The current issue and full text archive of this journal is available at

www.emeraldinsight.com/2040-4166.htm

IJLSS
3,3 Reducing electronic component
losses in lean electronics
assembly with Six Sigma
206
approach
Tan Ping Yi, Chin Jeng Feng, Joshua Prakash and Loh Wei Ping
School of Mechanical Engineering, Universiti Sains Malaysia, Penang, Malaysia

Abstract
Purpose – In electronics assembly, the losses of electronic components throughout the
surface-mounting process (including kitting and setup) are hard to trace. This affects accurate
material planning and manufacturing costing. This paper aims to investigate this issue and to
generate a suitable mixture of strategies for the relevant causes.
Design/methodology/approach – The project is executed by an undergraduate manufacturing
engineering student and several company engineers over a period of ten weeks. Define, measure,
analyze, improve, and control (DMAIC) approach delineates the project stages. The solutions devised
must be in agreement with lean philosophies and practices currently upheld in the company.
Findings – Component losses stem from multiple sources and are complicated by inherent
information inaccuracies. A right mixture of strategies is envisaged on analysis on these sources. An
average 18 percent of decrement in component losses in monetary value is achieved in the initial
16 weeks of the improvement phase.
Research limitations/implications – The DMAIC approach induces a focused, systematic and
thorough study on the selected area. For the limitations, this study is based on a single industrial case.
The evidence may be anecdotal and idiosyncratic to the electronics assembly industry. The final
solutions which emerged need to factor in the organization current maturities in Lean and Six Sigma
concepts.
Practical implications – Component loss is a common problem faced by electronics assembly
industries. In this paper, the nature of the problem and the related investigation are extensively
illustrated in the context of the case study. As many electronics assembly industries have embarked on
Lean or Lean Six Sigma journeys, the savings and data accuracy improvement achieved in this case
study provide valuable benchmarks.
Originality/value – The issues related to electronic component losses have not been reported in
established literature to date. This is also the first reported success case study of applying DMAIC to
address these issues in a lean company.
Keywords Electronics assembly, Component losses, Variability reduction, DMAIC, Electronics industry,
Electronic equipment and components, Lean production, Six Sigma
Paper type Research paper

1. Introduction
Lean manufacturing and Six Sigma are the two emergent production management
International Journal of Lean Six paradigms prevailing in recent years. Lean manufacturing concentrates on total
Sigma elimination of production wastes to create a production system highly adaptive to meet
Vol. 3 No. 3, 2012
pp. 206-230
q Emerald Group Publishing Limited
2040-4166
The authors wish to acknowledge Universiti Sains Malaysia (USM) for the full financial support
DOI 10.1108/20401461211282718 through the short-term grant no. 60311042.
market demands. Six Sigma seeks improvement on output qualities in customer Reducing
perspective, primarily by minimizing all forms of variability and occurrences of process component
errors. While both paradigms share many commonalities, they coexist in a management
system complementarily in many cases. Industrial enterprises adopted these paradigms losses
have gained substantially in efficiency, response speed, and production flexibility, and
consequently enable them to diversify products at low cost, with high productivity,
prompt delivery, minimum stock levels, and optimum quality (Cuatrecasas Arbós, 207
2002). One example of this gain is highlighted in Breyfogle (2010), where project
improvement efforts of a business management system are tackled. Similarly, in Sharma
and Chetiya (2010), success of the Six Sigma project selection approach in the Indian
manufacturing industry are studied and several critical factors are identified.
One of the common goals shared by both paradigms is to reduce variability.
Succinctly, variability or variation is any unwanted condition or the difference between
a current and a desired end state (Motley, 2005). In enterprises today, identifying and
reducing variability in key product features and manufacturing processes have
become imperative. The root cause of variability, nevertheless, is often multifold and
largely dependable on the technology used, types of processes, its development, and
handling.
In surface mounted technology (SMT) process, electronic components, from dozens
to hundreds, are interconnected and mounted on a printed circuit board (PCB). The
sizes of these electronic components are not homogeneous, with some measuring as
tiny as in millimetres. The process often results in a percentage of component losses,
which are difficult to trace. The losses have caused wastages as well as significant
discrepancies in material planning and manufacturing costs. To understand the extant
of the problem in this industry, we approached several electronic assembly plants with
previous established research relationships, at the early stage of the project. The
consensus corroborates the prevalence of component losses at SMT process. The losses
aggravate in high mix low volume productions due to increased changeovers and
higher component variety. Normally, customers have agreed levels of lead in (setup or
splicing loss) and wastage based on types of components, approximately 2-4 percent.
Company policies in general permit operators to directly dispose short strips as long as
the length is uneconomical for splicing. In management, inventories are over issued to
production according to the component sizes. These inventory levels are continuously
readjusted according to machine scrap rates and estimated component consumptions
or when significant shortages occur.
An extensive literature survey on recent journal publications on related topic was
unfruitful. This presages the research is new and worthwhile reporting. Electronics
are becoming increasingly vital in many product function designs. For example, in the
automotive industry, electronics are widely serving four main purposes: to improve
drivability, to enhance safety features, to lower environmental burden and to realize
greater operational reliability. According to Leen and Heffernan (2002) and
Sangiovanni-Vincentelli (2003), electronics comprise more than 25 percent of a normal
car’s cost. This percentage will continue to rise in the future based on the current demand
and technology trend.
In this paper, we reported a project spearheaded to reduce electronic component
losses in an electronics assembly plant. The project follows define, measure, analyze,
improve, and control (DMAIC), a problem solving methodology originated from
IJLSS Six Sigma concepts. The current paper is organized as follows: Section 2 introduces the
3,3 study of variability in literature and general procedures proposed to solve it. Section 3
presents DMAIC methodology followed in the current research. Section 4 presents the
company in the case study. Section 5 explains the problem statement of the case study.
Section 6 reveals the process flow of raw materials and the measuring method, and
Section 7 describes the root causes of the problem after analyzing the data. Section 8
208 provides a suitable solution for solving the root causes, and Section 9 presents the
results and the outcomes of the project. Finally, the research conclusion is provided in
Section 10.

2. Literature review
Variability naturally exists in every aspect of manufacturing activity. Motley (2005)
distinguish four main sources of variability in manufacturing processes:
(1) poor design practice and unrealistic requirements causing insufficient design
margins;
(2) poorly defined operating environments and inaccurate design reference;
(3) inherent variability of manufacturing processes; and
(4) manpower, materials, work methods, machinery, and measurement.

Quality management research has a definite view on variability: it is dichotomized into

usual or unusual. Usual variability is inherent and occurs by chance; therefore, there is
no noticeable pattern and it is uncontrollable. Unusual variability is one with
assignable causes, which can be managed or eliminated. Control of the process then
falls on maintaining its performance within the statistically calculated control limits
where only the common causes are exhibited (Summers, 2006).
In manufacturing, the impacts of variability, including environments and systems,
are often of interest. According to Heizer and Render (2008), variability deprives
manufacturing systems of the optimum process that consistently delivers perfect
products on time. The effects can be short term or long term. Variability entails
fluctuations in production system performances, giving rise to a higher number of
defective products or products not fitting customer specifications. The inherent
unpredictability in production processes complicates planning for throughput and
product quality control. The consequential build up of inventory, a common strategy to
counteract variability, decreases the capacity of the factory because the processes are
not fully exploited to produce in response to actual customer demands. The discrepancy
of planned and actual output adds a new dimension to the variability of the production
system. When multiple variability is present in the shop floor, launching improvements
become difficult because the root cause may be hidden.
Variability exists in all parts of manufacturing. The research on managing the
manufacturing variability is selective and topical. Jing (2003) investigated the effects of
reducing the setup/processing time variability of job shops and cellular manufacturing
with the kanban system. Through simulation, they found that reducing processing
time variability is better for cellular layouts and reducing setup time variability is good
for functional layouts or cellular layouts. On the other hand, Xu (1996) illustrated the
detection of the variability method using the clustering technique. Romero Rojo et al.
(2009) identified the sources of variability in cost estimation and developed a software
prototype that reduces its variability. He et al. (2007), based on the Markovian arrival Reducing
process (MAP) approximation method, determined production variability and claimed component
that the MAP is effective in detecting production variability in production lines
compared with the state-space based method. losses
In an empirical study carried out by Mapes et al. (2000) for process variability and
its effect on plant performance over 953 plants in the UK, variability was found to be
the key manipulator distinguishing high-performing plants from its low-performing 209
counterparts. The high-performing plants exhibited high adherence to schedule, low
process time variability, high reliability of delivery by suppliers, and low variability in
the process output. The less variability there is in a plant’s system, the better its
performance. These elements have also been identified in Shetty et al. (2010), based on
the assessments of a company’s lean initiative.
To control and reduce the different variability found in the
manufacturing system, numerous methods have been proposed in literature.
The prevalent ones include the Plan-Do-Check-Act (PDCA) cycle, DMAIC,
Initiating-Diagnosing-Establishing-Acting-Learning (IDEAL), ISO/IEC 15504, etc. While
DMAIC builds on earlier approaches to problem solving such as PDCA, the rigor required
throughput DMAIC projects is one of the aspects that differentiates DMAIC form earlier
approaches (Bertels, 2003). Despite this rigor, it provides value in optimizing repeatable
processes by way of reducing wastes and making incremental changes (El-Haik and
Shaout, 2010). In addition, a focus on quality achieved through variation reduction can be a
core element of a company’s philosophy and strategy (Goldsby and Martichenko, 2005).
Generally, DMAIC uses statistical, mathematical, engineering and management tools to
improve the process in a sequential way (Kumar et al., 2006). However, a team may have to
go through several iterations until the objectives of a project are met (Kumar et al., 2006).
The versatility of the DMAIC framework is epitomized in different cases of reported
applications; improvements include quality of PCB (Tong et al., 2004), supply chain
(Knowles et al., 2005), consumption of demineralized water in a thermal power plant
(Kaushik and Khanduja, 2009), heat efficiency and energy consumption in the steel
industry (Jang and Jeon, 2009), process variation and defect rates in engine manufacturing
(Antony et al., 2005), and customer relationship management (Pan et al., 2007).
General procedures in handling variability or variation can be simplified into four
steps. First, detect and determine the potential variability that exists in the production
system. Xu (1996) used clustering technique to screen out possible process variables,
and then used residual analysis to determine them. Second, analyze the potential
variability effects toward the system using different suitable measuring models.
Tan (1998) showed the effect of variability on the performance of unreliable production
systems using the Markov process. Third, choose and use suitable tools to eliminate or
reduce the effects of variability. For example, statistical process control (SPC) is
deployed to control product quality through random sampling. Jing (2003) used the pull
system to improve the performance of the production system. Romero Rojo et al. (2009)
proposed qualitative methods for earlier stage rough cost estimation and quantitative
methods when more accurate cost estimation is required. Lastly, evaluate the result of
the variability reduction and determine areas for future work. Many authors agree that
there is no best method for reducing variability, rather, the solution is largely
dependent on the situation. That is, addressing each type of variability requires an
understanding of the contextual issues involved.
IJLSS 3. Methodology
3,3 The DMAIC flow chart pertaining to the present case study is shown in Figure 1. The
methodology used in this project is broken into the respective five stages – DMAIC.
Each stage consists of a number of steps. The steps and their sequence are conceived
by considering into effective deployment of human resources available as well as based
on the nature of the project and the project’s permitted duration. In order to gain a
210 better understanding of the involved processes at the company, upon receiving the
assignment of the project, interviews are conducted with several staff mainly from
three levels: process manager, machine maintenance personnel and operators. Data
from different sources have to be collected individually. The data collected are
analyzed, cross-checked and used to create Ishikawa diagram, new standardized work
plans, and plant layouts. Elucidation of the framework on the undertaking of
individual stages is given next.

3.1 Define phase

During scope definition, the problem statement and project goal must be thoroughly
clarified. These include the impact of the problem toward the company in terms of cost
and quality.

Problem definition

DEFINE
Interview

Identify potential
areas of losses occur

Tracing model
MEASURE
selection

Tracing process and

data acquisition

Data/information
verification
ANALYZE
Data analysis

Improvement proposal IMPROVE

Figure 1.
Monitoring &
DMAIC methodology
continuous CONTROL
flow chart
improvement
3.2 Measure phase Reducing
In the measure phase, the process flow must be well grasped. The workstations component
and models that require focus must be identified. Data are collected. The measuring
method used in the process performance must be validated to ensure data accuracy and losses
suitability for analysis.

3.3 Analyze phase 211

The data (such as measurement) collected are analyzed and the source of variability is
determined. A number of quantitative and qualitative tools can facilitate further
understanding of the related cause and effect.

3.4 Improve phase

The solution for eliminating the major root cause is proposed in this phase.
Implementation of the solution is performed and its performance in terms of the
process is monitored.

3.5 Control phase

To ensure the sustainability of the improvement done, further monitoring and control
is essential. Control strategies are recommended for continuous improvement toward
the methods implemented.
The deployment of DMAIC to guide the project stems from the exploratory nature of the
project. The investigation of component losses requires a systematic and orderly procedure
to ensure a wholesome approach to the problem which involves multiple undetermined
variables. Specifically, the project is the first attempt by the company on the component
losses. A target has not been prescribed by the management and therefore will be set in the
first stage of the project. Moreover, as SMT is a complicated and advanced technology
coupled with manual supply of electronic components, many sources potentially
contribute to component losses. Decisions need to be reached on sources to be included and
methods of capturing data. Later, with the comprehensive amount of data collected, the
analysis and implementation demands a collective effort from a team. Finally, some forms
of control plan, particularly on dealing with enforcing change in practices may need to be
exercised to stabilize the losses produced at certain sources involving manual works. This
necessitates a distinctive control stage after implementation of solution.

4. Case study
4.1 Company profile
Established in the northern part of Malaysia, Company X assembles automotive
electronic components. Their products, including lamp control units, electronic parking
brakes, door lock modules, digital clocks, and so on, are supplied to renowned
automotive companies, both local and international. The plant operates three shifts
per day, six days per week (overtime on Sunday, if there is backlog), and has a total of
592 employees. An average of 10,000 products is assembled daily.

5. Define phase
5.1 Description of component losses problem
The tenet of the current project is to investigate the losses of electronic components
(such as IC, resistor, capacitor and so on) that occurs during the production process.
IJLSS As aforementioned, component loss is not unique in the electronics assembly industry.
3,3 Electronic components are retrieved from the storage to the front end processes prior to
the production based on the order received. Kit preparation is manual and the recipe of
the components is order specific. The processes take place at a sequential automatic
line with manual loading.
As a common practice, allowance for discrepancy is given, and, as long as loss is kept
212 within this allowance, the cost is absorbed into the production cost. The current problem
faced by the company’s production line is the high losses in electronic components. The
problem existed since the SMT was introduced to the company but was largely ignored.
Recently, it was bought to light due to several reasons. First, the prices for electronic
components have increased significantly in the past few years. Second, the losses on
electronic components escalate proportionally to the rising consumptions resulted from
more compact circuit designs and capacity expansion in the company. Third, stiff
industry competitions compel the company to save cost aggressively.
In a preliminary investigation ran by the company, the production department was
unable to clearly identify of which workstation or process generated the highest
percentage of losses as well as to relate one type of component loss to specific causes.
Their efforts were also hampered by serious inaccuracy in the inventory data. An
undergraduate student is sent to Company X for a period of ten weeks to investigate
the matter as internship project. The student is assisted by a team of mechanical and
industrial engineers from the company on non regular basis. The engineers have
knowledge with the relevant machines and the processes. One engineer from the team
has involved in Six Sigma projects in her previous working experience. The student is
equipped with the fundamental knowledge in quality management including Six
Sigma. Trainings on general lean concepts have been provided to him. Lean concepts
have been extensively practiced in Company X, exemplified by complete process
cellularization at backend processes and active kaizen project cycles for more than five
years. Six Sigma tools including SPC are exclusive for quality department.
Each department needs to improve process along selected key performance
indicators annually. The company suffered losses of over USD 300,000 the previous
year on the component losses. The persistence of the problem is perceived as a good
platform for investigation and improvement by the student dedicated fully on the
project. The management has not set any budget limit for the improvement, despite any
proposal require additional purchase of instruments must be financial justifiable.
Based on the previous year figure, the management suggests a 10 percent reduction in
component losses in term of monetary value over the first three months of project
completion. This is translated into saving of USD 800 per month.
Any improvement proposed should be in agreement with lean concepts. Any
solution resulting larger batch size in production, loading parts on rule beside first
come first served (FCFS) or increase in non value added time is not permitted. System
with self-monitoring and right first time, such as Poka-yoke is encouraged.

6. Measure phase
6.1 Process flow
The production of all models is relatively similar, albeit with different electronic
components. Each model goes through nine operations, which begins at Station 1: reel
weighing station and ends at Station 9: visual inspection station. The process flow of the
components during the surface-mounting process is shown in the raw material process Reducing
flow (Figure 2). The three process charts of electronic components repeating loading, component
in-process and unloading are shown in Figure 3. At the outset, the electronic components
are transported from the warehouse to the production department after the line losses
supervisor calls for the build number, which is an artificial parent determined by
grouping similar models scheduled to run. Each build number consists of dozens
213
Transporting partial reels from Transporting reels from warehouse
production department to warehouse. to production department.
Warehouse

Reel Amount Station Indicator

Flow of requested reel

CheckingReel
partial reels quantity.
Amount . Flow of unused reel
Machine
Reel Weighing Machine
Stencil
rack
Waiting for checking unused
reels quantity. Waiting for checking received
reels quantity. Standby
Standby Area Area
Fire Hose
Waiting for setup. Visual Quantity produced was backflush after
Waiting for return to Check visual inspection.
Setup Station Table
warehouse.Feeder Trolley Conveyor
Rack Table Loading reels into feeders.
Lifter
Trolley Table
PC
AOI
Table Table Table Table
Unloading reels from
feeders.
Rack Setup Station Conveyor
Table
Trolley
PC Table

Trolley Table
Feeder
Table Table Table Table
Rack

Reflow Oven

Table Table Table Table Table Table Table

Stencil Conveyor
Rack
Load reels into machine.
Raw PCB Standby Area
Placement Machine 2

Fire Hose
Placement Machine 1

Unload unused reels from

Conveyor
machine.
AOI Solder Paste
Conveyor
Dry
Cabinet Laser Marking Printer

loader

Manual
Label Figure 2.
PC Material flow of electronic
components investigated
Note: The terms and machine used is explained in the subsequent paragraph
IJLSS Raw materials/reels
(calls according to
Loaded trolleys will be setup
into placement machine.

3,3 Trans-
portation
Build Number)
moves from Placement
Placement process will be
done by picking and placing Trans-
portation
The trolleys will be
sent back to setup
warehouse to Machine the electronic components station
production on the printed circuit board
department (PCB)

PCB placed with electronic

Reel weighing/ Scanning and components will be
Material weighting the raw transferred into reflow oven The raw materials/
214 Acceptance
Station
materials/reels to
check the quantity
Reflow Oven
by conveyor belt. Solder
paste between electronic
Setup Station reels will be
unloaded from the
and type of components and PCB will be feeders
components received heated and seal the electronic
components on the PCB
Raw materials/reels Unused raw
Standby
wait at standby area Inspection for defect on the materials wait to be
Area
before setup PCB using automatic optical Standby Area weighed for quantity
inspection machine. and returned to
AOI Station
Defected PCB will be sent to warehouse
visual inspection station
Operators load the
through conveyor belt Unused raw
raw materials/reels
into the feeders on materials are
Setup Station Reel weighing/
each table weighted and
Materials Return
according to loading At visual inspection station, scanned to return
Station
list inspection operator visually the partial reels to
inspects the pin pointed warehouse
Visual Inspection
defected area (done by AOI
inspection station) to justify
The loaded trolleys the defect area Unused raw
Trans- will be moved to Trans-
materials sent back
portation SME line standby portation
to warehouse
area

The finished product will be

The loaded trolleys stored at standby area
between Front End process
Figure 3. will standby at SMD
Storage line and Back End process
Standby line standby area
Process flow of electronic Area awaiting previous Area line

components in three Build Number to

complete production
different tasks
(a) For loading (b) For in-process (c) For unloading

to hundreds of reels based on the corresponding bill of materials. Each reel holds
hundreds to thousands of electronic components in rolls of perforated tape. Normally,
the electronic components are sent to the production line seven up to 13 times a day
(Figure 4).
Upon arriving at the material acceptance station, a batch of electronic components
goes through the reel weighing machine. Handled by an operator, the machine weighs,
labels, and records the quantity of components on the reels to verify the actual quantity
received with information stored in a database. The components are then sent to the
setup station and loaded manually into detached feeders. Some of the feeders are shared
by different build numbers, thus, the feeders, each supported by a trolley, are
commonly loaded at the maximum level. Upon completion of the task, the trolley will
be pushed to the surface-mounted equipment (SME) line. An SME line contains a series
of processes in a largely synchronous continuous flow. The processes include solder
pasting, automatic optical inspection (AOI) solder paste, placement of electronic
components, heating from reflow oven, and, finally, AOI and visual inspection. Belt
conveyors interlink these processes.
Upon request, e.g. changeover, the trolleys are fitted accordingly into the electronic
components placement machines, which pick and place electronic components onto
solder-pasted PCBs using a vacuum mechanism. Depending on the design, a PCB
usually has to undergo four stages of placement process before entering the reflow
oven. The reflow oven slowly heats the solder paste to form a solder joint at the
electronic component’s legs. After an interval of rapid cooling, AOI follows. The solder
 Cover tape Reducing
component
losses
Carrier tape
215

Embossed
cavity tape

Figure 4.
Reel and its components
Bar-code label

joints formed will be inspected visually by computer imaging equipment. If a defect is

detected, the defect area will be identified and later verified by the operator at the visual
inspection station. The PCBs failing the verification at this stage are to be reworked or
scrapped, depending on its condition. The number of PCBs failing at the verification
stage is insignificant. The quantity of the PCBs produced, including defective ones, is
backflushed into the material requirement planning system to update the balance of
the electronic components kept as inventory.
Any unused reel will be retrieved from the feeders after the process and sent to the
reel weighing machine once again for determining the remaining component quantity
and relabeling. Upon collection of the reels, they are returned to the warehouse. The
latest information about the quantities of electronic components will be used to update
the warehouse database and a scrap cost report will be generated. The scrap cost
report shows the quantity of components lost during the process, as well as the cost of
total components lost.

6.2 Tracing process

The percentage of losses at each operation station is traced. The tracing concentrates
on four models. They are the high runners that had the most component losses from
January 2010 until April 2010. Each model comprises hundreds of components. There
is a need to further confine the scope of the tracing to electronic components with the
most significant losses in term of cost. The entire flow of the selective electronic
components is tracked, that is, from the arrival of these electronic components as raw
materials in the production line until they are returned to the warehouse. Some of the
electronic components selected are tiny and alike in appearance. To avoid confusion,
detailed features of the components are sketched. Six build numbers are followed over
a period of three weeks, a process that is extremely time-consuming. A large amount
of electronic components are frequently collected over a few days of the production
IJLSS period for a single batch alone. These components have to be sorted to filter the
3,3 studied ones.
The tracing process is explained below:
.
The quantity of each type of raw materials is manually counted and recorded
when they arrive at the shop floor.
.
During the setup of the feeders, electronic components dropped onto the floor are
216 collected. They are categorized as loading drop off. Factors causing the drop off
are observed.
.
The electronic components thrown out by the placement machines are acquired
from the line supervisor who performs the daily collection. The collected
electronic components are categorized as machine thrown out.
.
During unloading at the placement machines, any drop out component is gathered.
Sorting is required. The components are categorized as unloading drop off.
.
Before the components are sent back to the warehouse, the selected components
are counted again manually.
.
Backflush reports are retrieved from the computer system and compared with
the backflush quantity manually recorded by the operator.
.
The log recording the components thrown out by the placement machine is
obtained from the system.

7. Analyze phase
In this stage, the data and information collected are analyzed. The root causes of the
problems are identified using a Ishikawa diagram (cause-and-effect diagram).

7.1 Results and discussion

For illustration, results from one model (Model A) will be used. One batch from Model
A runs over two consecutive shifts of production. On the first shift, the production
starts after a model changeover and with the remaining time 200 are produced; on the
second shift, 300 are produced. The information obtained is shown in Table I. The
actual physical losses collected are shown in Table II.
System stated total loss: USD 67.21 2 USD 24.60 ¼ USD 42.61.
The scrap cost report generated by the machine is not tallied with the actual physical
losses collected. The actual losses were much higher (USD 91.95 . USD 42.61). This
underlines the inconsistency of the reel weighing machine upon which the scrap cost
report is based. Gain is shown using the reel weighing machine, physically though, they
are losses.
The percentages of losses categorized by process are shown in Table III. The
percentages show that machine thrown out contributes to almost 30 percent of the losses.
The thrown out happens because of the identification error of the component during
machine inspection. Another 11 percent goes to the unloading component process.

7.2 Ishikawa diagram

The root causes are categorized as human, machine, method, and material, as displayed
in the Ishikawa diagram (Figure 5). People are the operators involved in handling the
reels throughout the surface-mounting process. Machine is the SME machine.
 Reducing
No. Component System stated gain (RM) System stated loss (RM)
component
1 Capacitor 0 27.86 losses
2 Conditioning IC 0 18.75
3 IC flash 0 4.88
4 IC 0 4.47
5 Transistor 57.06 0.00 217
6 Capacitor 0 3.46
7 IC 8.52 0.00
8 Quartz crystals 0 0.70
9 General purpose transistor 0.62 0.00
10 Volt regulator 10.65 0.00
11 Bus driver IC 0 118.56
12 Transmitter 0 3.19
13 OP amplifier 0 0.00
14 Low side driver 0 28.17 Table I.
Total 76.85 210.04 Scrap cost report

M/C thrown out Unloading drop off Other Total losses

No. Component (RM) (RM) (RM) (RM)

1 Capacitor 36.62 2.39 1.20 40.21

2 Conditioning IC 2.68 2.68 16.07 21.43
3 IC flash 0 0.00 7.32 7.32
4 IC 0 0.00 8.94 8.94
5 Transistor 0 5.52 3.68 9.20
6 Capacitor 0 0.00 10.39 10.39
7 IC 0 8.52 8.52 17.04
8 Quartz crystals 6.26 0.70 0.70 7.66
9 General purpose transistor 0 3.09 0 3.09
10 Volt regulator 0.66 2.00 0 2.66
11 Bus driver IC 30.40 3.04 85.12 118.56
12 Transmitter 0 1.06 4.26 5.32
13 OP amplifier 1.82 0.91 1.82 4.55
14 Low side driver 0 0.00 30.99 30.99
Total 78.44 29.90 178.99 287.33 Table II.
Actual physical losses
Note: Losses during loading, splicing, and unloading components are not collected collected

Process Percentage

Unloading component 11
Placement process 27 Table III.
Others 62 Percentage of losses
Total 100 categorized by process

Method is the procedure of loading and unloading electronic components. In this study,
there is exclusion of two other commonly present categories in a Ishikawa diagram,
namely environment and measurement. Electronics assembly is run in a controlled
environment of which electronic components understudied and process equipments
IJLSS Machine Human
3,3 Setup
Placement Malpractice
machine
Inadequate oversight
from supervisors
218 Reel weighing machine
Human errors during
data input Electronic
Components
Loading & unloading Quantity Losses

Backflush Quality

Splicing Adhesive
Figure 5.
Ishikawa diagram Method Material

are protected. Reel weighing machine can be potentially categorized in measurement.

Instead, it is to be placed under machine category because the tendency of the machine to
produce errors is known and primarily reflects its capability and technological limit.
Table IV summaries the losses. Figure 6 shows the Pareto chart the root causes based on
percentage of losses:
(1) Human:
.
Setup (C1). During setup, operators are observed to seldom pay attention,
especially in their haste, to the components that drop off. Improper tools such as
scissors and pens are used, leading to component damage or drop off.
In addition, the operators’ attention deteriorates when they are involved in a
casual conversation. The situation seemingly worsens during the night shift,
when a high number of component misplacements occur, possibly due to
exhaustion.

Main root cause Sub-root cause Percentage of losses

Human Setup (C1) 11

Malpractice (C2) 20a
Inadequate oversight from supervisors (C3) 10a
Human errors during data input (C4) 3a
Machine Placement machine (C5) 27
Reel weighing machine (C6) 3 (virtual losses)
Method Loading and unloading (C7) 11
Backflush (C8) 3a
Splicing (C9) 5a
Material Quantity (C10) 3a
Quality (C11) 2a
Adhesive C12) 2a
Table IV. Total 100
Percentage of losses
by each root cause Note: aBased on observation, data measure is found infeasible
 Reducing
component
losses

219

Figure 6.
Pareto chart based on
percentage of losses

.
Malpractice (C2). A policy states that any length of reel that drops off from
the machine must be scrapped. In this situation, the reel cannot be reused
because mistakes can easily be made in identifying the components the reel
is actually carrying. Nevertheless, few operators take advantage of this
policy to save the trouble of splicing.
.
Inadequate oversight from supervisors (C3). The priority, as perceived by
supervisors, is to ensure the continuous feeding of components at the SME
lines. In fact, an increase in component losses is seldom reprimanded by the
management.
.
Human errors during data input (C4). The quantity produced is written and
keyed in by operators at the end of AOI, thus, human error is likely.
Occasionally, the system is not updated to account for PCBs sent back from
the rework department.
(2) Machine:
.
Placement Machine (C5). Defect components are removed by machine as scrap
during the placement process. Maintenance is done when there are machine
errors, during the placement process. Machine errors, such as machine
conveyor stop or high components throw out rate, are repaired by a technician.
However, the technician merely solves the problem, not the root cause.
.
Reel weighing machine (C6). The reel weighing machine is only accurate
when the input information of the reel is correct. Any uninformed change
of input information, e.g. packaging material, affects the reading. Manual
counting is impractical with the time required.
IJLSS (3) Method:
3,3 .
Loading and unloading (C7). The loading standard operating process (SOP)
does not emphasize the precautionary steps aimed to prevent component
losses during handling. In addition, the corresponding monitoring is weak,
resulting in noncompliance amongst operators.
.
Backflush (C8). The balance on component usage is estimated, not exact. The
220 system writes off the planned quantity of components of one batch, rather than
the actual consumption of these components. The defective components and
the drop off during setup and placement process are, in fact, not considered.
.
Splicing (C9). The final length of a reel at the feeder is combined with a new
reel to avoid machine stoppage in SME line. Splicing requires small overlaps
of the two reels. Drop off during the handling is not uncommon. There are a
number of steps involved, including bar code scanning verification. Without
enforcement, some steps, e.g. bar code scanning verification, are either omitted
or improvised without authorization. This complicates the investigation
because the information in the system is not updated accordingly.
(4) Material:
.
Quantity (C10). The quantity of components stored in each reel depends on
component size, supplier, packaging material, and previous usage. If a model
requires 1,000 component A, and with a standard quantity of 800, two reels,
with a total of 1,600 components are needed. The balance of 600 components
will be returned after production.
. Quality (C11). The quality of components may differ. If quality is poor,
machine throw out rate will inevitably increase.
.
Adhesive (C12). The adhesive protective cover of the components is
designed to withstand only a small amount of pulling force, as machine
requirements stipulate. Applying force, such as manual handling, usually
exposes the adhesive and the associated components prematurely.

8. Improve phase
8.1 Proposed solutions
Each root cause was carefully studied by the team consists of the investigator and
company engineers, accompanying by verification through Gemba walks (on-field
studies) when skepticism arises. Eight solutions are proposed. Some were suggested by
company engineers during brainstorming sessions.
8.1.1 Develops new standard operating procedure (S1). A new SOP for the loading
and unloading process is developed at the setup area. The new SOP is presented in
dual language, English and Bahasa Malaysia (local language), to accommodate local
and foreign operators.
Awareness is raised through poster displays, in which dos and don’ts are given. For
instance, tweezers should be used, not scissors, during the setup. Training sessions
would be duly provided.
8.1.2 Kaizen team (S2). A cross-functional kaizen team would be formed, consisting
of the setup and the machine operators. One example of the kaizen team structure is
shown in Figure 7. Shift leaders are appointed as kaizen team leaders, who are overseen
 Reducing
Kitting Supervisor
component
losses

Shift Leader H
(Kaizen Team
Shift Leader I
(Kaizen Team
Shift Leader J
(Kaizen Team
221
Leader) Leader) Leader)

Setup Operator Setup Operator Setup Operator

(Kaizen Member) (Kaizen Member) (Kaizen Member)

Setup Operator Setup Operator Setup Operator

(Kaizen Member) (Kaizen Member) (Kaizen Member)

Machine Operator Machine Operator Machine Operator

(Kaizen Member) (Kaizen Member) (Kaizen Member)

Machine Operator Machine Operator Machine Operator

(Kaizen Member) (Kaizen Member) (Kaizen Member)
Figure 7.
Kaizen team structure
Kaizen Team H Kaizen Team I Kaizen Team J

by the kitting supervisor. Weekly component losses during shifts are selected as one of
the key performance measurements. This would motivate operators to improve their
individual performances. Table V shows the schedule and tasks to be performed by the
kaizen team.
8.1.3 Install close-circuit television (S3). Close-circuit television (CCTV) cameras are
installed in the setup area to act as deterrent and to monitor the operators. Footages
provide a form of evidence for any misconduct, ensuring accountability and
responsibility from the operators. The assistant executive (head of line supervisor) is in
charge of the CCTV footage and will provide review reports to the production manager
regularly.

Daily task Weekly task Monthly task

1. 5Ss checklist 1. Review the top 20 component 1. Implement the

losses and the total cost generated solution approved by
the production
manager
2. Record errors during operation 2. Gather in groups to brainstorm the 2. Review the results of
root cause of the losses and the implementation
suggest suitable solutions
3. Gather all defective products and 3. Propose selected solutions for Table V.
component drop offs consideration by production Kaizen team task
managers and schedule
IJLSS 8.1.4 Re-layout setup area (S4). The setup area is re-laid out. The feeder tables are
3,3 repositioned from a straight line to an L-shape. The materials accepting station is
added to the setup area to smooth out the material flow. Electronic components
received are directly loaded into the feeder or to the standby trolley. The IT department
will install equipment to determine whether a reel is to be loaded into the feeders or
placed on the standby trolley after scanning. The new version of standby reel will have
222 its own slots to facilitate further off-line setup. Figures 8 and 9, respectively, show the
current layout and future layout of the setup area.
8.1.5 Backflush procedure (S5). The machine server data are included to backflush
the used component quantity. Hence, human errors that occur during the backflushing
process will be minimized, the quantity of defective component could be identified,
and the operation would be more autonomous and consistent. Figure 10 shows the flow
of the backflush process.
8.1.6 Preventive maintenance 6(S6). Preventive maintenance for machines,
scheduled once a week and monitored by the maintenance engineer, should be done
until all the root causes have been solved completely. Machine performance should be
at the optimal level after each scheduled maintenance. All machine errors should be
eliminated before the production process starts; otherwise, the machine vendor would
be called to assist in solving the errors.
8.1.7 FIFO system (S7). The dependency on the reel weighing machine could be
reduced by finishing off as much partial reels as possible during production. If a fewer
number of reels are returned to the warehouse, then the reweighing frequency is
reduced. To ascertain conformance by the operators to this new directive, loading
priority is given to a partial reel, which requires bar code scanning for verification. By
doing so, as a spillover benefit, dead inventory is also reduced.
8.1.8 Numbering reel (S8). Numbering the reels may be viewed as the ultimate
solution. Each reel pocket is labeled with a number in ascending order. The last pocket,
therefore, reveals the actual quantity remaining on the roll. This makes weighing on
the reel weighing machine a redundant process. For this to be successful, cooperation
by the suppliers must be forthcoming. The incurring investment in the suppliers is
justifiable because their effort largely benefits the electronics assembly industry
as a whole.

8.2 Study of the solutions

As shown in Table VI, the proposed solutions are assessed against four criteria: capital
investment, planning requirement (time requirement), implementation feasibility, and
maintenance difficulty. Solutions 1, 2, 4, 5, and 6 are in line with the lean principles due
to its focus on continuous improvement and standard work practice based on people
attitude and eliminating waste. Solution 3 is not in line with the lean principles
because it eliminates trust and respect in the operators by monitoring them.

9. Control phase
9.1 Kaizen team
The regularity of kaizen meetings is monitored. The meetings are used as venue to
reinforce the kaizen concept, seven wastes, value stream mapping, importance of
maintaining work quality, value added, and nonvalue added activities to the team
members. Some common practices in identifying waste are constantly explained,
 Setup Station Reducing
Feeder Table component
Rack
losses
Trolley
Table

PC
Trolley 223
Table

Trolley
Table

Trolley
Feeder
Table
Rack

Trolley
Table

Trolley
Table

Setup Station
Table

Trolley
Table

Trolley
Table

Trolley
Table

Trolley
Table

Trolley
Table

Trolley Figure 8.
Table Old setup station shop
floor layout

such as the observation method and quantifying the waste into the same standard
for measurement purposes. Aside from these, the brainstorming process is held to
identify potential improvements that may be carried out, e.g. changes on task
schedules.
IJLSS CCTV
3,3
Setup Station
Feeder Table
Rack
224 Trolley
Table

PC Table

Trolley
Table

Feeder

Table

Table

Table
Rack

Setup Station
Feeder
Rack Table

Trolley
Table

PC Table

Trolley
Table

Feeder
Table

Table

Table

Rack

Figure 9.
New setup station shop
Table

Table

Table

Table

Table

Table

Table

floor layout with CCTV

location

9.2 Performance audit

Certified operators are given the task to load and unload the reel according to the new
SOP. Their performances are to be monitored through individual performance audit by
an industrial engineering team also responsible for 5S audit in the production. The audit
is run at random interval, twice per shift for the first month. Upon stable, the frequency
will be reduced by stages to the normal rate of three times per month. Marks are given
on the spot following the marking scheme on the evaluation form. An example of
 Quantity of thrown Reducing
Assemble process
start
Placement Machine out component/ component
defect component
losses
Acquire the data
Quantity of product
Computer server after assemble
produced
process finish
225

Backflush the Figure 10.

components used Backflush process flow

the evaluation form is shown in Figure 11. The actual content of the form breaks into two
sections of each listing critical check points for procedure conformance. The reason of
any non-conformance of interest is collected by the additional follow-up questions, as
part of the continuous improvement process. The mark will attribute to the team’s 5S
performance score.

9.3 Results of implemented solutions

The weekly component losses are collected. Comparing the average weekly losses from
weeks 1 to 25 with weeks 26 to 41, losses were reduced from USD 7,680 to USD 6,400 or
about 18 percent. As shown in Figure 12, the imposed linear trend line onto the graph
reveals an aggregate of decreasing component losses.
In order to eliminate fluctuation in the raw materials inventory data, it is best to
implement the numbering system on all the reels. To ensure the sustainability of the
implemented solutions, the company is strongly recommended to form a kaizen team,
as stated in the control phase. Without further control and continuous improvement,
component losses cannot be reduced and maintained at a more satisfactory level.
However, the root cause of quality, quantity, and adhesive are still not resolved. The
company should cooperate with the suppliers of the electronic components to identify a
way to overcome these issues.
The positive result is promulgated to the whole organization and produces different
effects to individual levels of the organization. To the bottom level, operators are
shown that the minute activities performed daily have direct and cumulated impacts
on the production. Sufficient discipline control and continuous inculcation of
ownership among operators for the involved processes including the setup are
necessary. Enforcement of policies and frequent audit are importance in conjunction to
regular awareness training. For the middle and the top levels, the joint implementation
of the project has demonstrated by combining lean and Six Sigma is effective to
overcome the limitations of each program when implemented in isolation. This form of
project model is particularly suitable for exploratory projects where problems have not
been properly defined and perceived complex. Additionally, lean concepts traditionally
act as the backbone of any improvement initiative run in the company. Lean tools for
example value stream mapping are heavily employed, resulting in a streamlined
processing path of critical processes (ones that affect manufacturing lead time).
This practice has in hindsight sidelined two improvement opportunities. First,
processes on the subsidiary paths, e.g. off-line preparation steps are less improved.
 3,3

226
IJLSS

Table VI.
Evaluation of solutions
Cost
Implementation Causes solved
Solution Planning requirement feasibility Maintenance difficulty C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12

S1 Two weeks of Needs to ensure the new Requires training for all Requires time to time * * * * *
production engineer SOP is justified and setup operators monitoring
time to prepare the SOP accepted
S2 One week of production Cooperation from the Explanation and Self-motivation within * * * *
engineer time for management level to training of the kaizen the kaizen team member
conducting the training operators is essential concept is needed
course
S3 CCTV system cost and Suitable location for Personnel monitoring Willingness to * * * *
installation cost CCTV camera the footage recorded continuously review the
needs to be chosen footage recorded
wisely
S4 Five weeks of Requires justification of Operators require time Debugging required for * *
production engineer and the effectiveness of the to endure the changes in the software
software engineer time new layout and error the shop floor
for changing and checking in the software
preparing the layout written
software needed
S5 Four weeks of Checking accuracy of Train operators to Computer server needs * *
production engineer and server data acquire data from to be protected from
software engineer time server viruses and
breakdowns
S6 Require extra working Requires scheduling for Cooperation from the Management needs to *
hour from maintenance preventive maintenance maintenance ensure preventive
engineer and time slot department maintenance is running
technicians accordingly from time
to time
S7 Take four weeks of Needs to design the Train operators to use Authority is needed to *
software engineer and FIFO system flow the FIFO system ensure the operators
production engineer pattern follow the system all the
time time
S8 Patent cost Patenting the Request approval from Cooperation from all the *
numbering reel method the management using suppliers to use the
the patented method patent
 Certified Setup Operator Evaluation Form Reducing
Employee Name : Date : component
Employee Number : Score :
losses
24
Loading 227
Q1. Does all the 6 SOP steps being followed by the operator? Followed ___________/6 Steps

Q1a. Which step is not followed? Step _______

Q2. Does “Tweezer” used by the operator during the setup? Yes/No

Q2a. If Q2 answer is No, what tool was used?

Q3. Does the empty pocket have been positioned at the pickup window Yes/No

for all reels?

Q3a. If Q3 answer is No, state the reason.

_____________________________________________________________________

_____________________________________________________________________

Unloading

Q1. Does all the 9 SOP steps being followed by the operator? Followed___________/9 Steps

Q1a. Which step is not followed? Step _______

Figure 11.
Q2 Does the empty pocket length left 9 cm? Yes/No Example of certified setup
Yes/No operator evaluation form
Q3 Does the plastic length left 45 cm?

17,045
Start of S1 and S4 implementation
14,205
Losses (USD)

11,364

8,523

5,682

2,841

0
Week1
Week2
Week3
Week4
Week5
Week6
Week7
Week8
Week9
Week10
Week11
Week12
Week13
Week14
Week15
Week16
Week17
Week18
Week19
Week20
Week21
Week22
Week23
Week24
Week25
Week26
Week27
Week28
Week29
Week30
Week31
Week32
Week33
Week34
Week35
Week36
Week37
Week38
Week39
Week40
Week41

Figure 12.
Week Graph of total losses from
weeks 1 to 41 (2010)
Total losses Linear (Total losses)
IJLSS Many problems occurred on these areas have not been studied as intensely as the
3,3 critical processes. Second, variability and its impact on costing and management are
largely overlooked. The project reveals that these improvement opportunities are
significant particularly on cost savings and the undertakings are likely to be highly
rewarded.

228 10. Conclusions

The current paper presented a systematical approach in solving electronic component
losses issue in an electronics assembly company. DMAIC is a well-established
methodology in reducing variability in a process. In the improve phase, the shop floor
layout of the setup station was changed, and new loading and unloading procedures
were adopted by the setup operators. From the results shown, the solutions
successfully reduced component losses to a lower level. Few key lessons were learned.
First, for business excellence, variability in all forms must be continuously detected
and limited. It is best tackled in an independent project-based improvement system
with a dedicated team and budgets. Therefore, second, it is important to build a critical
mass for these types of projects by starting to train suitable staff and to create
awareness by demonstrating success cases regularly. Third, DMAIC methodology
enhances problem study. The methodology comprises an array of tools to facilitate
comprehensive data collection and analysis. Resulted from the analysis, 12 root causes
and their significances were determined in this case study. This was instrumental to
compose appropriate set of strategies at the later phase. The distinctive control phase
matched the need of this project because the inclusion of performance audit was
necessary to ensure conformance of manual operations to the standards. Finally, the
success introduction of a Six Sigma tool into lean manufacturing provides confidence
to management to be receptive to alternative improvement methods.

References
Antony, J., Kumar, M. and Tiwari, M.K. (2005), “An application of Six Sigma methodology to
reduce the engine-overheating problem in an automotive company”, Proceedings of the
Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, Vol. 219
No. 8, pp. 633-46.
Bertels, T. (2003), Rath and Strong’s Six Sigma Leadership Handbook, Wiley, Hoboken, NJ.
Breyfogle, F.W. III (2010), “Process improvement projects shortcomings and resolution”,
International Journal of Lean Six Sigma, Vol. 1 No. 2, pp. 92-8.
Cuatrecasas Arbós, L. (2002), “Design of a rapid response and high efficiency service by lean
production principles: methodology and evaluation of variability of performance”,
International Journal of Production Economics, Vol. 80 No. 2, pp. 169-83.
El-Haik, B.S. and Shaout, A. (2010), Software Design for Six Sigma: A Roadmap for Excellence,
Wiley, Hoboken, NJ.
Goldsby, T. and Martichenko, R. (2005), Lean Six Sigma Logistics, J. Ross Publishing,
Boca Raton, FL.
He, X.F., Wu, S. and Li, Q.L. (2007), “Production variability of production lines”, International
Journal of Production Economics, Vol. 107 No. 1, pp. 78-87.
Heizer, J. and Render, B. (2008), Operations Management, 9th ed., Prentice-Hall,
Upper Saddle River, NJ, p. 643.
Jang, G.S. and Jeon, J.H. (2009), “A Six Sigma methodology using data mining: a case study on Reducing
Six Sigma project for heat efficiency improvement of a hot stove system in a Korean steel
manufacturing company”, Communications in Computer and Information Science, Vol. 35
component
No. 3, pp. 72-80. losses
Jing, W.L. (2003), “Improving the performance of job shop manufacturing with demand-pull
production control by reducing set-up/processing time variability”, International Journal
of Production Economics, Vol. 84 No. 3, pp. 255-70. 229
Kaushik, P. and Khanduja, D. (2009), “Application of Six Sigma DMAIC methodology in thermal
power plants: a case study”, Total Quality Management & Business Excellence, Vol. 20
No. 2, pp. 197-207.
Knowles, G., Whicker, L., Femat, J.H. and Canales, F.D.C. (2005), “A conceptual model for the
application of Six Sigma methodologies to supply chain improvement”, International
Journal of Logistics Research and Applications, Vol. 8 No. 1, pp. 51-65.
Kumar, U., Crocker, J., Chitra, T. and Saranga, H. (2006), Reliability and Six Sigma, Springer,
Berlin.
Leen, G. and Heffernan, D. (2002), “Expanding automotive electronic systems”, Computer, Vol. 35
No. 1, pp. 88-93.
Mapes, J., Szwejczewski, M. and New, C. (2000), “Process variability and its effect on plant
performance”, International Journal of Operations & Production Management, Vol. 20
No. 7, pp. 792-808.
Motley, B. (2005), “Introduction to variability and variation reduction”, Defense AT&L Magazine,
May/June, pp. 53-5.
Pan, Z., Ryu, H. and Baik, J. (2007), “A case study: CRM adoption success factor analysis and
Six Sigma DMAIC application”, Proceedings of 5th ACIS International Conference on
Software Engineering Research, Management & Applications, IEEE Computer Society,
Busan, pp. 828-38.
Romero Rojo, F.J., Roy, R. and Shehab, E. (2009), “A methodology for variability reduction in
manufacturing cost estimating in the automotive industry based on design features”,
Proceedings of the 19th CIRP Design Conference – Competitive Design, Cranfield
University, Cranfield, pp. 197-202.
Sangiovanni-Vincentelli, A. (2003), “Electronic-system design in the automobile industry”,
Micro, IEEE, Vol. 23 No. 3, pp. 8-18.
Sharma, S. and Chetiya, A.R. (2010), “Six Sigma project selection: an analysis of responsible
factors”, International Journal of Lean Six Sigma, Vol. 1 No. 4, pp. 280-92.
Shetty, D., Ali, A. and Cummings, R. (2010), “Survey-based spreadsheet model on lean
implementation”, International Journal of Lean Six Sigma, Vol. 1 No. 4, pp. 310-34.
Summers, D.C.S. (2006), Quality, 4th ed., Pearson Education, Upper Saddle River, NJ.
Tan, B. (1998), “Effects of variability on the due-time performance of a continuous materials flow
production system in series”, International Journal of Production Economics, Vol. 54 No. 1,
pp. 87-100.
Tong, J.P.C., Tsung, F. and Yen, B.P.C. (2004), “A DMAIC approach for printed circuit board
quality improvement”, International Journal of Advanced Manufacturing Technology,
Vol. 23 Nos 7/8, pp. 523-31.
Xu, J. (1996), “Variability detection and robustness design in complex production system”,
Computers & Industrial Engineering, Vol. 31 Nos 3/4, pp. 775-8.
IJLSS About the authors
Tan Ping Yi is a Graduate Student from School of Mechanical Engineering, Universiti Sains
3,3 Malaysia. Part of the work presented here belongs to her final year project.
Chin Jeng Feng obtained his Doctor of Philosophy (PhD) from University of Birmingham, UK
and currently is a Senior Lecturer in the School of Mechanical Engineering, Universiti Sains
Malaysia. The research areas of interest include production planning and control, lean
manufacturing and intelligent manufacturing system. Chin Jeng Feng is the corresponding
230 author and can be contacted at: [email protected]
Joshua Prakash is a Postgraduate Student from School of Mechanical Engineering, Universiti
Sains Malaysia undertaking a Doctor of Philosophy (PhD) in Mechanical Engineering majoring
in Lean Manufacturing. He received his Bachelor Degree (BSc) in Manufacturing Engineering
with Management from School of Mechanical Engineering, Universiti Sains Malaysia.
Loh Wei Ping obtained his Doctor of Philosophy (PhD) from Universiti Sains Malaysia,
Malaysia and currently is a Senior Lecturer in the School of Mechanical Engineering, Universiti
Sains Malaysia. The research areas of interest include data mining, mathematical modeling and
applied biomechanics.

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints