Professional Documents
Culture Documents
by
Nationality: Myanmar
Previous Degree: B.S.E Industrial Engineering
Asian Institute of Technology
Thailand
I would like to present special thanks to Associate Professor Ir. Erik L.J. Bohez, the chairman
of examination committee for his constant, expert guidance, encouragement and valuable
suggestions from the very beginning of the thesis. I would like to express special thanks to
the exam committee members, Dr. Huynh Trung Luong, Dr. Than Lin and Dr. Mongkol
Ekpanyapong for their suggestions and encouragements. I would like to express my special
gratitue to Mr. Daniel Reinle, Vice President of Siemens Limited, Thailand, who supported
me to use the Tecnomatix Plant Simulation Software and allowed me to have very
supportive technical discussions with specialists and host factory. Im also deeply indebted
to Mr. Ittipol P. (SLT), Miss Napa Hoonpanich(Premium PLM Co. Ltd) and Mr. Chaowat
Kruttin(Premium PLM Co. Ltd) for their kind helps and technical consultations. Moreover,
thanks to all my teachers and lab supervisors for their kind teaching, guidance and help
throughout my time in AIT. Finally, words are never enough to express my deep gratitude to
my family and friends whose love for their inspiration and encouragements to complete my
masters program.
ii
ABSTRACT
In order to survive in aggressively competing markets of current age, companies need to look
for the ways to improve their productivity continuously. It is also critical to ensure the
efficient use of resources such as, Man Power, Machine, Materials, Time and Money to
achieve the favourable outcomes. In order to reach the maximum production rate at minimal
cost, it is essential to have a balanced line configuration. Line balancing is about arranging a
production line so that there is an even flow of production from one work station to the next.
Line balancing also a successful tool to reduce bottleneck by balancing the task time of each
work station so that there is no delays and nobody is overburden with their task.
This thesis presents a case study on an assembly line optimisation problem in an existing
assembly section of a Power Stearing Column production line for Honda Fit. The rationale of
this project is to build the model of current system, to simulate the entire processes in
assembly line, improve the line efficiency and also to recommend improvement activities to
fulfil the fluctuating demand through the analysis done in the simulation model. The
simulation is done by using Tecnomatix Plant Simulation. All the data needed for the line
balancing analysis are collected. This data gathered is then simulated in Tecnomatix Plant
Simulation.
iii
Table of Contents
2 Literature Review 3
2.1 Assembly Line Balancing 3
2.2 Terminology Used in Assembly Line Balancing 3
2.3 Assembly Line Balancing Problems 4
2.4 Aspects of Line Balancing 5
2.5 Discrete Event Simulation 7
2.6 Modeling and Simulation 7
2.7 Tecnomatix Plant Simulation 11
3 Case Studies 17
3.1 Modelling, Simulation and Optimization of Process Planning 17
(Jovisevic, S., Jovisevic, V., &Jokanovic, S., 2009)
3.2 Kanban Simulation Model for Production Process 21
Optimization. (Golchev, R., et.al, 2015)
iv
Chapter Title Page
References 70
Appendix 71
v
List of Figures
vi
List of Figures
Figure Title Page
5.1 Modelling Steps Flow Chart 41
5.2 Station Allocation 42
5.3 Processing Time Configuration Window 42
5.5 Resources Menu Bar 43
5.6 Layout After Workplaces and Footpaths are Added 43
5.7 Creation Table of Workers in Workerpool 44
5.8 Workplace Window 44
5.9 Supported Service Window 45
viii
CHAPTER 1
INTRODUCTION
1.1 Background
In the contemporary, dynamic changing world, access to production data in real life is
necessary to properly plan, simulate and supervise production. Companies operating in
different sectors of the economy are more and more commonly using IT solutions to optimize
manufacturing systems by balancing workflows and process lines.
Technically, line balancing is one of the components of cellular manufacturing which
consist of major tenants in lean manufacturing. The concept of line balancing itself is
everyone is working together in a balance fashioned where everyone doing the same amount
of work, the variation is smoothened, no one is overburden, no one is waiting and the work is
done in a well single piece flow. Line balancing is also can be defined as the allocation of
sequential work activities into a line called work stations in order to achieve best utilization
of labor and equipment thus minimizing idle time. In addition, balancing may be achieved by
rearrangement of the work stations and by equalizing the workload among assemblers so that,
all operations take about the same amount of time. Furthermore, line balancing benefits an
assembly area in many ways, as it minimizes the amount of workers and work station which
can reduce cost and space for the assembly area. Line balancing also benefits in a way that it
can identify the process which causes bottleneck and standardization of work between the
operators can ease the bottleneck problem. Aside from identification of bottleneck, line
balancing equalized the workload among the workers so that there is no worker which is
overburden.
Finally, line balancing helps to assist the plant layout which will lead to the reduction
of production cost by the reduction of worker and the reduction of idle time. Assembly line
grouped in their respective workstations is represented manually in the form of a precedence
diagram.
Commonly, a traditional way of doing assembly line balancing is by using precedence
diagram. A precedence diagram specifies the order or sequence in which the activities must
be performed. Each circle is a node, and the number inside each circle identifies particular
operation. The number outside circle represents duration of operation or the cycle time.
Arrow represents directions of flow of operation. In contrary, simulation model is a new and
effective way to build the real life situation of the assembly process. There are many types of
simulation model that can help not only to identify and reduce bottleneck but also can build
the exact plant layout virtually. The more realistic the simulation model the more accurate and
effective the design for implementation on the assembly area.
Simulation technology is an important tool for planning, implementation, and
operating complex technical system. There are many simulation software created just to build
the virtual layout of the assembly area such as WITNESS and ARENA software.
1
However, the simulation used in this project is the Tecnomatix Plant Simulation
software created by SIEMENS. Tecnomatix Plant Simulation can increase profitability of a
facility by increasing throughput and utilization of resource and facility. Plant Simulation is
also able to decrease lead times, required resources and storage requirement provided that are
all accurate data inserted in the analysis. Furthermore, Plant Simulation is able to identify the
bottlenecks, reduce WIP, evaluate the effects of capital investments or changes in processes
and avoid planning errors as the simulation was done virtually without applying to the
facilities first.
1.2 Problem Statement
After visiting a Tier 1 automobile supplier factory, it was found out that the assembly
section producing Electric Power Steering Columns is working on the Kanban System(Pull
System) to fulfil the fluctuating demand which causes delivery shortages sometimes. At the
same time, types of products produced on it are also changing frequently. The company
would like to evaluate the line layout and settings of the assembly line for each product type
whether these are resulting the optimal outcomes
before changing.
To achieve an optimal assembly line, it firstly needs to analyze it to know where are
the bottlenecks and drawbacks for the current line. And then try to find the alternative
configurations to achieve the favourable results. The company would like to find the optimum
assembly line model with optimal inventory setup that is efficient and able to fulfil the
fluctuating demands economically by using Tecnomatix Plant Simulation before they setup a
new line.
1.3 Objectives
The objectives for this project are to:
a) Simulate the existing pull system assembly line
b) Determine the optimal minimum stock level and maximum stock level at the
warehouse to avoid
Delivery Shortage and
Over Stock
1.4 Scope and Limitation
Scope
The study is based on understanding of the theory and concept of assembly line
balancing and optimisation. A case study is conducted at a Tier 1 automotive supplier factory.
The data of an existing assembly line, producing Electric Power Steering Column Assembly
is collected, modelled and simulated using Tecnomatix Plant Simulation software. The
research is done with the help of Premium PLM Co. Ltd. in Bangkok, Thailand.
Limitation
This study emphasizes only on the final assembly line section though the entire
manufacturing line contains Production, Sub-assembly, Assembly and Warehouse sections.
Though they have many different product types, this product only consider a single product,
EPSC(Electric Power Steering Column) for Honda Fit.
2
CHAPTER 2
Literature Review
2.1 Assembly Line Balancing
Line balancing is about arranging a production line so that there is an even flow of
production from one work station to the next. Line balancing also a successful tool to reduce
bottleneck by balancing the task time of each work station so that there is no delays and
nobody is overburden with their task.
According to Falkenauer (2000) Assembly Line Balancing, or simply Line Balancing
(LB), is the problem of assigning operations to workstations along an assembly line, in such a
way that the assignment be optimal in some sense. Furthermore, an assembly line can also be
defined as a system which is formed by arranging workstations along a line. At these
workstations, work pieces can be transferred by using labor force as well as equipment, and
tasks are assembled taking into consideration precedence constraints and cycle time. The
decision problem of optimally balancing the assembly work among the workstations is
pointed out by M.Baskak (2008) as the assembly line balancing problem.
5
Falkenauer (2000) listed a few of the difficulties that must be tackled in a line
balancing tool in order to be applicable in the industry. Those difficulties are:
I. Workstation cannot be eliminated. Since each workstations has their own
identities, it is obvious that the workstations cannot be eliminated unless the
workstations were in front or at the end of the line. The elimination of any
workstation at the middle will create a gap or holes in the assembly line.
II. The load needs to be equalized. A small increase in the maximum lead time may
yield a substantial reduction in load misbalanced. Takt time is normally set by the
companys marketing that sets production target. The cycle time must not exceed
the given Takt time. But, it is normally useless to reduce the lines cycle time
below that value. Minimizing the cycle time is only required as long as it exceeds
the Takt Time. Once, the objective is met, equalization of workload should be
pursued instead.
III. Multiple operators. Once a workstation features more than one operator, the
workstations lead time ceases to be a simple sum of durations of all operations
assigned to it. Firstly, the whole workstation need to have time equal to the
slowest operator to complete all operations assigned to the workstation. Since
different workstation has different workload, hence it is surely not equal to the
sum of durations divided by the number of operator. The precedence constraint
that nearly exists among the workstations may introduce idle(waiting) time
between operations. This idle time reduces efficiency of the workstation and must
be reduced as much as possible.
In assembly lines, where specific operation robots are used, the importance of simultaneously
balancing of the resources and the assembly line can be understood better (Agpak and
Gokcen, 2005)
2.4.1 Bottleneck
Line balancing problem is the bottleneck of flow production. According to theory of
constraints (TOC) by (Goldratt and Cox, 1986) the throughput of manufacturing systems is
constrained by the capacity of bottleneck machines. In most situations, the final throughput of
manufacturing systems could be notably improved if the bottleneck machines are well
scheduled and controlled. However, how to define the bottleneck and how to design an easily-
implemented bottleneck detection method are still problematic at present.
In intermittent manufacturing, it is almost impossible to balance the available capacity
of the various workstations with the demand for their capacity. As a result, some workstations
are overloaded and others are under-loaded. The overloaded workstations are called
bottlenecks. Throughput is the total volume of production passing through a facility.
6
Bottlenecks control the throughput of all products processed by them. If work centers
feeding bottlenecks produce more than the bottleneck can process, excess work-in process
inventory is built up. Work centers fed by bottlenecks have their throughput controlled by
bottleneck and their schedules should be determined by that of the bottleneck.
2.5 Discrete Event Simulation
Simulation is a simplified system representation, made and understood by a
specialist, who intends to identify its potentials improvements (TAKO, ROBINSON, 2010).
The simulation aims to obtain results that can be transferred into reality (VDI, 1993). With
the help of simulation there can be tested factories even before the equipment is purchased
this generating information about the number of machines needed, cost, space required and
the optimal location of equipment. Simulation and modeling was increasingly used as a
decision-helping tool; its most important feature, which awakes an interest for simulation, is
the prospect of working with complexes systems and the possibility of analysis of the
dynamics behavior (BANKS et al., 2010).
According to (Sandanayake et al., 2008) discrete event simulation allied with
production system analysis, aiming at performance improvement, became more relevant in
the last decades. Together with advancements on computers, discrete event simulation helps
specialists in visualizing, analyzing and optimizing complex production processes, in a
reasonable period of time and with a reasonable investment. For example, Hernandez and
Librantz (2013) used discrete event simulation to evaluate 12 scenarios behaviour of supply
chain process in sugar cane exportation. In the end, they chose the last scenario because it
provided a better productivity.
Simulation is a reality imported to a controlled environment, where its behaviour may
be studied under a sort of condition, without involving physical risks and/or costs (BANKS et
al., 2010). Those conditions can be studied aiming at productivity and quality improvement,
machine acquisition, changes in layout and process parameters. Simulation is one of the most
used research tools, mainly due to its versatility, flexibility and power of analysis (RYAN;
HEAVEY, 2006). So discrete event simulations are a powerful tool, which can be applied to
investigate any stochastic system (HILLIER; LIEBERMAN, 2010).
2.6 Modeling and Simulation
According to Hillier and Lieberman (2010) computational simulations should be used
to solve complex problems. According to Leal et al. (2011) a simulation model development
presents three steps:
Conception or Problem Formulation,
Implementation and
Results analysis.
In the conception phases, the researchers define the simulations main objectives such
as physical layout planning, maximizing the efficiency, defining the suitable buffer space, etc.
Then, a conceptual model is built after all required data such as: batch size, production
frequency for each product, setup time, downtime and time to machine be repaired are
7
collected. The researcher should plan the simulation experiments and then run them
iteratively to verify the model.
Once the conceptual model is verified, implementation phase can be initiated. In this
stage, the conceptual model is transformed in a computational model, by setting up all the
methods and configurations in detail according to all collected data. After that, the
computational model must be verified and statistically validated. Verification and validation
methodology (Sargent, 2012) is shown in figure 2.1. The main different points in verification
and validation of a model can be compared as follow.
Table 2.1: Verification vs. Validation
Verification Validation
8
Simulation Performance and Uses
Increase
Throughput
Resource Utilization
Facility Utilization
Determine
Decrease
Throughput times
Required resources
Storage requirements.
Reduce WIP
Protect investments
Simulation can be used to plan a new facility by performing the following tasks:
9
Determine manpower requirements
Train the operators of the machines in the different states of the facility
10
Definition of Problem and Target
Analyze System
Acquire Data
Create Model
Validate Model
Evaluate Results
Recommendation
14
Pro Std App Run. Re Edu Stu
Feature T s d
Limited Functionalities
Build simulation models with any of the X X X(+) - X X X
built-in and with user-defined objects
Build simulation models with objects from X X X - X (X)* (X)*
libraries and any of the built-in material
flow objects (licenses for the libraries are
required)
Model with length-oriented objects X X X - X X X
Create and modify class objects, X X X - X X X
especially user-defined objects
Create and modify control codes X X - - X X X
employing the programming language
SimTalk
Create user-defined dialogs X - - - X X X
Create libraries X - - - X - X
Use the merge mechanism X - X X X X X
Import CAD files (DWG, DXF, DGN) as the X - X X X X X
background image of the Frame
Profiler X - X X X X X
Objects and functionalities which are not
available for all license types
Confidence analysis X - X X X X (X)*
DataFit X - X X X X (X)*
Experiment Manager X - X X X X (X)*
Factorial analysis X - X X X X (X)*
Neural Networks X - X X X X (X)*
Sequential Sampler X - X X X X (X)*
Statistical tools X - - - - - -
Attribute Explorer X - X X X X X
Card File X - X X X X X
Queue File X - X X X X X
Stack File X - X X X X X
Time Sequence X - X X X X X
Flow Control X - X X X X X
Optimization objects based on genetic
X - X X X X X
algorithms
Optimization Wizard X - X X X X (X)*
Generator X - X X X X X
HTML Wizard X - X X X X X
Layout Optimizer Wizard X - X X X X (X)*
Teamcenter Interface X - - - X - -
Trigger X - X X X X X
Variants Generator X - X X X X (X)*
XML Interface X - X X X X X
License Types:
Pro: Professional License
Std: Standard License
App: Application License
Run.T: Runtime License
Res: Research License
Edu: Education License
Stu: Student License
X Included
- Not available
(X)* The object can be loaded, the maximum number of objects might be exceeded if
you do so. In this case you can use the object, but you cannot save the model any
more.
O Optional product
(+) Except for the object Method
MaterialFlow objects, libraries, and modeled objects: EOM library; Conveyor library,
Cross-sliding Car library, High Bay Warehouse library, Kanban Objects library, Portal Crane/
Cranes library, and TransferStation.
User Interface objects: Comment, Display, Chart, Button, Checkbox, and Drop-Down List.
16
Chapter 3
Case Studies
Objective of Case Studies
The objective of this case study is to enhance the knowledge of properties, features and
different application of Tecnomatix Plant Simulation for different purposes.
3.1 Modeling, Simulation and Optimization of Process Planning (Jovisevic, S., Jovisevic,
V., & Jokanovic, S., 2009),
3.1.1 Overview
This paper is focusing on modeling of the manufacturing and assembly planning for
crank shafts using Tecnomatix Plant Simulation in 2D and 3D environment and determining
optimal parameters of production.
17
Figure 3.1: Segment From Simulation Model Shown in 2D
After designing satisfactory model of the manufacturing and assembly process plan
for crankshafts in the 2D environment, it is transformed into a model of process plan in 3D
environment. The purpose of transforming into 3D environment is primarily to determine the
spatial layout of machinery and equipment in the production plant.
18
Figure 3.3: Simulation model located in the corresponding production plant
Simulation model and production plant of the manufacturing and assembly process plan for
crankshafts by the system Tecnomatix Plan Simulation in 3D is shown in Figure 3.12 and
Figure 3.13. This model is founded on the basis of previous information on the number,
layout and positions of the necessary technological systems and equipment.
3.1.3 Simulation results
Improving the simulation model is performed through initial process simulation and analysis
of simulation results. In the model after the analysis of results, bottlenecks of production are
identified and accumulation of work in process(WIPs) in each station of the process plan are
also observed. Efficiency of machines performing the initial process of simulation, i.e. before
performing the optimization process, is unsatisfactory as shown in Figure 3.14.
19
After analysis of influential factors, optimization of process plan was performed,
where goal is set as a function of high efficiency machines and minimum time duration of
production cycle, while the limits were defined by the process plan. Resolving the above-
mentioned shortcomings, after introduction of buffer zones and increasing the number of
machines, it achieved a significantly shorter production cycle time. For example batch of 100
pieces of crankshafts, can save 40% of time in manufacturing process after the simulation.
After completion of design and simulation of process plan in the 3D environment, model
needs to determine the exact location of machines and devices in the spatial layout of
production facilities. Based on this information, manufacturing cycle is reduced by 5% as a
result of savings in time in transport parts between machines. Based on the layout and
location of machines and devices, it was designed a preliminary solution for production plant
with appropriate characteristics (departments, roads, entrances, exits, etc.). Results from
simulation of process plan provide cycle times generated for the appropriate batch size and
machine utilization of machines in the manufacturing process in percentage. According to
optimal designed simulation model, the utilization of machines for the batch size of 100
crankshafts is shown in Figure 3.15.
3.2.1 Overview
This article is based on a basic research for possibilities for KANBAN
implementation in one metal-working company. In that direction, the basic aim of the article
is to present benefits of combined implementation of KANBAN system and methodology for
Discrete Event Simulation (DES). After giving the brief theory on KANBAN and DES, the
article discusses the options for their integration. The case at the end is focused on one
production line in one metal-working company and experiments with the container capacity
which directly affects the number of KANBANs.
21
Data Collection and
Analysis
Kanban Number
Calculation
No
Is Kanban
Designation of Kanban designed properly?
Yes
Training
Improvement of the
system
Figure 3.6: Methodology for Kanban Implementation (Gross & McInnis, 2003)
22
3.2.3 Kanban Simulation Model
This section presents the application of the KANBAN system in a simulation model
made with the software suite Technomatix Plant Simulation by Siemens. Using Bankss
methodology, each and every problem is defined during the first step. Then, goals are set and
a model is conceptualized according to the acquired information. After formulating the
model, experiments which in normal circumstances might last for days, months or even years
are created in just few minutes. If after the result analysis it is concluded that the experiment
data are sufficient and correct, records are prepared and the final step, implementing the
solution, can be made.
24
Chapter 4
Description of System Under Study
25
Figure 4.1: Electric Power Steering System
Cover
Cover contains cover assembly which is produced in the sub-assembly line by pull
system and a retention ring.
27
Bracket Assembly Tilt
The bracket assembly is the section to attach the EPS Column to the rest of the system.
Bracket assembly contains:
- Lever Assembly Tilt
- Fixed Cam
- Washer
- Thrust Bearing
- Nylon Nut
- Bolt
- Bracket Assembly Tilt
- Bracket Assembly
- Spring and
- Retention Ring
Other Components
Items in other component group are the components that are assembled on the EPS
column other than the previous sub-assembly groups such as:
- Motor
- Electrical Control Unit
- Bracket Harness
- Harness Assembly
- Flang and
- Cover Connector
28
Figure 4.7: Bracket Harness
29
4.2 Production Process and Stations
4.2.1 Electrical Power Steering Column Production Line
Electric Power Steering Column production line contains 4 major sections:
1. Manufacturing Section
2. Sub-Assembly Section
3. Assembly Section and
4. Warehouse Section
Among these, only Manufacturing Section is run in Push system. All the other downstream
sections are run in Pull system.
Work Flow
Sub- Final
Manufacturing
Assembly Assembly Warehouse
Section
Section Section
Figure 4.12: Overview of the Electric Power Steering Column Production Line
4.2.2 Process Flow of Electrical Power Steering Column Assembly Line
This study will be emphasized on the Final Assembly Section in which all the
subassemblies and components are assembled as the final product. The line contains 25
processes ran by 17 human workers. According to the collected data in August 2016,
production capacity of the line is 721 finished goods per day. The production system is
running in 2 normal shifts(Day and Night) and 2 overtime shifts. Although the main flow of
the assembly line is sequenced to move the parts from a station to the downstream adjacent
station, some parallel stations(GA-4,GA-7,GA0-8,GA-15,FA-1) that are adding
supplementary components to the main assembly are configured to transport the processed
workpieces to next stations by human workers.
Table 4.1: Shift Schedule
Working time Day Night
31
The process flow chart along with the detail description of the activities and added
components in each station is described in figure 4.13.The process flow with each process
performed by particular worker are described as follow.
Worker1
At the beginning of the final assembly process, the Worker1 assembles the Worm End
at the Worm End Assembly station (GA-1) with Worm, 2 spacer rubbers, C-Ring and Flang.
Then the workpiece is carried by the worker to the Worm End Assembly Fitting
station(GA-2) and assembled with Gear Box followed by ball bearing fitting at the Gear Box
Bearing Fitting station(GA-3).
Worker2
In parallel, the Worker2 caulks 2 sets of bush and ball bearings at the Bush Caulking
station(GA-04). The workpiece at station GA-3 is then taken to the Lower Bracket Assembly
station(GA-05) to combine with caulked bushes.
Worker3
The Worker3 attaches the sensor column to the workpiece at the Sensor Shaft
Assembly station(GA-7).
Worker5
At the same time, the Worker5 is preparing the Cover at Shaft Cover Assembly
station(CA-07) combining Cover Assembly and Retention ring and also prepare the worm
wheel assembly at the Worm Spacer Caulking station(GA-8) which are then fed to the Worm
Assembly station.
Worker4
Worker4 at the Worm Assembly station(GA-9) combines the cover and worm wheel
assemblies with the main assembly. Then the assembly is checked whether the fittings and
the gears are working properly at the Motoring station(GA-10)
Worker6
The sensor shaft is then calibrated to the neutral voltage at the Neutral Voltage Adjust
station (GA-11) by Worker6. Seal and No. Label are also attached to the assembly at this
station.
Worker7
Worker 7 prepares the housing assembly at the Housing Assembly satation(GA-13)
and tests the mechanical movement of the worm gear at Worm Movement Test
station(GA-12).
32
Worker8
The tilt-lever assembly is prepared and combined with the housing assembly from
GA-13 in Tilt-PreAssembly station(GA-15) by Worker8. The sub assembly is then carried to
Tilt Assembly station(GA-16) and combined with the main assembly by the Worker8.
Worker9
Worker9 tests the function of the main assembly at the Tilt&Telescopic Operation Test
Station(GA-17).
Worker11
Worker11 combines the Electrical Control Unit(ECU) and electric Motor at
ECU&Motor Assembly station(FA-1). The Worker11 also perform the torque test of the
main assembly at the Rotational Torque Test station(FA-3).
Worker10
Worker10 attaches the assembly of Motor and ECU to the main assembly at the Motor
Assembly station(FA-2).
Worker12
Worker12 also performs the ECU&Motor assembling function at FA-1 station and
performs the performance test at the Performance Test station(FA-04).
Worker13
Worker13 inspects the assembly at the Pre-Shipping Inspection station(FA-5) and
attaches the Cover Connector at Bunded Table(FA-5.5).
Worker14
Worker14 performs the shipping inspection at (FA-6) and also perform the cover
attaching function at FA-5.5.
Worker15
Worker15 performs the overall inspection at the Final Check station.
Worker16,17
Worker16 and Worker17 perform the packing at the Packing station.
33
4.2.3 Plant Layout of Electrical Power Steering Column Assembly Line
The plant plan of the EPS column final assembly line is described in Figure 4.14. As it
can be seen in the figure, all the stations are ran by human workers. And even some of the
transportation of the workpieces between stations are also done by human workers. As it was
illustrated in the figure, the line includes the inventory stalls for the dedicated items in
between the particular stations. To be easier to understand, the description of the station
names according to each stations number are listed in Table 4.2.
Machining
7 0.79 5.8 8.9
Time
1 GA-1 1
Handling &
12 0.56 11.4 13.1
Oper:Time
Machining
21 0.4767 20.7 22.3
Time
2 GA-2 1
Handling &
6 0.6 5.2 7.2
Oper:Time
Machining
8 0.57411 7.2 9.1
Time
3 GA-3 1
Handling &
10 0.49 9.4 11.1
Oper:Time
Machining
20 0.872 18.9 21.9
Time
4 GA-4 2
Handling &
24 0.42 23.6 25.2
Oper:Time
Machining
0 0 0 0
Time
5 GA-5 2
Handling &
43 0.35 42.4 43.8
Oper:Time
Machining
4 0.5 3.8 5.2
Time
6 GA-7 3
Handling &
29 0.64 28.3 30.4
Oper:Time
Machining
8 1.08 6.5 10.2
Time
7 CA-7 5
Handling &
35 0.79 33.8 36.9
Oper:Time
Machining
30 0.56 29.4 31.1
Time
8 GA-8 5
Handling &
19 0.47 18.7 20.3
Oper:Time
Machining
0 0 0 0
Time
9 GA-9 4
Handling &
22 0.88 21.2 23.7
Oper:Time
Machining
45 0.59 44.2 46.2
Time
10 GA-10 4
Handling &
10 0.87 8.9 11.9
Oper:Time
Machining
7 0.49 6.4 8.1
Time
11 GA-11 6
Handling &
46 0.5 45.8 47.2
Oper:Time
37
Table 4.4: Time Study of the EPS Column Final Assembly Line(Cont.)
Machining
38 0.43 37.6 39.2
Time
12 GA-12 7
Handling &
16 1.08 16.5 20.2
Oper:Time
Machining
5 0.36 4.4 5.8
Time
13 GA-13 7
Handling &
32 0.48 31.3 32.9
Oper:Time
Machining
0 0 0 0
Time
14 GA-15 8
Handling &
44 0.5 43.1 44.9
Oper:Time
Machining
0 0 0 0
Time
15 GA-16 8
Handling &
46 0.64 45.4 47.5
Oper:Time
Machining
8 0.7 7 9.7
Time
16 GA-17 9
Handling &
26 1.11 24.9 28.4
Oper:Time
Machining
0 0 0 0
Time
17 FA-1 11/12
Handling &
24 0.37 23.2 24.8
Oper:Time
Machining
0 0 0 0
Time
18 FA-2 10
Handling &
46 0.42 45.4 46.9
Oper:Time
Machining
23 0.59 22.2 24.2
Time
19 FA-3 11
Handling &
17 0.43 16.3 17.9
Oper:Time
Machining
21 0.36 20.3 21.9
Time
20 FA-4 12
Handling &
22 0.51 21.5 23.2
Oper:Time
Machining
5 1.11 3.9 7.4
Time
21 FA-5 13
Handling &
55 1.13 53.4 57.2
Oper:Time
Machining
0 0 0 0
Time
22 FA-5.5 13/14
Handling &
15 0.69 14 16.7
Oper:Time
Machining
0 0 0 0
Time
23 FA-6 14
Handling &
32 0.48 31.3 32.9
Oper:Time
Machining
0 0 0 0
Time
24 QA 15
Handling &
51 0.64 49.6 51.9
Oper:Time
Machining
0 0 0 0
Time
25 PACKING 16/17
Handling &
48 0.57 47.2 49.1
Oper:Time
38
4.2.5 Kanban System
In order to limit accumulation of inventory in WIP and keep the efficient stock level in
the warehouse, final assembly line of Electrical Power Steering Column is running in pull
system. According to the stock level in the Ware House section after Assembly section, the
Kanban cards are circulated to assemble the required number of finished products to fulfil
the empty containers. A Kanban box contains 12 identical finished products. Here it can be
considered a box Kanban represents 12 identical Kanban cards. For the Electrical Power
Steering Column assembly, the product under study, the maximum stock level is 5 boxes(60
pcs) and the reorder stock level is 2 boxes(24 pcs). Once the stock level in the warehouse
falls under re-order level(2 boxes), the warehouse sends out the Kanban cards to produce
finished products and refill up to the maximum stock level.
Similarly, Kanban systems for the sub-assembly lines are also triggered by the stock
levels in the buffers on the final line, for example Sensor Shaft Assembly buffer on GA-07
station triggers the sub-assembly line for Sensor Shaft Assembly. But in this paper, these sub-
assembly lines are not considered and these buffers are assumed to avoid the shortage.
39
Kanban card in the EPS Column Final Assembly Line describes the following
specifications of the product which will be produced.
1. Product Type
2. Product Picture
3. Production Line to be produced on
4. Product ID
5. Package Bar Code describing the package that the product belongs to
6. Kanban ID
7. ID of Sensor Shaft Assembly used in this product
8. ID of Electrical Control Unit used in this product
9. Quantity of products in this box
6.
1.
7.
2. 5.
8.
3. 4. 9.
40
Chapter 5
Model Development and Optimisation
The simulation model of the current assembly line is developed according to the
collected data. At this point, in addition to the shop-floor data, background production
management practices and arrangements that are applied on the production system such as,
push or pull system of production, allocation of stores and buffers between the sections etc.,
will also need to be feed in. The model will be developed in following steps:
Basic
Model
Development
Worker
Configuration
System
Performance
Analysis
System Calibration
Verification
and
Validation
Not
Validated
Validated
Pull
System
Configuration
Optimisation
Recommendation
Figure 5.1: Modelling Steps Flow Chart
41
5.1 Basic Model Development
Basic model development phase contains the development of the stations with
particular settings and material flow routes. First of all single-proc stations are placed
according to the plant plan. For the stations that have both operation time and machining
time, a pre-station is added in front of the machining station as a separate station to improve
the flexibility in modelling. Station allocation in the assembly line plan is shown in figure
5.2.
To declare the number of workers who will be working in the model, configure the
workerpool. Open the creation table in Attribute tab. Key in the 17 workers as shown in
figure5.7.
43
Figure 5.7: Creation Table of Workers in Workerpool
In the object workplace on Attribute tab, the respective station is described and in the
supported service window, key in the StandardService to operate the station. Incase, if the
worker is supposed to perform the other services such as servicing, repairing etc. the service
has to be declared in this window.
44
Figure 5.9: Supported Service Window
In order to declare the station that the operation at this station will be done by human
worker, in the tab of Importer, processing must be activated. Key in the name of Broker.
Priority of the station can be set if single worker is working for multiple stations. Importer
window of the process CA7-Pre is shown in figure 5.10 for example.
45
Figure 5.11: Exit Window
For applying the working shifts to the model, a Shift Calendar is added. Shift details
that are discussed in section 4.2.2 are added to the calendar. And then in the worker pool,
name of the shift calendar that the workers will work accordingly is added.
46
Figure 5.13: WorkerPool Window
5.3 Verification and Validation
Before the model is configured to the pull system as a part of verification and
validation, it is analysed to measure its performance in push system and approach of
Correlated Inspection(Law and Kelton, 1991) as shown in figure(5.14) is applied. The system
is run for a day and analysed the behaviour of the results. All the configurations such as:
- Shift Schedule(Shifts, Overtime, Breaks)
- Process Flow
- Workers Operation Time(Travelling, Handling, Machining)
- Workpiece Transportation, etc. are tuned up to get the closest representing model
compared to the real system.
After fine tuning, it was found out that the number of throughput is 724 per day while the
daily production rate of the real system is 721 per day.
48
5.4 Pull System Configuration
As it was discussed in section 4.2.5, the EPS Column final assembly line works as a
pull system. There is a warehouse after the packing station. In the model, after the finished
goods are packed into boxes(12 pcs/box) in packing station, these boxes are sent to the
kanban buffer. In Kanban buffer, minimum stock level, maximum stock level and initial
stock level are defined. Stocks from the Kanban buffer is withdrawn by the Kanban
singleproc. station in which the withdrawal rate, in another word, order arrival rate can be
defined.
49
Once the model is run, the Kanban buffer will have the initial stock and the Kanban
Singleproc will withdraw a box from Kanban buffer. Kanban buffer will withdraw one box in
each hour. Once the stock level in Kanban buffer gets to minimum stock level(2 boxes), the
Kanban source will produce a Kanban box that contains 12 Kanban cards. The assembly line
will start producing the finish goods and the packing station will collect 12 finished goods to
pack in a box. Afterwards, the box is forwarded to the Kanban buffer and the stock level will
raise. Transitions of the stock levels are shown in figure 5.19. When the stock level gets back
to maximum stock level(5 boxes), Kanban source will stop producing Kanban boxes and the
assembly line will stop producing finished goods.
50
Figure 5.20: Stock Level Transitions in Kanban Buffer
In this model, the initial stock level is set as 5 boxes.And the withdrawal rate is set as
1 hour constantly. So, In the first stage, once the model is run, a box from the buffer is
withdrawn. Therefore stock level at time zero show 4. After an hour as another box is
withdrawn again, the stock level drops down to 3. After another hour, as the stock level
reaches the minimum level (2 boxes), the Kanban source produces the Kanban boxes to start
the assembly operation. After producing a box of finished good, it is transferred to the
Kanban buffer and the stock level raised again to 3. The assembly line keeps producing the
finished goods till the stock level gets back to the maximum level(5 boxes). As the
withdrawal rate is set as 1 hour constant, at each hour, another box of stock will be taken
from the Kanban buffer. Here in this point, as the withdrawal rate in real system is working
on the production plan discussed in section 4.2.6, the model is modified to withdraw the
finished goods according to the production plan on next section.
51
5.5 Implementation of Production Plan
As the final assembly line is run according to the order plan, production plan table by
which the finished Electric Steering Columns from the warehouse will be withdrawn and
delivered to the customer is implemented in table file as shown in table 5.1. The production
plan table contains the planned orders(no. of finished goods) on each day of a month. The
finished goods withdrawal time is set to be at 3:00 pm (or) 15:00 for every day in the first
column of the delivery summary table as shown in table 5.2.
Table 5.1: Production Plan Table.
ITEM 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31TOTAL
PT22XX0264S
38 0 36 36 36 36 36 0 0 36 36 36 36 36 0 0 36 36 36 36 36 0 0 24 24 0 24 24 24 0 0 698
Once the model is initialized, the initial stock in the warehouse is created at the
maximum stock level, min and max values in the warehouse are defined and the piece count
in the delivery station is set as zero. First six columns of delivery summary table is created
for the planned orders. The delivery status is recorded in the delivery summary table
describing:
- Delivery Time
- Object Name in Model
- Number of Orders
- Product Name
- Attributes
- Order Quantity
- Delivered Quantity
- Delivery Shortage Check.
In order to control the delivery, a method is implemented in which the current stock in
the warehouse and delivery station are checked whether the order can be fulfilled or not.Till
in the warehouse, finished goods are counted in dozen packages. To deliver the finished
goods in piece count, the packages are unpacked into pieces in the delivery station after taken
out from the warehouse. The method checks the order and the stock level and if the stock is
enough for current order, the required number of boxes from the warehouse are requested and
unpacked and sum with the remaining quantity at the delivery station. The order is then
subtracted from the unpacked quantity. The remaining quantity is kept at the delivery station.
If the stock is enough to fulfil the requested order quantity, Delivery Shortage Check
is remarked as No meaning no delivery shortage in the delivery summary table. If the stock
is not enough to fulfil the requested order quantity, only the in hand stocks are delivered.
Delivery Shortage Check is remarked as Yes meaning there is a delivery shortage and the
value for Delivery shortage counter is counted up. When the stocks are available in
warehouse, a package is requested and unpacked to keep at the delivery station. Whenever
the stock level falls under Minimum Stock level, the assembly line starts production to
replenish the warehouse up to the maximum stock level.
52
The reorder quantity is defined as:
Reorder Quantity = Maximum Stock - Current Stock
When the model is ready to run, the event controller is set to run for a month(31 days). In the
early models, the minimum stock level is set to be 2 boxes and the maximum stock levels is
set to be 5. In this configuration, the model was observed to have many delivery shortages as
presented in table 5.2. The delivery status on each day of the month are described in the
Table 5.2: Delivery Summary Table Before Optimisation
delivery summary table as described. As it can be seen in the delivery summary table, 7
delivery shortages are observed in a month.
53
Figure 5.22: Method for Delivery Control
5.6 Optimisation Using Genetic Algorithm Wizard
In order to optimise the model to minimise the number of delivery shortages, genetic
algorithm wizard in Plan Simulation software is used. In the Genetic Algorithm Wizard, it
needs to describe the objective, Minimization or Maximization. And to define the number of
simulation runs, it needs to describe,
- Number of Generation
- Size of Generation and
- Observations per Generation
54
results from this formula:
Number of simulation runs = observations per individual * (generation size + 2 *
generation size * (number of generations - 1))
For the case, it is using 20 observations for evaluating an individual and create 5
generations with a frequently used Size of generation of 10, Plant Simulation has to execute
1800 simulation runs (20(10+2104)).
For the optimisation parameter, lower bound, upper bound and increment value of
the variables in the optimisation process are described as shown in figure 5.23.
55
Figure 5.26: Optimised Results
56
Figure 5.27: Stock Level Graph After Optimisation
As the initial stock level is set at maximum stock level, at the beginning, the stock
level starts at 6 boxes of finished goods. On the first day, 48 pieces are withdrawn. To deliver
48 pieces, the warehouse needs to provide 4 boxes delivering 3 boxes and 2 pieces to the
customer and keeps the 10 rest pieces in the delivery station. As the inventory level fell down
to 2 boxes under minimum stock level(3boxes), the assembly line starts producing.
For the first day, reorder quantity is 4 boxes of finished goods. On the second day
there is no activity as there is no planned order for that day. On the third day, the planned
order, 3 boxes were withdrawn and the stock level fell down to the minimum stock level. The
warehouse was replenished again up to 6 boxes. The system goes on in similar behaviour for
the following days as it can be seen in figure 5.27. Day by day records of the inventory levels
in the warehouse are described in the following figures.
57
Figure 5.28: Inventory Level on Day1-Day3
58
Figure 5.29: Inventory Level on Day4-Day6
59
Figure 5.30: Inventory Level on Day7-Day9
60
Figure 5.31: Inventory Level on Day10-Day12
61
Figure 5.32: Inventory Level on Day13-Day15
62
Figure 5.33: Inventory Level on Day16-Day18
63
Figure 5.34: Inventory Level on Day19-Day21
64
Figure 5.35: Inventory Level on Day22-Day24
65
Figure 5.36: Inventory Level on Day25-Day27
66
Figure 5.37: Inventory Level on Day28-Day30
67
Figure 5.38: Inventory Level on Day31
According to the delivery summary table and the daily stock level graphs, it can be
concluded that the model is optimised for no delivery shortage at the minimal inventory in
stock to minimise the inventory cost. Here the optimisation is considered based on monthly
basis. Even in case for the quarterly basis or batch by batch basis, the basic model can be
used to play with the desired scenario for optimising the most suitable inventory level.
68
Chapter 6
Conclusion And Recommendations
6.1 Conclusion
The purpose of this study is to analyse the performance measures of the assembly line
and find out the optimum inventory setting to minimise the Delivery Shortages. Over the
study during this project, it can be concluded as follow:
1. All the line information of the final assembly line under study are collected from a Tire 1
Power Steering Supplier Company including some historical data and observed data.
Samples of the observations were processed to have appropriate input data for simulation
model in chapter 4.
2. Tecnomatix Plant Simulation is used to model the assembly line based on the processed
data. Initially, it observes the performance of the final assembly line at full capacity
(modelled in Push System). In this phase, the model is calibrated to verify and validate with
the real system as discussed in section 5.3.
3. The line performance of the assembly line in pull system is observed using Plant
Simulation (modelled in Pull System) as described in section 5.4.
4. Real Scenario of production according to monthly delivery plan is implemented and
observed the weaknesses(Delivery Shortages) in section 5.5.
5. Optimisation of the model was performed and an optimised inventory setup to minimise
the Delivery Shortages is proposed in section 5.6.
6.2 Recommendation
Some possible future studies on assembly line analysis and improvement are
recommended as follow:
1. Sub-assembly lines for the sub assemblies supplied to the final assembly line can be
modelled and integrate with the existing final assembly line to observe more realistic
behaviour.
2. Warehouse with the inventory management model supplying the outsourced components
can be implement and integrate with the existing final assembly line to observe more realistic
behaviour.
69
REFERENCES
- Jovisevic, S., Jovisevic, V., & Jokanovic, S. (2009, December). Modeling, Simulation and
Optimization of Process Planning. Journal of Production Engineering, 12(1), 87-90.
- TAKO, A. A.; ROBINSON, S. Model development in discrete-event simulation and system
dinymics: an empirical study of experts modellers. European Journal of Operational
Research, v. 207, n. 2, p. 784-794, 2010.
- BANKS, J.; CARSON II, J. S.; NELSON, B. L.; NICOL, D. M. Discrete-event simulation.
5th New Jersey: Prentice-Hall, 2010.
- SANDANAYAKE, Y. G.; ODUOZA, C. F.; PROVERBS, D. G. A systematic modeling and
simulation approach for JIT performance optimization. Robotics and Computer- Integrated
Manufacturing, v. 24, n. 6, p. 735-743, 2008.
- HERNANDEZ, M. A. G.; LIBRANTZ, A. F. H. Improvement of the supply chain for the
sugar cane exportation process employing discrete event simulation techniques. Acta
Scientiarum. Technology, v. 35, n. 4, p. 637-643, 2013.
- RYAN, J.; HEAVEY, C. Process modeling for simulation. Computers in Industry, v. 57, n.
5, p. 437-450, 2006.
- HILLIER, F. S.; LIEBERMAN, G. J. Introduction to operations research. 9th ed. New
York: McGraw-Hill, 2010.
- LEAL, F.; COSTA, R. F. S.; MONTEVECHI, J. A. B.; ALMEIDA, D. A.; MARINS, F. A.
S. A practical guide for operational validation of discrete event simulation models. Pesquisa
Operacional, v. 31, n. 1, p. 57-77, 2011.
- PEREIRA, T. F.; MOTEVECHI, J. A. B.; MIRANDA, R. C.; FRIEND, J. D. Integrating
soft systems methodology to aid simulation conceptual modeling. International
Transactions in Operational Research, v. 22, n. 2, p. 265-285, 2015.
- SARGENT, R. G. Verification and validation of simulation models. Journal of Simulation,
v. 7, n. 1, p. 12-24, 2012.
- LAW, A. M., KELTON, W. D., 1991, Simulation Modeling and Analysis. Mc. Graw - Hill.
Inc
70
Appendix A
Init Method
Var tb:table[string,integer,boolean,string,real,time,date,datetime,object,length,weight,speed]
Delivery_Order.Delete
KanbanBuffer.Init_Kanban := KanbanBuffer.Max_Kanban
Kanban_Chart.MyAnnotation["Value",1] := KanbanBuffer.Min_Kanban
Kanban_Chart.MyAnnotation["Value",2] := KanbanBuffer.Max_Kanban
Kanban_Chart.Annotations := Kanban_Chart.MyAnnotation
Kanban_Chart.YScaleMax := KanbanBuffer.Max_Kanban * 1.25
Kanban_Plotter.YScaleMax := KanbanBuffer.Max_Kanban * 1.25
Kanban_Plotter.XRange := EventController.End
for var i := 1 to KanbanBuffer.Init_Kanban
.SNSS_MU.Box_PT22XX0264S.Create(KanbanBuffer)
next
Packing.Flag := 0
Delivery.Part_Unbox := 0
EventController.Date := str_to_datetime(to_str(Order_Plan[2,1]) + "/" + Order_Plan[1,1] + "/
1" + " 00:00:00")
for var i := 3 to Order_Plan.XDim
for var j := 1 to Order_Plan.YDim
if Order_Plan[i,j] /= 0
Delivery_Order[6,Delivery_Order.YDim + 1] := Order_Plan[i,j]
Delivery_Order[4,Delivery_Order.YDim] := Order_Plan[0,j]
Delivery_Order[3,Delivery_Order.YDim] := 1
Delivery_Order[2,Delivery_Order.YDim] :=
str_to_obj(".SNSS_MU.Box_" + Order_Plan[0,j])
Delivery_Order[1,Delivery_Order.YDim] :=
str_to_datetime(to_str(Order_Plan[2,j]) + "/" + Order_Plan[1,j] + "/" + Order_Plan[i,0] + "
15:00:00")
tb.create
tb.ColumnWidth := 10
tb[1,1] := "No"
tb[2,1] := Delivery_Order.YDim
tb[1,2] := "QTY"
tb[2,2] := Order_Plan[i,j]
Delivery_Order[5,Delivery_Order.YDim] := tb
tb.forget
end
next
next
71
Delivery Control Method
72