charts, diagrams, tables, and data can even be exported to Excel, Word, CVS and other software supporting the study and analysis To improve the productivity of production line, Flexsim is a suitable choice because it can show the whole process with sufficient parts and detail, e.g., conveyor, operator, processor, queue, forklift Besides, experiments can

be deployed on computer quickly, they can be completed within a month, a week, or a few hours depending on the complexity of the model and the skill of the engineer Therefore, users can reduce the time and cost requirements compared with physical testing

Figure 2 4 A logo image of Flexsim software

Source: Flexsim.com

2.3.4.2 Advantages and disadvantages of Simulation

Heilala (1999) stated that there are some advantages and disadvantages of simulation

shown in the Table 2.3 Through the information in the table, we can realize how

effective the simulation is and these are some reasons why Simulation can be a good choice for manufacturing system analysis

Table 2 3 The advantages & disadvantages of Simulation

 Study the behavior of a system

flexibly without building the

real one

 Reduce manufacturing cost and

risk through thorough testing

scenarios by software

 Results are visual in general,

compared to other model

 Find out problems existing in

the system easily

 Simple to perform ``What-If''

CHAPTER 3: CURRENT PROBLEMS AND THE IMPLEMENTATION OF MODELING & SIMULATION

OF THE ACTUAL MATTRESS PRODUCTION LINE

3.1 Mattress production process description

The manufacturing process of the company is batch processing Because of the attributes

of mattress production, there are a lot of designs and size leading to hundreds of mattress types In general, all mattresses have the same process, the differences between them are size, materials, components and the way to make them, so they require highly skilled workers To meet the characteristics of similar basic process and the variety of products, the suitable layout for the manufacturing process of the company is a combination of functional and product layout to take the advantages of both types A combination layout

is possible where an item is being made in different types and sizes The main sequence

of operation follows the product layout, but in each section, machinery is arranged in a functional layout to meet the variety

In the mattress production system of Dunlopillo Vietnam Ltd, the production line is divided into 3 major Sections including Cover, Spring and Finishing The process of making a mattress has a series of steps as follow:

In the Spring Section: the spring frame of the mattress is created The first step in this

section is to produce coils from steel wire to create 3 types of spring mattress i.e Bonnel, CPS and IPS Three types of coil are produced by 3 automatic Coiler machines At the following stage, coils for Bonnel Spring are assembled, bend helical without border wire

by semi-automated machine, coils for CPS Spring are assembled and folded into woven, coils for IPS Spring are cut and assembled as well This stage also includes 3 semi-automated machine Subsequently, assembled springs are arranged and clipped into border wire and stabilizer at 2 processing tables

non-In the Cover Section: parallel to the Spring Section, cover of the mattress is also

produced The cover comprises of 3 parts i.e Top, Bottom and Border To begin with, Cover is quilted by several layers consisting of fabric, PU, fiber and non-woven Depending on the type of mattress, processing time, quantity and type of layers will vary The quilted cover is produced in 2 automated Quilting machines In the next,

quilted Border is sent to 1 cutting machine to cut into accurate size Top and Bottom is cut automatically in the quilting machines After that, Top, Bottom, Border continuously are sent to a series of tailoring machine (including 2 panel hemming machines, 1 border hemming machines, 1 embroidery border machine, 3 panel sewing machines, 5 border sewing machines) to hem and sew (sewing labels, flysheet, rope, ) Finally, all of them are checked by QC workers and transferred to Finishing Section to attach to the spring

In the Finishing Section: this Section comprises of 4 steps as 4 stations: Glue-box,

Tape-edge, QC and Packaging Spring frame from Spring Section is transferred to box station to glue with different layers of PU side, PU sheet, felt, latex on 4 Glue-box processing tables Next, Glued mattress is inserted cover and taped edge manually by 4 operators with 4 Tape-edge machines After this step, the mattress is almost finished, then it is sent to QC station with 2 tables to check all dimension parameters,

glue-characteristic, visual of Top, Bottom, Border following to “The Instruction of finished

mattress quality inspection” from QA-QC department The process finishes with

Packaging stage at a processing table along with 2 workers

Figure 3 1 Mattress structure

Source: Author’s compilation

Figure 3 2 Mattress production process

Source: Author’s compilation

Due to the different characteristic and the capacity imbalances in Spring, Cover and

Finishing Section, working time for each Section is different and it is shown as Table

3.1 below:

Table 3 1 The working schedule Shift 1

06:00am to 02:00pm

Shift 2 02:00pm to 10:00pm

Shift 3 10:00pm to 06:00am

Day shift 08:00am to 04:00pm

3.2 Current problem description

3.2.1 Current production perspective

Currently, production is mainly for export Due to the bulky product size that takes a lot

of space to store and the variety types of products, the major business model is MTO (Make to order) Besides, some big markets such as Malaysia, Singapore, China, Hong Kong have large orders and have the certain number of key products, so the company also combines MTS (Make to stock) as well to meet the needs of these markets In recent years, the purchase order has been relatively stable within the capacity, so production and business activities have been controlled quite well

According to Monthly Production Report (MPR) in 3rd quarter, 2020, the average daily

yield is approximately 138 mattresses More detail is shown in Table 3.2

Table 3 2 The average daily yield of mattress production system in third quarter, 2020

Name Month

Finished Goods (pcs)

Spring Unit (Bonnel, IPS & CPS)

Matt Cover (set)

Source: Monthly production report

In general, the current quantity of finished goods can meet the domestic and foreign orders When the machine has problems, any unexpected things occur that delay the production schedule, or the number of orders increases, the company still handle and meet the delivery schedule by adjusting its internal capability However, the production

is not really efficient, so it will be very hard to deal with unforeseen situations if there

are more orders Looking at the output in the Table 3.2, the problem here is balancing

loss leading to accumulation phenomenon: Finish Section compared to the remainders

Moreover, based on the MPR (Table 3.3), Production productivity rate of Finishing and

Cover Section are very low There are a lot of reasons can cause this problem such as low resource utilization, inappropriate resource allocation leads to the appearance of

bottlenecks, problems in planning, or downtime of machinery, etc This matter is pretty serious which can make wasted resources and high cost Therefore, it need more detailed analysis to define thoroughly what problems are actually happening in the process and what solutions are optimal to solve them

At the DVL, the Production Productivity Rate (PPR) is calculated by formula:

𝑇𝑜𝑡𝑎𝑙 𝑤𝑜𝑟𝑘𝑖𝑛𝑔 ℎ𝑜𝑢𝑟𝑠 =

𝑄𝑡𝑦 𝑜𝑓 𝑢𝑛𝑖𝑡 × 𝑐𝑦𝑐𝑙𝑒𝑡𝑖𝑚𝑒

𝑁𝑜 𝑜𝑓 𝑤𝑜𝑟𝑘𝑒𝑟𝑠 × 6.83The Average Production Productivity Rate per day for three sections are show in the

Table 3.3 below

Table 3 3 The Average Production Productivity Rate per day

Finished Goods (unit) Spring Unit (Bonnel,

PPR (%)

Total working hours (hour)

Total cycle time (hour)

PPR (%)

Total working hours (hour)

Total cycle time (hour)

PPR (%)

162.42 81.92 50.47 140.18 105.01 74.98 109.28 67.37 61.65

Source: Monthly production report

3.2.3 Future demand forecast

As mentioned briefly about the problem of the company in the Introduction The company within the next 2 years when the Covid19 pandemic will be gradually controlled, the company will expand more its business in Vietnam and Indonesia leading

to the demand forecast for the following 2 years will increase as shown in Table 3.4

below

Thus, the existing production capacity needs to be calculated and measured to optimize production efficiency Then compare it with the forecast demand to determine how much the factory needs to produce more accurately? Is the current capacity enough to meet the demand? If not, what solutions do the company need to take to improve?

Table 3.4 The demand forecast for 2023 Unit: pcs

Day

Total Contract Duchess Marilyn

Month Total Contract Duchess Marilyn

Year Total Contract Duchess Marilyn

Quantity 65.400 34.500 16.800 14.100

Source: Sale department

In the following sections, all of the problems posed in Section 3 will be solved in turn

by applying simulations on the flexsim 2019 software platform Simulation execution process, improvements results are also presented in detail

3.3 Data collection and analysis

Collecting data for the simulation model is not an easy task, because the simulation only has meaningful and analytical value when the input data is measured relatively exactly,

so researchers will have to spend a lot of time and efforts Identifying the number of

machines and workers is an integral part of building simulation model, Table 3.5 shows

them Each machine and processing table goes along with 1 worker, except for Combiner IPS, Packaging which are handled by 2 workers

At the present, mattress production system has variable automated, semi-automated machine and manual operation Thus, measuring working time for each kind of machine

is different After one month of collecting historical data, observation, and using stopwatch to measure working time of each step for three type of mattresses All of them

are displayed in Table 3.6 As mentioned in the Introduction, because of the limitation

of this study, 3 model of mattresses are Contract, Duchess, Marilyn representing 3 types

of spring structure (Bonnel, IPS, CPS) and segment (Low-end, high-end and end) chosen for this simulation

middle-Table 3 5 List of machines and processing tables in the mattress production line Unit: pcs

No Type Name of

machine Qty No Type

Name of machine Qty

SPRING SECTION 3 Manual IPS Glue-box 1

1 Automated Bonnel

Source: Maintenance department