SUPPLIER DEFECT DATA ANALYTICS IN LEAN SUPPLY CHAIN … · al.(2008) worked on responsive supply chain with a competitive strategy in a networked economy. Supply chain management

JOURNAL OF INTERNATIONAL ACADEMIC RESEARCH FOR MULTIDISCIPLINARY Impact Factor 2.417, ISSN: 2320-5083, Volume 4, Issue 4, May 2016

30 www.jiarm.com

SUPPLIER DEFECT DATA ANALYTICS IN LEAN SUPPLY CHAIN WITH A CONTINUOUS IMPROVEMENT APPROACH

OZLEM AKÇAY KASAPOĞLU, PHD*

*Dept. of Operations Management, Faculty of Business Administration, Istanbul University, Turkey

ABSTRACT

In today’s environment supply chains are in the conflict of being responsive and being

efficient. It is not an easy task to be responsive while being efficient. In the efficiency efforts

defect reduction is crucial. Defects are inevitable but their impact and the sources of them

should be determined so that their impact on the processes could be decreased. To be able to

do this, process engineers need to implement a defect management process that focuses on

preventing defects, catching defects as early in the processes as possible, and minimizing the

impact of defects, can yield significant returns. In today’s’ global world with the huge supply

chains this could not be done without the help of our suppliers. Suppliers with a long term

commitment and relationships help the factories in their process improvement and in cleaning

the defects. Once the defects are identified then the related suppliers are identified. Then the

quality improvement efforts could be done in the first tier supplier and the other suppliers

upstream. Once there is a belief and knowledge on lean manufacturing in the supply chain the

lean supply chain would be the naturally happy result. This paper provides a framework for

the discussion about lean understanding and continuous improvement, how defect analysis

initiates linked to the data analytics are broadening to enable improved lean supply chain,

how decision trees as data mining tools support the decision on the branches of materials and

their related suppliers that should be concentrated on.

KEYWORDS: Tqm, Defect, Continuous Improvement, Lean, Data Mining

INTRODUCTION

The best and the most direct route to world class competitiveness for the products and

services is through defect reduction on the processes. If the defect rates decreases, yields are

improved, rework is reduced, warranty costs and the difference between customer needs and

process performance decreases so the market potential increases. Productivity gains through

continuous process improvement can contribute to profitability by supporting both revenue

growth and cost reduction. In this context, leading organizations increasingly emphasize


31 www.jiarm.com

continuous process improvement through “data driven decision-making” to achieve these

benefits.

In manufacturing companies there are big databases, various kinds of data are

sometimes unorganized. If those data bases are not strategically managed, they stay as

meaningless junk data not capable of producing knowledge. Some companies are

considering quality as a cost, not collecting and not analyzing data and they are not capable

of producing information for the continuous improvement purposes. Continuous

improvement understanding must be supported by management and should be in companies’

strategies. A strategy may be defined as a plan or method for obtaining some goal or result.

Unlike other quality initiatives, continuous improvement has a strategic component aimed at

not only developing management’s commitment and their active involvement. The steps

involved in creating this strategy include identifying the key processes that affect the strategic

business objectives of the organization highest impact processes are then targeted for

improvement. Data mining tools are applied to the manufacturing data for the define stage of

the quality efforts and to able extract knowledge to able to build up a strategy for the quality.

Defects are considered as the one of the seven wastes in lean manufacturing and an element

in a hidden factory concept. Continuous improvement on our defects should be considered

after the detection of them. Continuous improvement efforts are common both in TQM and

lean manufacturing, also in 6 sigma practices. We can call it Kaizen, we can call it DMAIC

the concept should be with us in all our journeys in production. Defect detection starts with

the discovery of the defects which includes the implementation techniques, methodology and

standard processes to reduce the risk of defects. A defect is only discovered when it has been

documented and acknowledged. Identification and analysis of the process in which a defect

originated to identify ways to improve the process to prevent future occurrences of similar

defects. The next step should be process improvement. Defect prevention should begin with

an assessment of the critical risks associated with the system. Getting the critical risks

defined allows people to know the types of defects that are most likely to occur and the ones

that can have the greatest system impact. Strategies can then be developed to prevent them.

The major steps for defect prevention are as follows: Identify the critical risks facing the

project or system. These are the types of defects that could jeopardize the successful

construction, delivery and/or operation of the system. For each critical risk, make an

assessment of the financial impact if the risk becomes a problem. Once the most important

risks are identified try to eliminate each risk. For risks that cannot be eliminated, reduce the


32 www.jiarm.com

probability that the risk will become a problem and the financial impact should that happen.

( Defect management, 2016)

Lean supply chain management represents a new way of thinking about supplier

networks. Lean principles require cooperative supplier relationships while balancing

cooperation and competition. Cooperation involves a spectrum of collaborative relationships

& coordination mechanisms. Supplier partnerships & strategic alliances represent a key

feature of lean supply chain management. Lean Supply Chain Management principles are

driven from basic lean principles, they focus on the supplier network value stream The seven

important lean principles are: Eliminate waste, synchronize flow, minimize both transaction

and production costs, establish collaborative relationships while balancing cooperation and

competition, ensure visibility and transparency, develop quick response capability, manage

uncertainty and risk, align core competencies and complementary capabilities and foster

innovation and knowledge-sharing.

Synchronized production and delivery throughout the supplier network is a central

lean concept. Important concepts that should be considered are: Integrated supplier lead times

and delivery schedules, flows from suppliers pulled by customer demand (using takt time,

load leveling, line balancing, single piece flow), minimized inventory through all tiers of the

supply chain, on-time supplier delivery to point of use, minimal source or incoming

inspection, effective two-way communication links to coordinate production & delivery

schedules, greater efficiency and profitability throughout the supplier network and striving

for zero quality defects essential to success. (Nightingale,2016)

Literature review

There are studies that are done on lean and responsive supply chains. Jurado and

Fuentes (2014), studied lean management, supply chain management and sustainability. You

and Grossmann, (2008) worked on the design of responsive supply chains under demand

uncertainty. Cox and Chicksand (2005), studied the limits of lean management thinking.

Folinas et al., (2013), explored the greening of the food supply chain with lean thinking

techniques. Roh, et al.(2013) worked on the implementation of a responsive supply chain

strategy in global complexity with the case of manufacturing firms. They mentioned that

although a responsive supply chain is an integral part of order-winning manufacturing

strategies, it has not been clear how firms build a responsive supply chain in global

manufacturing environments. Firms also need to provide key implementation practices


33 www.jiarm.com

sharing of information with customers, collaboration with suppliers, use of advanced

manufacturing technology to achieve pull production to achieve responsiveness to the market.

Purvis et al., (2014) studied the development of a lean, agile and leagile supply network

taxonomy based on differing types of flexibility. The paper explores the meaning of

flexibility in the context of lean, agile and leagile supply networks and articulates a supply

network flexibility framework. Two key ‘sources’ of flexibility are investigated: vendor

flexibility and sourcing flexibility. The paper introduces an extension of the ‘leagility’

concept beyond the simple material flow decoupling point concept. Two new types of

leagility are put forward: (1) leagile with vendor flexibility systems, which combine the use

of agile vendors with lean sourcing practices and (2) leagile with sourcing flexibility systems,

which combine the use of lean vendors with agile sourcing practices. Naylor et al.(1999)

studied leagility also, they integrated the lean and agile manufacturing paradigms in the total

supply chain. As the lean thinking and agile manufacturing paradigms have been developed

there has been a tendency to view them in a progression and in isolation. Gunasekaran et

al.(2008) worked on responsive supply chain with a competitive strategy in a networked

economy. Supply chain management (SCM) has been considered as the most popular

operations strategy for improving organizational competitiveness in the twenty-first century.

In the early 1990s, agile manufacturing (AM) gained momentum and received due attention

from both researchers and practitioners. In the mid-1990s, SCM began to attract interest.

Dües et al (2013) studied green as the new Lean: how to use Lean practices as a catalyst to

greening your supply chain. Chen, and Cheng (2013), studied Supply chain management with

lean production and RFID application they did a case study in expert systems with

applications. Mohammaddust et al.(2015), developed lean and responsive supply chains.

They did a robust model for alternative risk mitigation strategies in supply chain designs.

Lemmens et al., (2016) made a review of integrated supply chain network design models and

worked on key issues for vaccine supply chains.

There are studies done on the lean manufacturing and continuous improvement.

Behrouzi, et al., (2010) studied lean performance evaluation of manufacturing systems with a

dynamic and innovative approach. Many organizations around the world have attempted to

implement it but the lack of a clear understanding of lean performance and its measurement

will contribute to the failure of lean practices. This paper presents an innovative approach to

measure the lean performance of manufacturing systems by using fuzzy membership

functions. Kull et al (2014), worked on the moderation of lean manufacturing effectiveness


34 www.jiarm.com

by dimensions of national culture: Testing practice-culture congruence hypotheses. The

successful use of lean manufacturing (LM) practices requires more than the use of tools.

Although manufacturing facilities worldwide use LM practices, dimensions of a nation’s

culture may moderate LM׳s effect on operating performance. Yang et al. (2014), worked on

the lean production system design for fishing net manufacturing using lean principles and

simulation optimization. Büyüközkan et al. (2015), studied the assessment of lean

manufacturing effect on business performance using bayesian belief networks. Fullerton et

al., (2014) studied lean manufacturing and firm performance with the incremental

contribution of lean management accounting practices. Manufacturing firms operating in

rapidly changing and highly competitive markets have embraced the continuous process

improvement mindset. They have worked to improve quality, flexibility, and customer

response time using the principles of Lean thinking. To reach its potential, lean must be

adopted as a holistic business strategy, rather than an activity isolated in operations. The lean

enterprise calls for the integration of lean practices across operations and other business

functions. Arslankaya. et al. (2015), worked on the maintenance management and lean

manufacturing practices in a firm which produces dairy products. This study described the

implementation of maintenance management and lean manufacturing techniques at the

maintenance workshop in order to eliminate the losses due to breakdowns and to enhance

productivity and the motivation of employees at an enterprise which produced dairy products.

Oleghe et al. (2016) studied on the variation modeling of lean manufacturing performance

using fuzzy logic based quantitative lean index. Nelson (2016) studied pull versus push:

lessons from lean manufacturing, ın becoming a lean library. Starting in the 1940s and

continuing to the present, the Japanese automotive industry, led by Toyota, created a new

method of manufacturing based on a demand-driven, or “pull,” model. In this model

inventory is not stored in large warehouses or parking lots but supplied when a new car or

product is ordered by downstream consumers. This method is contrasted with the traditional,

or “push,” method of manufacturing where products are constructed on an assembly line

without regard for the upstream demand by the end user or consumer. A pull model, in

contrast, depends on accurate and timely metrics about what materials or services are

necessary to deliver the end product with minimal wastage in the manufacturing workflow.

Indrawati and Ridwansyah, (2015) studied manufacturing continuous improvement using

lean six sigma: with an Iron Ores Industry Case Application. In order to improve the

manufacturing process capability, this research is conducted using lean six sigma method.


35 www.jiarm.com

The first part is focused on waste analysis using process activity mapping. Then

manufacturing process capability is evaluated. Further, failure mode and effect analysis is

used as a basic consideration in developing the continuous improvement program. Based on

the analysis, product defects, inappropriate processing and waiting are type of manufacturing

waste that frequently occurs. A continuous improvement program is develop to overcome

that problem.

Hartin and Ciptomulyono (2015) examined the Relationship between Lean and

Sustainable Manufacturing on Performance. The aim of this paper is to explore and evaluate

previous work focusing on the relationship and links between Lean and sustainable

manufacturing. This paper also examines impact of lean and sustainable manufacturing to

improve performance. Many evidences suggested that Lean is beneficial for Sustainable

manufacturing, dominantly on perspective environment and economic aspect. This paper

identify major research gaps for integrated lean and sustainable manufacturing to improve

performance business and modeling as a methodology approach. This paper provides a

quantitative descriptive analysis and qualitative thematic analysis to provide an analysis of

relationship lean and sustainable manufacturing and its impact on performance.

Choomlucksana and Ongsaranakorn (2015) improved the productivity of sheet metal

stamping subassembly area using the application of lean manufacturing principles.

Companies frequently find techniques and tools to enhance productivity and quality for

success in the long-term in order to maximize competitive advantage. To date, lean

manufacturing principle is one of the successful improvement concepts that have been

applied to eliminate waste and non-value added activities that occur in many companies. This

paper explores a real work case study of the manufacturing sheet metal stamping process to

demonstrate how lean manufacturing can help improve work efficiency. Lean and other

improvement tools and techniques such as visual control, Poka-Yoke, and 5s were applied to

help companies identify areas of opportunity for waste reduction and improve the efficiency

of production processes. There are research done on continous improvement and TQM.

Petruţa and Roxana (2014), Integrated six sigma with quality management systems for the

development and continuous improvement of higher education institutions. Katzman and

Paushter (2016) studied on the building a culture of continuous quality improvement in an

academic radiology department.

In the literature there are studies on the defect management and analysis. S. Ferretti, et

al., (2013), monitored systems for zero defect manufacturing. Marcello et al., (2014), worked


36 www.jiarm.com

on the design and management of manufacturing systems for production quality. They

believed that manufacturing companies are continuously facing the challenge of operating

their manufacturing processes and systems in order to deliver the required production rates of

high quality products, while minimizing the use of resources. Production quality is proposed

in this paper as a new paradigm aiming at going beyond traditional six-sigma approaches.

Ahmad, et al., (2012) mentioned that in today’ s highly competitive market, the demand for

quality is the single most critical factor for companies to survive in the ever-expanding global

marketplace. The purpose of their paper is to propose relationship between TQM practices

and business performance with mediators of Statistical Process Control (SPC), Lean

Production (LP) and Total Productive Maintenance (TPM) based on extensive review of the

literature. David (2005) studied on the defect analysis and basic techniques for management

and learning. Defect analysis is defined as the study of the properties of defects themselves,

as opposed to methods for predicting the number and nature of defects based on other

properties of software, such as complexity models. Sarkar and Saren (2016), made a product

inspection policy for an imperfect production system with inspection errors and warranty

cost. During manufacturing of products, all produced items are considered as perfect in

general. This viewpoint of taking all finished products are perfect is not correct always.

Defective items may occur during the production process for several reasons. Their paper

describes a deteriorating production process which randomly shifts to out-of-control state

from in-control state. In case of full inspection policy, expected total cost together with

inspection cost results higher inventory cost. Therefore, product inspection policy is better to

use for reducing inspection costs. During product inspection process, inspectors may choose

falsely a defective item as non-defective and vice-versa. Type I and Type II errors are

incorporated in this model to make more realistic rather than existing models. Baykasoglu et

al. (2011) classified defect factors in fabric production via Difaconn-miner, they did a case

study on expert systems with applications. In their paper a data mining based case study is

carried out in a major textile company in Turkey in order to classify and analyze the defect

factors in their fabric production process. It is aimed to understand the causes of the defects

in order to minimize their occurrence. The main motivation behind this study is to minimize

scrap loses in the company and enabling more sustainable production via data mining. Dhafr

et al.(2006) tried to improve the quality performance in manufacturing organizations by

minimization of production defects. Their methodology comprises a model for the

identification of various sources of quality defects on the product; this model would include


37 www.jiarm.com

an analysis tool in order to calculate defect probability, a statistical measurement of quality,

and a lean manufacturing tool to prevent the presence of defects on the product. The

attribution of defects to their source will lead to a fast and significant definition of the root

cause of defects. The techniques described in this paper were developed for an improvement

project in a plastic parts painting manufacturing facility of a first-tier supplier to the

automotive industry. Perner, et al.,(2016) tried to avoid defects in manufacturing processes,

they did a review for automated CFRP production, robotics and computer-integrated

manufacturing. Linear robots, numerically controlled machines and articulated robots are

some of the major components in automation. In various applications, these single systems

are linked together to become advanced mechatronic systems since they can realize special

and general tasks.

Data mining and the decision trees

Data mining has attracted a great deal of attention in the information industry and in

society as a whole in recent years, due to the wide availability of huge amounts of data and

the need for turning such data into useful information and knowledge. The information and

knowledge gained can be used for applications ranging from market analysis, fraud detection

and customer retention, to production control and science exploration (Han J., Kamber M.,

2006). Data mining can be viewed as a result of the natural evolution of information

technology. The database system industry has witnessed an evolutionary path in the

development of the following functionalities: data collection and database creation, data

management and advanced data analysis. Data mining is an integral part of knowledge

discovery in databases, which is the overall process of converting raw data into useful

information. This process consists of a series of transformation steps, from data preprocessing

to post processing of data mining results (Tan P.N. et al., 2006). Data mining is widely used

in diverse areas. There are a number of commercial data mining systems available today and

yet there are many challenges in this field. (data mining, 2016). Data mining is widely used

for: Financial data analysis, retail industry, telecommunication industry, logistics sector,

biological data analysis, other scientific supplications, intrusion detection. With the rapid

development of database technology and the wide application of database management

system, data mining is more and more used in logistics enterprise. It is a key problem how to

use data mining with low cost and high efficiency (Fei Z., et al. 2010).


38 www.jiarm.com

Databases are rich with hidden information that can be used for intelligent decision making.

Classification and prediction are two forms of data analysis that can be used to extract models

describing important data classes or to predict future data trends. Many classification and

prediction methods have been proposed by researchers in machine learning, pattern

recognition, and statistics. Most algorithms are memory resident, typically assuming a small

data size. Recent data mining research has built on such work, developing scalable

classification and prediction techniques capable of handling large disk- resident data.

Classification and prediction have numerous applications including fraud detection, target

marketing, performance prediction, manufacturing, and medical diagnosis. (Han, Kamber,

p.285). Decision tree is one of the most widely used techniques for classification which is the

task of assigning objects to one of several predefined categories. In practice, one wants to

have a small and accurate tree for many reasons. A smaller tree is more general and also

tends to be more accurate. It is also easier to understand by human users. In many

applications, the user understanding of the classifier is important (Liu, p. 59-60). The

Decision Tree procedure creates a tree-based classification model. It classifies cases into

groups or predicts values of a dependent (target) variable based on values of independent

(predictor) variables. The procedure provides validation tools for exploratory and

confirmatory classification analysis. The procedure can be used for:

Segmentation: Identify persons who are likely to be members of a particular group.

Stratification: Assign cases into one of several categories, such as high-, medium-, and low-

risk groups.

Prediction: Create rules and use them to predict future events, such as the likelihood that

someone will default on a loan or the potential resale value of a vehicle or home.

Data reduction and variable screening: Select a useful subset of predictors from a large set of

variables for use in building a formal parametric model.

Interaction identification: Identify relationships that pertain only to specific subgroups and

specify these in a formal parametric model.

Category merging and discretizing continuous variables: Recode group predictor categories

and continuous variables with minimal loss of information.

There are exponentially many decision trees that can be constructed from a given set

of attributes. While some of the trees are more accurate than others, finding the optimal tree

is computationally infeasible because of the exponential size of the search space.

Nevertheless, efficient algorithms have been developed to induce a reasonably accurate,


39 www.jiarm.com

albeit suboptimal, decision tree in a reasonable amount of time. These algorithms usually

employ a greedy strategy that grows a decision tree by making a series of locally optimum

decisions about which attribute to use for partitioning the data. ID3, C4.5, C&RT, QUEST

are examples for the decision tree algorithms. (Tan, Steinbach, Kumar, p.151-152). There are

growing methods in that field:

Chaid: Chi-squared Automatic Interaction Detection. At each step, CHAID chooses the

independent (predictor) variable that has the strongest interaction with the dependent

variable. Categories of each predictor are merged if they are not significantly different with

respect to the dependent variable.

C&RT: Classification and Regression Trees. C&RT splits the data into segments that are as

homogeneous as possible with respect to the dependent variable. A terminal node in which all

cases have the same value for the dependent variable is a homogeneous, "pure" node.

Quest: quick, unbiased, efficient statistical tree. A method that is fast and avoids other

methods' bias in favor of predictors with many categories. Quest can be specified only if the

dependent variable is nominal. (IBM, p.1)

Application

Problem definition

In the analysis the distribution of the defects and the sources of those defects are

wanted to be analyzed. On the number of defects the material type, the production month and

the day, the shift that is produced and the controller is considered for their effects. The

material types and their suppliers are so important, it is effected by the supplier selection and

procumbent decisions. In today’s supply chain understanding the suppliers are considered as

partners. So before changing the supplier the improvement studies on the suppliers

production processes should be done. One of the wastes in lean manufacturing is defects and

those defects are consequence of the defects from the supplier so if the quality improvements

are done cooperatively the lean success in the supply chain is inevitable.

This study is done on an electronics factory, data are driven from an assembly line of a type

of television. There are 12 fields and 25311 records. Data set is consisted of month, firm no,

supplier, day of inspection, shift, code, material description, place, controller, status and day.

1. DATA PREPARATION

In the Table 1, the distribution of the variables used in the analysis could be seen.

Table 1- Data Audit


40 www.jiarm.com

Figure 1 is about the distribution of the months. The 9th month has more defects when it is

compared to the seventh, eighth and the nineth month. It can be seen in the Figure 2 as well.

Figure 1- Distribution of month

Figure 2- Histogram of day of inspection

The Figure 3 shows the distribution of the days that the defects are done. 1 is coded as

Sunday and it is the day that the most of the defects are done.

Figure 3- Distribution of day


41 www.jiarm.com

The Figure 4 shows the distribution of the shift. In the 2. shift the biggest number of defects

are found out from the analysis. The factory’s first shift starts at 08:00 ends in16:00, 2. Shift

is between 16:00- 24:00 and the 3.shift is between 24:00- 08:00.

Figure 4- Distribution of shift

In Figure 5, one can see the number of defect in the two factories of the company. In the

analysis it is found that the first factory has 63% defect rate which is more than the second

factory’s 36.95% defect rate.

Figure 5– Distribution of place

In Figure 6 distribution of the materials can be seen. Logos are the material that contains

biggest number of defects.

Figure 6- Distribution of material

The pare to chart of the defected materials can be seen in Figure7. According to the Paretos

80/20 rule, if 20 % of the defects could be improved than the rest of them could be solved.


42 www.jiarm.com

C o u n t 280268265245245240206205195195491 193186180178153144142138130112407 1091081061051039997969393400 918888416399374315307289P erc en t 3 3 3 3 3 3 2 2 2 26 2 2 2 2 2 2 2 2 2 15 1 1 1 1 1 1 1 1 1 15 1 1 1 55 4 4 4 3C u m % 384144475053555860626 6467697172747677798010 8283848586888990919215 9394951002024283135

Valu eO th er

C r ysta

l /Osila

t o r /Rezo

n ator

C o ils

M etal M

at er ia

ls 4

Sa tellit e

Re c eiv

er

C able s

N e two rk

Kab le sSm ts

D io des

F e r r ite s

Jac k s

C apa ci t

o r Ch ip

Ba si l

O the r Cab

le

Me tal M

ater ia

ls

D io de chip

PVC c able s

C a b le c o nne c to

r

L oudspe a k er

s

T r ansis

to r base

cuu r e

n t HFE

C able s w

i th c onne c to

r s

C o ils c h ip

C ables

tra p

Po ly ur etha ne

Scr ew W

ashe r

Po we r ad ap

to r ope n

R esista

nc e Ch ip

Rem o te Con tr o

l ler

LCD fr

on t Panel

C o v e r

N y lon Bag s

D ispla y tu

be

Re sistan

ces

C om plete LC

D P C B disp

la yPC Bs

p lack e t fo

nk sion

C omp le te n etw

o r k c a b le

v izol in

co t to

ns

C apa ci t

o r Ela

Su ppo r ts si z

e 3

Sc r ews

O t her M at e

r ials

L ogo s

9000

8000

7000

6000

5000

4000

3000

2000

1000

0

100

80

60

40

20

0

Coun

t

Perc

ent

P a re to C har t of Va lue

Figure 7- Pareto chart of the defected materials

Figure 8- Distribution of controller

The controller SD found so many defaults. (Figure 8) But in the Table 2 it is found that this is

the total amount that he inspected. In the Table 2, the matrix shows that the number of

defects that the controlled, which is coded as binary,1 is coded for defected and the 0 code is

for the undefected parts.

Table 2- Matrix of controller by status

In the Table 3 one can see the number of defects coming from the suppliers. The biggest

number is from the supplier B, 693 defected units. The name of the suppliers is confidential

information so that their names are coded. In the decision tree the biggest number of defects


43 www.jiarm.com

are coming from the “logo material group” which are procured from the supplier B. “Solder

materials” as defected products are procured from supplier A.

Table 3- Defect amount in supplier

Supplier 0 1

B 17 693

C 2 448

X 13 407

Z 4 323

A 0 32

2. MODELING

Figure 9- Model

In the Figure 9 the constructed model can be seen. In the data mining analysis,

decision tree technique is used. Since the data was big, sampling is done for the decision tree.

In practice, one wants to have a small and accurate tree for many reasons. A smaller tree is

more general and also tends to be more accurate. It is also easier to understand by human

users. In the Figure 1 it is seen that the number of controlled units were highest in the 9th

month. In the 9th month more inspection is done. In the decision tree also it is found that

month was important on the number of defects.


44 www.jiarm.com

Figure 10– Decision tree

In both branches of the decision tree 99.744% defective units are found in the 9th month and

after that month 93.151% defects is found out. The defects after the 9th month can be see in

Table 4. Logos have the highest number of defects in this branch. Also controller effect is

also important AD and SD controllers’ (%100) of the parts are found to be defective.

Table 4- Matrix of material by status

Material 0 1

Boxes 0 3

Basilide 1 446

Bobines 0 3

Entegrated circuits 0 1

Cables 0 32

Chemical Materials 0 10

Connectors 0 120

Lcd mechanical equip. 0 14

Solder 0 22

Logos 16 236

Placets 0 1

Placet function 0 9

Polyuretans 0 1

Satelite receiver 0 26

Scart Soccets 0 52

Bobins 0 2

Remote controllers 0 5


45 www.jiarm.com

Conclusion

Manufacturing firms operating in rapidly changing and highly competitive markets

have embraced the continuous process improvement mindset. Defect Rate is a KPI for

quality. This is used to measure the amount of product faults. Depending on the type of

product manufactured and the volume at which it is manufactured, this KPI may be measured

as faults per unit or defects per million. Data were collected from the manufacturing plant,

which indicated that the daily defect rates were significant. These figures gave a clear

indication that the number of defects could be significantly reduced to a few parts within the

total production. After the analysis collaborative quality improvement studies are done on the

first tier suppliers for the defected units and 60% improvements are done on the defect rate

KPI. Quality is a predictable degree of uniformity and dependebility, at low cost and suited

to the market. Improvement and innovation are both required if a form is to be healthy in the

future. The purpose of process improvement is to modify current methods to continuously

reduce the difference between customer needs and process performance. Continuous quality

improvement and cost reduction are necessary for an organization’s health and competitive

economy. Quality improvement requires never ending reduction of variation in product and

or process performance around nominal values. Supplier partnerships and strategic alliances

bring important mutual benefits. Reduced transaction costs (cost of information gathering,

negotiation, contracting, billing), improved resource planning & investment decisions, greater

production predictability and efficiency, improved deployment of complementary

capabilities, greater knowledge integration and R&D effectiveness, incentives for increased

innovation (through cost-sharing, risk-sharing, knowledge-sharing) and increased mutual

commitment to improving joint long-term competitive performance. Lean supplier networks

offer significant competitive advantages: Exhibit superior performance system-wide, greater

efficiency, lower cycle time, higher quality, not an accident of history but result of a dynamic

evolutionary process, not culture dependent but are transportable worldwide, can be built

through a proactive, well-defined, process of change in supply chain management

(Nightangale, 2016).

This paper focuses on the lean supply chain management systems with its emphasis on

continuous improvement and defect reduction. Identification of opportunities for eliminating

defects as a means of achieving continuous improvement was considered to be important.

Data Mining is used for data analytics, to make this knowledge actionable within the

enterprise to improve business processes and identification of the defects.


46 www.jiarm.com

References

1. Adina-Petruţa P., S. Roxana, 2014. Integrating Six Sigma with Quality Management Systems for the Development and Continuous Improvement of Higher Education Institutions, Procedia - Social and Behavioral Sciences, Volume 143, Pages 643-648, ISSN 1877-0428.

2. Ahmad M.F., N. Zakuan, A. Jusoh, J. Takala, 2012. Relationship of TQM and Business Performance with Mediators of SPC, Lean Production and TPM, Procedia - Social and Behavioral Sciences, Volume 65, 3 Pages 186-191, ISSN 1877-0428.

3. Al-Shdaifat E. A., 2015. Implementation of total quality management in hospitals, Journal of Taibah University Medical Sciences, Volume 10, Issue 4, Pages 461-466, ISSN 1658-3612.

4. Arslankaya S., H. Atay, 2015, Maintenance Management and Lean Manufacturing Practices in a Firm Which Produces Dairy Products, Procedia - Social and Behavioral Sciences, Volume 207, Pages 214-224, ISSN 1877-0428.

5. Baykasoglu A., L. Özbakir, S. Kulluk, 2011. Classifying defect factors in fabric production via DIFACONN-miner: A case study, Expert Systems with Applications, Volume 38, Issue 9, September Pages 11321-11328, ISSN 0957-4174.

6. Behrouzi, K. Y. Wong, 2011. Lean performance evaluation of manufacturing systems: A dynamic and innovative approach, Procedia Computer Science, Volume 3, Pages 388-395, ISSN 1877-0509.

7. Büyüközkan G., G. Kayakutlu, İ. S. Karakadılar, 2015. Assessment of lean manufacturing effect on business performance using Bayesian Belief Networks, Expert Systems with Applications, Volume 42, Issue 19, Pages 6539-6551, ISSN 0957-4174.

8. Card D. N., 2005. Defect Analysis: Basic Techniques for Management and Learning, Advances in Computers, Elsevier, Volume 65, Pages 259-295, ISSN 0065-2458, ISBN 9780120121656.

9. Chen J. C., C. Cheng, P.T.B. Huang, 2013. Supply chain management with lean production and RFID application: A case study, Expert Systems with Applications, Volume 40, Issue 9, Pages 3389-3397, ISSN 0957-417.

10. Choomlucksana J., M. Ongsaranakorn, P. Suksabai, 2015. Improving the Productivity of Sheet Metal Stamping Subassembly Area Using the Application of Lean Manufacturing Principles, Procedia Manufacturing, Volume 2, Pages 102-107, ISSN 2351-978.

11. Colledani M., T.Tolio, A. Fischer, B. Iung, G. Lanza, R. Schmitt, J. Váncza, 2014. Design and management of manufacturing systems for production quality, CIRP Annals - Manufacturing Technology, Volume 63, Issue 2, Pages 773-796, ISSN 0007-8506

12. Cox A., D. Chicksand, 2005. The Limits of Lean Management Thinking:: Multiple Retailers and Food and Farming Supply Chains, European Management Journal, Volume 23, Issue 6, Pages 648-662, ISSN 0263-2373.

13. Data Mining Applications and Trends, http://www.tutorialspoint.com/data_mining/dm_applications_trends.htm.

14. Defect management, 2016.http://www.defectmanagement.com/defectmanagement/modelmain.htm14.04.2016.

15. Dhafr N., M. Ahmad, B. Burgess, S. Canagassababady, 2006. Improvement of quality performance in manufacturing organizations by minimization of production defects, Robotics and Computer-Integrated Manufacturing, Volume 22, Issues 5–6, Pages 536-542, ISSN 0736-5845.

16. Dües C. M., K.H Tan, M. Lim, 2013.Green as the new Lean: how to use Lean practices as a catalyst to greening your supply chain, Journal of Cleaner Production, Volume 40, Pages 93-100, ISSN 0959-6526.

17. Fei Z., J.Zhang and X Zhou.,. 2010. Research on the application of data mining in logistics enterprise, ICLEM, American Society of Civil Engineers, pp. 2288-2295.

18. Ferretti S., D. Caputo, M. Penza, D.M. D’Addona, 2013.Monitoring Systems for Zero Defect Manufacturing, Procedia CIRP, Volume 12, Pages 258-263, ISSN 2212-8271.

19. Folinas D., D. Aidonis, D. Triantafillou, G. Malindretos, 2013. Exploring the Greening of the Food Supply Chain with Lean Thinking Techniques, Procedia Technology, Volume 8, Pages 416-424, ISSN 2212-0173.

20. Fullerton R.R., F. A. Kennedy, S. K. Widener, 2014. Lean manufacturing and firm performance: The incremental contribution of lean management accounting practices, Journal of Operations Management, Volume 32, Issues 7–8, Pages 414-428, ISSN 0272-6963

21. Gunasekaran A., K. Lai, T.C. E. Cheng, 2008. Responsive supply chain: A competitive strategy in a networked economy, Omega, Volume 36, Issue 4, , Pages 549-564, ISSN 0305-0483.

22. Han J.and M.Kamber. 2006. Data mining concepts and techniques, Morgan Kauffman, p.1-285. 23. Hartini S., U. Ciptomulyono, 2015. The Relationship between Lean and Sustainable Manufacturing on

Performance: Literature Review, Procedia Manufacturing, Volume 4, Pages 38-45, ISSN 2351-9789.


47 www.jiarm.com

24. IBM SPPS, Decision trees 22, p.1. 25. Indrawati S., M. Ridwansyah, 2015, Manufacturing Continuous Improvement Using Lean Six Sigma:

An Iron Ores Industry Case Application, Procedia Manufacturing, Volume 4, Pages 528-534, ISSN 2351-9789.

26. Kull T. J., T. Yan, Z. Liu, J. G. Wacker, 2014. The moderation of lean manufacturing effectiveness by dimensions of national culture: Testing practice-culture congruence hypotheses, International Journal of Production Economics, Volume 153, Pages 1-12, ISSN 0925-5273.

27. Lemmens S., C. Decouttere, N. Vandaele, M. Bernuzzi, 2016. A review of integrated supply chain network design models: Key issues for vaccine supply chains, Chemical Engineering Research and Design, Volume 109, Pages 366-384, ISSN 0263-8762

28. Liu B.2007. Web data mining, Springer, p.59-60. 29. Martínez-Jurado P. J., J. Moyano-Fuentes, 2014. Lean Management, Supply Chain Management and

Sustainability: A Literature Review, Journal of Cleaner Production, Volume 85, Pages 134-150, ISSN 0959-6526

30. Mohammaddust F., S. Rezapour, R. Z. Farahani, M. Mofidfar, A. Hill, 2015. Developing lean and responsive supply chains: A robust model for alternative risk mitigation strategies in supply chain designs, International Journal of Production Economics, ISSN 0925-5273.

31. Naylor J B., M. M Naim, D.Berry, 1999.Leagility: Integrating the lean and agile manufacturing paradigms in the total supply chain, International Journal of Production Economics, Volume 62, Issues 1–2, , Pages 107-118, ISSN 0925-5273.

32. Nelson J., 2016, Chapter 3 - Pull versus Push: Lessons from Lean Manufacturing, In Becoming a Lean Library, Chandos Publishing, Pages 29-49, ISBN 9781843347798.

33. Nightingale D, 2005 Massachusetts Institute of Technology, http://ocw.mit.edu/courses/aeronautics-and-astronautics/16-852j-integrating-the-lean-enterprise-fall-2005/lecture-notes/7_lean_sup_ch_mg.pdf 14.04.2016

34. Oleghe O., K. Salonitis, 2016, Variation Modeling of Lean Manufacturing Performance Using Fuzzy Logic Based Quantitative Lean Index, Procedia CIRP, Volume 41, Pages 608-613, ISSN 2212-8271,

35. Perner M., S. Algermissen, R. Keimer, H.P. Monner, Avoiding defects in manufacturing processes: A review for automated CFRP production, Robotics and Computer-Integrated Manufacturing, Volume 38, April 2016, Pages 82-92, ISSN 0736-5845.

36. Purvis L., J. Gosling, M. M. Naim, 2014.The development of a lean, agile and leagile supply network taxonomy based on differing types of flexibility, International Journal of Production Economics, Volume 151,Pages 100-111, ISSN 0925- 5273

37. Robeson J. F. , 1994. Logistics handbook, Simon and Schuster, ISBN 1439106258, 9781439106259. 38. Roh J., P.Hong, H. Min, 2014. Implementation of a responsive supply chain strategy in global

complexity: The case of manufacturing firms, International Journal of Production Economics, Volume 147, Part B, Pages 198-210, ISSN 0925-5273.

39. Sarkar B., S. Saren, 2016. Product inspection policy for an imperfect production system with inspection errors and warranty cost, European Journal of Operational Research, Volume 248, Issue 1, Pages 263-271, ISSN 0377-2217.

40. Tan P.N., M.Steinbach and V.Kumar.2006. Introduction to data mining, p. 3,151-152,327. 41. Yang T., Y. Kuo, C.T. Su, C.L. Hou, 2015. Lean production system design for fishing net

manufacturing using lean principles and simulation optimization, Journal of Manufacturing Systems, Volume 34, , Pages 66-73, ISSN 0278-6125.

42. You F., I. E. Grossmann, 2008. Design of responsive supply chains under demand uncertainty, Computers & Chemical Engineering, Volume 32, Issue 12, Pages 3090-3111, ISSN 0098-1354.

Documents

SUPPLIER DEFECT DATA ANALYTICS IN LEAN SUPPLY CHAIN … · al.(2008) worked on responsive supply chain with a competitive strategy in a networked economy. Supply chain management