chapter 2 - inflibnetshodhganga.inflibnet.ac.in/bitstream/10603/35278/10/10_chapter2.pdf ·...

29

1

Chapter: 2 SIX SIGMA METHODOLOGY IN INDIAN INDUSTRIES –

OPPORTUNITIES AND CHALLENGES

2.1 Introduction and methodology

Six Sigma is a systematic methodology aimed at operational excellence through

continuous process improvements (Pande et al., 2003). It is a process-focused and data

driven methodology aimed at near elimination of defects in all processes which are critical

to customers (Montgomery and Woodall, 2008). Six Sigma allows for more careful

analysis and more effective decision-making aiming for the optimal solution rather than

what is simply ‘good enough’. This methodology has been successfully implemented

worldwide for over 25 years, producing significant improvements to the profitability of

many large and small organisations (Brady and Allen, 2006). It has been considered as a

strategic approach to improve business profitability and to achieve operational excellence

through effective application of both statistical and non-statistical tools/techniques (Hoerl,

2004).

The application of Six Sigma methodology is growing almost every day (Halliday,

2001). Moving from the manufacturing industry to service, transactional, administrative,

research and development, sales and marketing, healthcare, construction projects and

software-development industries with the aim of continuously improving the processes and

The results included in this chapter are discussed in Gijo (2011), Gijo (2012), Gijo and Antony (2013), Gijo et al (2013), Gijo and Sarkar (2013), Gijo and Scaria (2010; 2013b) and Gijo et al (2011).

Chapter: 2

30

thereby improving the customer satisfaction (Banuelas et al, 2005; Stewart and Spencer,

2006; Mahanti and Antony, 2009; Firka, 2010; Kaushik et al, 2012).

A four stage approach is used for this study, namely problem definition, literature

survey, case study design and the data analysis. A detailed literature review was

undertaken in Six Sigma with an objective of identifying the type of improvements carried

out by different people in various organizations to address process related problems. The

researcher worked with the companies to provide support for the project in Six Sigma

techniques, whilst recording data about the exercise from which to develop a case study. A

case study entails the detailed and intensive analysis of a single case - a single

organisation, a single location or a single event (Bryman and Bell, 2006). Yin (2003)

describes a case study as an empirical inquiry that investigates a contemporary

phenomenon within its real-life context. According to Lee (1999), the unit of analysis in a

case study is the phenomenon under study and deciding this unit appropriately is central to

a research study. In this chapter, few case studies are designed to study the underlying

process problems so that solutions can be implemented for process improvement. The

extent to which generality can be claimed from few case studies are limited, but by

documenting case experiences in the light of existing literature, each case adds to the sum

of knowledge available for future practitioners and researchers (Cooper and Schindler,

2006).

Based on the available data on the process, studies were conducted to understand

the baseline status of the processes and drafted a project charter for each case study, which

explain the details of the problems. The collected data were analysed using descriptive and

inferential statistics. The work described in this chapter was conducted in various

Six Sigma Methodology in Indian Industries – Opportunities and Challenges

31

organizations in India, from manufacturing to health care and even organizations like non-

conventional energy. The purpose of this study is to demonstrate the power of application

of Six Sigma methodology in industrial scenario and identify the general drawbacks of this

methodology during implementation.

The rest of this chapter is arranged as follows. In section 2.2, a study on application

of Six Sigma in reducing the defects of a grinding process is discussed. Section 2.3

discusses the integration of Taguchi’s Beta correction technique with Six Sigma

methodology. The summary of the applications of Six Sigma methodology in a

manufacturing company, a super-specialty hospital and a wind mill company along with

conclusions are presented in section 2.4. The shortcomings in implementation of Six Sigma

methodology is discussed in section 2.5. The chapter ends with discussions in section 2.6.

2.2 Application of Six Sigma methodology to reduce defects of a grinding process

The present study deals with the reduction of defects of a fine grinding process in

an automobile part manufacturing company in India. The company with manpower of

approximately 2550 people is manufacturing common rail direct injection (CRDI) system

pumps for vehicles. These pumps were used in cars, trucks and buses throughout the

world. An injector primarily consists of nozzle and nozzle holder body. A schematic view

of fuel injector is given in Figure 2.1. The components used in fuel injector and their

functions are as follows. Distance piece aligns the high-pressure fuel lines of nozzle holder

body and nozzle. It’s both the sides are fine ground precisely to ensure sealing of the high-

pressure fuel coming from holder body to nozzle. Cap nut retains the nozzle and distance

piece with the holder body with sufficient torque to ensure sealing. Spring and pressure

Chapter: 2

32

bolt ensures the functioning of injector with set opening pressure and timely delivery of

fuel.

The current project was undertaken in the distance piece fine grinding process,

which is done by fine grinding machine. Different types of distance pieces were fine

ground in this machine. This is a sophisticated and very expensive CNC fine grinding

machine. It finishes both faces of distance pieces in batches precisely with sub-micron

flatness values.

After fine grinding, the distance pieces were inspected visually to find various

defects. Since the production of distance pieces were in thousands per shift, it was not

practically possible to do 100% inspection of these components by objective methods.

Hence visual inspection was carried out for all the components with reference to

masterpieces and visual limit samples. Since rejection level of distance pieces after fine

grinding process was very high and the function of the component in the product was

highly critical, it was essential to do 100% inspection. Under these circumstances, the

project was of highest priority to the management as it was clear that an effective solution

to this problem would have a significant impact in reducing rework/ rejection and

improving productivity. In the past, many attempts were made to solve this problem by

using different methodologies, which were unsuccessful. Six Sigma problem solving

methodology (DMAIC) was recommended when the cause of the problem is unclear

(Breyfogle, 2003). Hence it was decided to address this problem through the Six Sigma

DMAIC methodology.


33

Distance piece

Figure 2.1: Schematic view of fuel injector

2.2.1 The define phase

A team was formed by involving the people associated with the process to execute

this study. The team along with champion and MBB developed a project charter (Table

2.1) with all necessary details of the project.

It was decided to consider rejection percentage of distance pieces after fine grinding

process as the Critical to Quality (CTQ) characteristic for this project. The goal statement

was defined as reduction in rejection of distance pieces by 50% from the existing level,

Nozzle holder body

Spring

Distance piece

Nozzle

Cap nut

Chapter: 2

34

which should result in large cost saving for the company in terms of reduction in rework

and scrap cost.

Table 2.1: Project charter - grinding process improvement

Project Title: Reducing rejection in distance pieces after fine grinding process

Background and reasons for selecting the project:

The rejection of distance pieces in the fine grinding process was as high as 16.2%.

Approximately 4200 components are machined during every shift. The cost of

components rejected due to defects was approximately $US1.6 million per annum. In

addition to this, there was loss associated with tool, machine and man-hour related to

rejection of components.

Aim of the project:

To reduce the rejection of distance pieces by 50% from the existing level.

Critical to Quality characteristic:

Rejection percentage of distance pieces after fine grinding process.

Project Scope Fine grinding process

Project Champion: Head - Manufacturing

Project Leader: Senior Manager - Manufacturing

Team Members: Planning Manager, Maintenance Manager,

Quality Control Senior Engineer,

Machine Operator.

Expected Financial

Benefits:

A saving of approximately $US one million in terms of

reduction in rejection and tool cost.

Expected Intangible

Benefits:

Reduction in rejection will lead to increased output to

meet the market demand and thereby increase in turnover

and reduction in operational expenses.

Expected customer benefits: Improving on time delivery.

Schedule: Define: 2 Weeks, Measure: 2 weeks

Analyze: 3 weeks, Improve: 4 weeks

Control: 4 weeks.


35

Since there was a cross-functional team for executing this project, the team felt that

it was necessary to perform a SIPOC analysis to have a better understanding of the process.

The team focused on the fine grinding process for improvement, which is defined as the

scope of the project. The process mapping along with SIPOC is presented in Table 2.2.

Table 2.2: SIPOC for fine grinding process

Supplier Input Process Output Customer

Vendor Pre-finished

parts

Fine

Grinding

Process

Finished parts Assembly shop

Sand blasting

process

Shot blasted

parts

Planning

Department

Setting

parameters

Production

reports

Manufacturing

Department

Planning

Department

Dressing method

Planning

Department

CNC program Quality reports Quality

Department

Planning

Department

Visual limit

sample

Planning

Department

Tooling

Process Steps

Visual Inspection

Cleaning & Arranging

Sand Blasting

Fine Grinding

Pre-Grinding

Chapter: 2

36

2.2.2 The measure phase

The CTQ considered in this case was the rejection percentage of distance pieces

after the fine grinding process. These rejections were mainly due to the occurrence of

different types of defects like burr, shades, deep lines, patches and damage on the

component after machining. The schematic representation of these defects is presented in

Figure 2.2. These defects create an uneven surface in the component that could lead to fuel

leakages in pumps. After machining, the components were visually inspected for these

defects. Master samples were provided for identifying each of these defects and inspectors

did the inspection. Since there was no instrument involved in the inspection process and

only visual inspection was performed, before going ahead with further data collection, the

team decided to carryout Attribute Gauge Repeatability and Reproducibility (Gauge R &

R) study to validate the measurement system. In such studies, intra-inspector agreement

measures repeatability (within inspector), inter inspector agreement measures the

combination of repeatability and reproducibility (between inspectors). The non-chance

agreement between the two inspectors, denoted by Kappa, defines as

agreementschanceofNumbernsobservatioofnumberTotalagreementschanceofNumberagreementsobservedofNumber

−−

=κ

For conducting the study, 100 components were selected and they were classified as

good or bad independently by two inspectors. From the resulting data, the Kappa value

was calculated and was found to be 0.814 with a standard error of 0.0839. Since the Kappa

value was more than 0.6, the measurement system was acceptable (Landis and Koch,

1977).


37

Deep lines

PatchesBurr

Damage

Shade

Figure 2.2: Schematic representation of defects

After the measurement system study, a data collection plan was prepared with

details of type of data, stratification factors, sampling frequency, method of measurement

etc. for the data to be collected during the measure phase of this study. The data collection

plan thus prepared is presented in Table 2.3. Data were collected as per the plan to

understand the baseline status of the process. During the defined period of data collection,

368219 components were inspected and 61198 components were rejected due to various

defects. Each one of the rejected components was having one or more defects. Detailed

data on the type of defects were collected and the same was graphically presented as a

pareto diagram (Figure 2.3). The collected data shows that the rejection in the process was

166200 PPM (parts per million). The corresponding sigma rating of the process can be

approximated to 2.47.

Chapter: 2

38

Table 2.3: Data collection plan - distance piece rejection

Characteristic Data type How

measured

Sampling

notes

Related

conditions

Rejection

percentage of

distance pieces after

fine grinding

process

Attribute Visual

checking by

comparing with

visual limit

samples

100% of units

in all the three

shifts for one

month

Shift wise and

defect type

wise

Cou

nt

Perc

ent

DefectCount

0.2Cum % 42.3 75.5 99.8 100.0

76539 60267 43995 318Percent 42.3 33.3 24.3

OthersPatchesDeep linesShades

200000

150000

100000

50000

0

100

80

60

40

20

0

Figure 2.3: Pareto diagram for visual defects


39

For any improvement initiative in this organization, the general goal set by the

management was to reduce the rejection by 50 percent from the existing level. Based on

this policy, the target set for the study was to reduce the rejections at the fine grinding

process to 83100 PPM from the existing level of 166200 PPM.

2.2.3 The analyze phase

After mapping the process, the team proceeded to analyze the potential causes of

defects. A cause and effect diagram was prepared after conducting a brain storming session

with all the concerned people from the process along with the project team, Champion and

MBB. The output of the cause and effect diagram depends to a large extend on the quality

and creativity of the brain storming session (Ishikawa and Lu, 1985). Figure 2.4 represents

the cause and effect analysis prepared during the brain storming session.

HighRejection inGrindingProcess

Measurements

Method

Material

Machine

Man

Repair batches mix up

Improper cleaning afterdressing

Wheel straightness not OK

Program parameters not Optimum

Material removal rate not OK

Supplier to suppliervariation

Presence of sandblasting dust

Product family to familyvariation

Variation in size ofinput parts

Loading/unloading system not OK

Improper setting

Inspector to inspectorvariation

Figure 2.4: Cause and effect diagram for rejection in grinding process

Chapter: 2

40

The next step in this phase was to gather data from the process in order to obtain a

better picture of the potential causes, so that the root cause/s can be identified. The team

had detailed discussion with the process personnel to identify the possible data that can be

collected on the potential causes in the cause and effect diagram. After getting the detailed

picture of availability of data on causes, the team discussed with the MBB to identify the

type of analysis possible on these causes. Based on this discussion, a cause validation plan

was prepared to detail the type of data to be collected and the type of analysis possible for

each of these causes. The potential causes like ‘variation in input parts’, ‘supplier material

variation’, ‘program parameters not OK’ etc. can be validated by statistical analysis on the

data collected from the process. But potential causes like ‘improper cleaning after

dressing’, repair batches mix up’ etc have to be validated only by process observation,

popularly known as GEMBA investigation. In GEMBA method of cause validation, one

need to visit the workplace and observe the practices followed and compared with the

specification and/or desired method (Womack, 2011). Based on the observations during

the period of GEMBA study, a decision can be made about the presence or absence of a

cause. Hence for few causes, detailed data were collected and statistical analyses were

planned and for the remaining causes, GEMBA was planned to validate the causes. Table

2.4 summarizes the potential causes and the type of analysis planned for each cause.

Since three suppliers provide the raw material, it was suspected that there was

possibility of supplier-to-supplier variation with respect to the thickness of input raw

material. To study this variation, data on thickness of components from all the three

suppliers were collected and ANOVA was performed in the data and p-value was

observed. In this case, the p-value was found to be 0.407, ruling out the possibility of

significant difference between the suppliers. Further to validate the potential cause of

variation in size of input parts, a batch of 57 distance pieces was selected and thickness

measured. The data were subjected to Anderson Darling Normality test, and found to


41

follow normal distribution (Montgomery and Runger, 2007). Process capability study was

carried out on this data and found to be capable, confirming that the thickness variation in

input part was not a root cause.

Table 2.4: Cause validation plan for high rejection in grinding process

Sl. No Cause Plan for validation

1 Improper cleaning after dressing GEMBA

2 Repair batches mix up GEMBA

3 Variation in size of input parts Process capability analysis

4 Product family to family variation Chi-square test

5 Presence of sand blasting dust GEMBA

6 Supplier to supplier variation ANOVA

7 Material removal rate not OK ANOVA

8 Process parameters not optimum Design of Experiments (DOE)

9 Improper setting DOE

10 Loading/unloading system not OK GEMBA

11 Wheel straightness not OK GEMBA

12 Inspector to inspector variation Gauge R & R

Since different families of products were produced, data were collected for material

removal rate as well as defects with respect to various families of components to test their

significance. Data on material removal rate (MRR) was collected on three types of

components to study the effect of type of component on MRR. Based on the quantity of

material removed from the component during machining, the MRR value was calculated

Chapter: 2

42

by an inbuilt software program in the machine. These MRR data were recorded in

millimeter/minute. ANOVA was performed on this data and p-value was found to be

0.085, not showing significance at 5% level. To test whether product family-to-family

variation affects the defects, a chi-square test was carried out between defect type and

family of components. For each of the defect types, viz, patches, shades and deep lines,

separate chi-square test was done with three different families of components. The details

of chi-square test are given in Table 2.5. From Table 2.5, it was clear that except for

shades, family-to-family variation does not affect visual defects. The machining program

and machine parameters for each family and type was different. The team thought, it was

better to have a uniform machining program and parameters for all the family of

components so that the process can be better managed. Hence for validating the process

parameters and identifying the optimum operating conditions, the team decided to conduct

a design of experiment (DOE) during the improve phase.

Table 2.5: Test statistic values for Chi-square test

Defect type Chi-Square statistic Degrees of freedom p-value

Patches 2.509 2 0.285

Deep lines 2.398 2 0.301

Shades 452.256 2 0.000

The other causes listed in the cause and effect diagram were validated by GEMBA

analysis. Some of such details are presented in Table 2.6. The detail of validation of all

causes in the cause and effect diagram was summarized in a tabular format and is given in

Table 2.7.


43

Table 2.6: GEMBA observations - grinding process

Sl. No. Cause Observation / Conclusion

1 Improper cleaning after

dressing It was observed that cleaning done after dressing.

Not a root cause

2 Repair batches mix up Repair batches were found mixed with other

batches during the random visit to the process.

Root cause.

3 Presence of sand blasting

dust Traces of shot blasting dust found in the input

batch during inspection. Root cause

4 Loading/unloading system

not OK Loading table wear out observed at the edges.

Root cause.

5 Wheel straightness not OK Wheel straightness found OK. Checking frequency

followed as per procedure. Not a root cause.

Table 2.7: Summary of validation of causes for high rejection in grinding process

Sl. No Cause Tools used for validation Results

1 Improper cleaning after dressing GEMBA Not a root cause

2 Repair batches mix up GEMBA Root cause

3 Variation in size of input parts Process capability analysis Not a root cause

4 Product family to family

variation Chi-square test Root cause

5 Presence of sand blasting dust GEMBA Root cause

6 Supplier to supplier variation ANOVA Not a root cause

7 Material removal rate not OK ANOVA Root cause

8 Process parameters not optimum DOE Root cause

9 Improper setting DOE Root cause

10 Loading/unloading system not

OK GEMBA Root cause

11 Wheel straightness not OK GEMBA Not a root cause

12 Inspector to inspector variation Gauge R & R Not a root cause

Chapter: 2

44

2.2.4 The improve phase

At this stage, as decided earlier, a DOE was planned for optimizing the process/

machine parameters. The parameters (factors) selected for experimentation were load

applied, initial load setting, coolant flow rate, upper wheel rpm, lower wheel rpm and cage

rpm. Since the relationship between these parameters and material removal rate was not

known, it was decided to experiment all these parameters at three levels (Taguchi, 1988).

The existing operating level was selected as one level for experimentation. Based on

various operational feasibilities the other two levels were selected. The parameters and

levels selected for experimentation are presented in Table 2.8. Also, during the discussion

with the technical personnel, it was felt that there is a possibility of interaction between

load applied with upper wheel rpm, load applied with lower wheel rpm and load applied

with cage rpm. Hence it was decided to estimate the effect of these three interactions also.

Six factors, each at three levels and three interactions with replications require a huge

number of components for conducting a full factorial experiment, which would be costly

and time consuming exercise. It was possible to estimate the effect of these selected

parameters and interactions using the 27 experiments with the help of Orthogonal Array

(OA). Hence for conducting an experiment with six parameters and three interactions,

L27(313) orthogonal array was selected. Experiments using orthogonal arrays play a crucial

role in achieving additivity of the model effects (Phadke, 1989). The design layout

prepared as per L27(313) orthogonal array is given in Table 2.9. The response of the

experiment was decided as MRR.


45

Table 2.8: Process parameters and their levels - grinding process

Sl. No. Factor Levels 1 Load applied 170 200* 230 2 Initial load setting Minimum Medium Maximum* 3 Coolant flow rate 8 LPM 12 LPM 16 LPM* 4 Upper wheel RPM 70* 90 110 5 Lower wheel RPM 50* 60 70 6 Cage RPM 20* 30 40

* Existing levels

Table 2.9: Design layout for experimentation - grinding process

Exp. No. Load Setting Coolant UW RPM LW RPM C RPM 1 170 Min. 8LPM 70 50 20 2 170 Min. 8LPM 90 60 30 3 170 Min. 8LPM 110 70 40 4 170 Med. 12LPM 70 60 40 5 170 Med. 12LPM 90 70 20 6 170 Med. 12LPM 110 50 30 7 170 Max. 16LPM 70 70 30 8 170 Max. 16LPM 90 50 40 9 170 Max. 16LPM 110 60 20 10 200 Med. 16LPM 70 50 20 11 200 Med. 16LPM 90 60 30 12 200 Med. 16LPM 110 70 40 13 200 Max. 8LPM 70 60 40 14 200 Max. 8LPM 90 70 20 15 200 Max. 8LPM 110 50 30 16 200 Min. 12LPM 70 70 30 17 200 Min. 12LPM 90 50 40 18 200 Min. 12LPM 110 60 20 19 230 Max. 12LPM 70 50 20 20 230 Max. 12LPM 90 60 30 21 230 Max. 12LPM 110 70 40 22 230 Min. 16LPM 70 60 40 23 230 Min. 16LPM 90 70 20 24 230 Min. 16LPM 110 50 30 25 230 Med. 8LPM 70 70 30 26 230 Med. 8LPM 90 50 40 27 230 Med. 8LPM 110 60 20

Chapter: 2

46

As per the design layout given in Table 2.9, the experiments were conducted after

randomizing the sequence of experiments, and data were collected. The experimental data

were analyzed by Taguchi’s Signal-to-Noise (S/N) ratio method. The S/N ratio can be

treated as a response (output) of the experiment, which is a measure of variation when

uncontrolled noise factors are present in the system. Since the requirement of this process

was to remove material in a uniform rate from distance pieces to achieve specified

dimension, the S/N ratio of nominal-the-best type was selected for analysis. The S/N ratio

values were estimated for all the 27 experiments and ANOVA was performed on these S/N

values to identify the significant parameters and interactions. From the ANOVA table

(Table 2.10), it was clear that the interaction effect between load applied and upper wheel

rpm was significant at 5 percent level. Also, factors setting and lower wheel rpm was

significant at 10 percent level. The main effect and interaction plots of S/N ratio values

were made and are presented in Figures 2.5 and 2.6 respectively. From these plots, the best

levels for parameters were identified as the level corresponding to highest value of S/N

ratio. Thus, the best levels for load applied and upper wheel rpm were selected from the

interaction plot and the best levels for the other parameters were selected from the main

effect plot. The optimum combination for process parameters thus identified is given in

Table 2.11.


47

Table 2.10: ANOVA table for S/N ratio – grinding process

Source DF Seq. SS Adj. SS Adj. MS F p - value Load 2 0.0031630 0.0031630 0.0015815 07.00 0.125

Setting 2 0.0073185 0.0073185 0.0036593 16.20 0.058**

Coolant 2 0.0020963 0.0020963 0.0010481 04.64 0.177

UW RPM 2 0.0005852 0.0005852 0.0002926 01.30 0.436

LW RPM 2 0.0050296 0.0050296 0.0025148 11.13 0.082**

C RPM 2 0.0018296 0.0018296 0.0009148 04.05 0.198

Load x UW RPM

4 0.0198815 0.0198815 0.0049704 22.00 0.044*

Load x LW RPM

4 0.0057037 0.0057037 0.0014259 06.31 0.141

Load x C RPM 4 0.0062370 0.0062370 0.0015593 06.90 0.131

Error 2 0.0004519 0.0004519 0.0002259

Total 26 0.0522963

*Significant at 5 percent level, **Significant at 10 percent level

Mea

n of

SN

rat

ios

230200170

7.8

7.6

7.4

7.2

7.0

Max.Med.Min. 16LPM12LPM8LPM

1109070

7.8

7.6

7.4

7.2

7.0

706050 403020

Load Setting Coolant

UW RPM LW RPM C RPM

Figure 2.5: Main effects plot (data means) for S/N ratios

Chapter: 2

48

Load

LW RPM

8.0

7.2

6.4

706050

C RPM

UW RPM

1109070 403020

8.0

7.2

6.4

8.0

7.2

6.4

230200170

8.0

7.2

6.4

Load

230

170200

UW

110

RPM7090

LW

70

RPM5060

C RPM

40

2030

Figure 2.6: Interaction plot (data means) for S/N ratios

Table 2.11: Optimum combination for process parameters for grinding process

Sl. No. Factor Optimum level

1 Load applied 170

2 Initial load setting Medium

3 Coolant flow rate 12

4 Upper wheel RPM 90

5 Lower wheel RPM 60

6 Cage RPM 30

These optimum levels in Table 2.11 were taken as solutions for the causes related to

process parameters. Finally, the list of selected solutions is presented in Table 2.12.


49

Table 2.12: Cause - solution matrix for the root causes of grinding process

Sl. No Cause Solution

1 Repair batches mix up New storage system for repair parts

introduced in the process

2 Product family to family

variation

Process parameters were optimized as per

result of DOE

3 Presence of sand blasting dust Cleaning method after sand blasting

introduced

4 Material removal rate not OK Reference table prepared for adjusting

load

5 Process parameters not OK Optimum factor level combination from

DOE

6 Improper setting Optimum factor level combination from

DOE

7 Loading/unloading system not

OK

Conditioning of grinding wheel-loading

table is done.

A risk analysis was carried out to identify any possible negative side effects of the

solutions during implementation. The team concluded from the risk analysis that there

were no significant negative impacts associated with any of the selected solutions. Hence,

an implementation plan was prepared for the above solutions with responsibility and target

date for completion for each solution. A time frame of two weeks was provided for

implementing these solutions. All the solutions were implemented as per the plan and

results were observed. A graphical presentation of the comparison of results before and

after the study is provided in Figure 2.7.

Chapter: 2

50

Days

Perc

enta

ge

343128252219161310741

20

15

10

5

0

Before After

Figure 2.7: Rejection percentages - before and after the project

2.2.5 The control phase

Standardization of the solutions was taken care by affecting necessary changes in

the process procedures that was a part of the quality management system of the

organization. The quality plans and control plans were revised as per the solutions

implemented and issued to the corresponding users. As a part of ISO 9001 implementation,

once in three months internal audits were carried out in the process. The CTQs of the

projects were added to the internal audit checklist so that verifications can be performed

during the audits. Implementing appropriate control chart can do future monitoring of the

process for assignable causes. Since there was possibility of different types of defects like,

shades, deep line, patches etc., can appear after grinding, a control chart need to be

introduced for monitoring the process. Since the defect related data were collected from the


51

process, the most appropriate control chart for this situation was the u chart (Grant and

Leavenworth, 2000). Hence, u chart was introduced for monitoring the process along with

a reaction plan. Every shift, data were collected on number of defects observed during

visual inspection and these values were plotted on the u chart by the quality control

inspector attached with this process. When any signal for assignable cause appears in the

control chart, the quality control inspector discusses this issue with the operator and

immediate action was initiated on the process. The reaction plan displayed near to the

machine gives direction for identifying the action required for addressing the assignable

cause. Also, training was provided for the people associated with the process about the

improved operational methods so that they are able to manage the process effectively.

After implementation, data were compiled from the fine grinding process with

respect to the defects for one month and the rejection percentage was found to be 1.19.

Hence, as a result of this project, the rejection percentage of the distance pieces at the fine

grinding process reduced from 16.6 percent to 1.19 percent. The corresponding

approximate sigma level was estimated at 3.76. Thus, the sigma level of the process has

improved from 2.47 to 3.76. This shows significant improvement in terms of sigma rating

as well as defect percentage.

2.2.6 Observations

This section presents the step-by-step application of the Six Sigma methodology for

reducing the rejection level of a fine grinding process. Several statistical tools and

techniques were effectively utilized to make inferences during the project. As a result of

the project, the rejection level of distance pieces after the fine grinding process has been

reduced substantially. Once the results were observed, with the help of the finance

Chapter: 2

52

department, a cost-benefit analysis for the study was performed. Due to improvement in

the process, cost associated with rejection, repair, scrap, re-inspection and tool came down

drastically. The annualized savings resulted from this project was estimated and found to

be about US$2.4 million. This is another highlight of the study. This has given an

encouragement for the management to implement Six Sigma methodology for all

improvement initiatives in the organization.

2.3 Integration of Taguchi’s Beta correction technique with Six Sigma- methodology

In today’s fast-paced global economy, markets demand that companies produce

their products more quickly with better quality and at the same time with lesser cost. In

order to meet these requirements, organizations adopt various methodologies for process

improvement. This section deals with the first pass yield improvement of a grinding

operation in a large manufacturing company in India. They were involved in

manufacturing of automobile parts, components and subassemblies for various original

equipment manufacturers (OEMs) in India and abroad. Because of the complexity of the

manufacturing process and accuracy requirements for the products, the organization is

equipped with high-precision machineries and highly competent work force of around

1850 personnel. This organization was having a system of performing value stream

mapping (VSM) for the entire process (i.e., from customer order to delivery of product to

customer) for identifying the bottleneck areas in the process. VSM not only handles a

specific process, but also provides an overall view of the entire system, seeking to optimise

it as a whole (Jiménez et al, 2012). This VSM helps the organization to highlight the

problem areas and bottleneck processes which restrict the process flow for meeting the


53

customer requirements (Braglia et al, 2006). During this analysis, it was identified that the

plunger machining process have very low first pass yield as the rejection and re-work was

very high in the process. Hence this was identified as a priority area for the organization to

focus their efforts for improving the process.

In this part of the chapter, the study of machining of ‘plunger’, which is a critical

component in fuel injectors used in diesel engines is discussed. The first pass yield of the

plunger manufacturing process was as low as 94 percent, leading to rejection of

approximately 1900 components per month. This was creating huge loss to the company in

terms of rejection/ rework / scrap and delay in product delivery to customer. The estimated

financial loss due to rejection and scrap alone was around US$ 95,000 per annum. In

addition to this, the dissatisfaction of customers due to violations in on-time delivery was

creating negative impact on the reputation of the company. Hence it was decided to

address this problem through Six Sigma DMAIC methodology. All the five phases of the

Six Sigma DMAIC methodology was successfully implemented in this study as explained

in the following sections.

2.3.1 Different phases of DMAIC

(i) The define phase

Similar to the previous case, during the define phase, a team was formed for

executing this study. The team prepared the project charter defining the details of the

project (See Table 2.13 for Project charter).

Chapter: 2

54

Table 2.13: Project charter - plunger manufacturing process improvement

Project Title: First pass yield improvement in the plunger manufacturing line.

Background and reasons for selecting the project:

This is a high volume production process utilizing costly equipments and tools. The

first pass yield of the process is as low as 94 percent resulted in scrapping of around

1900 components every month. The problem is very complex and there are too many

variables affecting the taper and foot face thickness. The estimated financial loss due to

rejection and scrap alone was around US$ 95,000 per annum.

Aim of the project:

To improve the first pass yield from 94 percent to 99 percent in plunger manufacturing

line.

Project Champion: Manager – Production

Project Leader: Assistant Manager – Production

Team Members: Engineer – Maintenance, Engineer – Product Planning

Supervisor – Production, Engineer – Quality Control

Operator – Shift I, Operator – Shift II, Operator – Shift

III

Expected Benefits: A saving of approximately US$ 100,000.

Expected customer

benefits:

Reduction of customer complaints related to field failure

and delay in delivery.

Schedule: Define: 2 Weeks, Measure: 3 weeks, Analyse: 4 weeks,

Improve: 4 weeks, Control: 8 weeks


55

A process flow chart along with a SIPOC mapping was prepared to understand the

details of the process (See Table 2.14 for details).

Table 2.14: SIPOC along with process map for foot face and finish size grinding

Supplier Input Process Output Customer

Vendor Pre-machined

parts

Foot face

grinding &

Finish size

grinding

Finished parts Stores

Planning

Department

CNC program Production and

Quality reports

Manufacturing

Department

Calibration

Department

Gauges

Planning

Department

Tooling

Process Steps

Past data for one year was collected to understand the type of defects leading

towards low first pass yield. During the one-year period under study, 23453 components

were rejected due to different problems. Out of the rejected components, 97.3 percent were

rejected due to problems at finish size grinding and foot face grinding processes. During

finish size grinding, the plunger taper was measured and during foot face grinding, the foot

thickness was measured (See Figure 2.8). Hence it was decided to consider the ‘plunger

taper’ and ‘foot thickness’ as the critical to quality (CTQ) characteristics. The specification

limits for these characteristics were 0.5 to 1.5 microns and 3.45 to 3.5 microns

Receive OD pre-grind

components

Foot face grinding

Finish size grinding

Inspection

Chapter: 2

56

respectively. Since the tolerance for these characteristics were very narrow, it was a big

challenge for the organization to maintain the manufactured components within these

tolerances.

Taper

Foot thickness

Figure 2.8: Schematic diagram of plunger

(ii) The measure phase

In this study, the CTQs considered were ‘plunger taper’ and ‘foot thickness’. Since

the tolerances for both the characteristics were very narrow, it was decided to conduct a


57

measurement system analysis to validate the measurement process related to these

characteristics. Two inspectors and 10 components were identified for conducting this

measurement system analysis study. These two inspectors measured the plunger taper and

foot thickness for all the 10 components twice. These data were analyzed and the total

gauge repeatability and reproducibility (GR&R) values were found to be 9.71 percent and

4.1 percent respectively for foot thickness and plunger taper. The output of one of the

GR&R study is presented in Table 2.15. Since percentage GR&R values were less than

10% for both the cases, it was concluded that the current measurement system was

adequate for further data collection (AIAG, 2002).

Table 2.15: Results of Gauge R & R study

Source Standard deviation % Study variation

Total Gauge R & R 0.0002509 9.71

Repeatability 0.0002108 8.16

Reproducibility 0.0001361 5.27

Part-to-part 0.0025717 99.53

Total variation 0.0025839 100.00

After the measurement system study, a data collection plan was prepared with all

details of the data required to be collected, including sample size, frequency of sampling

etc. As per the data collection plan, a sample of size 1500 components was collected from

the process across a period of one month. The taper and foot thickness were measured for

those 1500 components and data were recorded. The next step in the measure phase was to

evaluate the base line performance for the selected CTQs with the collected data. For this

purpose, the data were tested for normality by Anderson-darling normality test and the p-

values of both the data set were found to be less than 0.05 with A-D statistic values of

Chapter: 2

58

3.5103.4953.4803.4653.4503.4353.420

LSL USLProcess Data

Sample N 1500S tDev (O v erall) 0.0140109

LSL 3.45Target *U SL 3.5Sample Mean 3.47362

O v erall C apability

C pm *

Pp 0.59PP L 0.56

PP U 0.63Ppk 0.56

O bserv ed PerformanceP PM < LSL 20666.67P PM > USL 14666.67P PM Total 35333.33

Exp. O v erall P erformanceP PM < LSL 45927.60P PM > USL 29852.41P PM Total 75780.02

89.827 and 10.403 respectively for foot thickness and plunger taper respectively, indicating

that the data were from a process that was not normally distributed. Further, the data were

tested for all known distributions with the help of Minitab software, but failed to identify

any specific distribution for this data. The Box-Cox transformation also was tried for the

data but was unsuccessful in transforming the data to normality.

Process capability analysis was carried out for both the characteristics to get an

estimate of the baseline performance of the process in terms of these CTQs. The output of

the process capability analysis is presented in Figures 2.9 and 2.10. The ‘observed

performances’ from these analyses were as follows: The PPM total for foot thickness and

plunger taper were 35333 and 51333 respectively. The approximate sigma level

corresponding to these PPM values were 3.31 and 3.13 respectively. This provides a

baseline for both the CTQs.

Figure 2.9: Process capability for plunger foot thickness


59

3.63.02.41.81.20.60.0-0.6

LSL USLP rocess D ata

S ample N 1500S tD ev (O v erall) 0.359057

LS L 0.5T arge t *U SL 1.5

S ample M ean 1.04179

O v erall C apability

C pm *

P p 0.46P P L 0.50

P P U 0.43P pk 0.43

O bserv ed P erform anceP P M < LS L 18666.67P P M > U S L 32666.67P P M T otal 51333.33

E xp. O v erall P e rform anceP P M < LS L 65660.53P P M > U S L 100949.09P P M T otal 166609.62

Figure 2.10: Process capability for plunger taper (iii) The analyze phase

In the process of identifying root causes of the problem, the potential causes

generated through a brainstorming session were listed separately for both the CTQs. Those

identified potential causes were presented in the form of cause and effect diagrams, as

given in Figures 2.11 and 2.12.

plunger taperVariation in

M ethods

M aterial

M achine

M an

W rong input by operator

W rong taper correction

Unnecessary correction

M achine not taken the proper correction

T ailstock Not OkAlignment between headstock &

C entre hole not OK in plunger

B ent P lunger

P repart m aterial taper Not Ok

Dressing frequency not OKGauge R & R not OK

Dressing depth not OKDressing feed not OK

Dressing speed not OKGrinding Speed not OK

Grinding Feed not OKGrinding S tock not OK

Figure 2.11: Cause and effect diagram for variation in plunger taper

Chapter: 2

60

thicknessfootVariation in

Methods

Material

Machine

Man

Wrong correction

Unwanted correction

Untrained operator

Wrong reading taken

Machine not taking correction

Jerky movement in spindle

Less stock for grinding

Dressing frequency Not Ok

properly in fixtureComponent not getting clamped

Wrong setting master

Gauge GRR Not OK

Figure 2.12: Cause and effect diagram for variation in foot thickness

From the potential causes listed in the cause and effect diagrams, the root causes to

be identified by data based validation of causes. Further, a cause validation plan was

prepared after discussing with the people working in the process. During these discussions,

it was identified that there are two categories of causes. For one category measurable data

was possible to collect from the process where as for the other category, direct measurable

data was not possible to collect. Wherever measurable data was available, appropriate

statistical techniques were selected for analysis and validation of the potential causes.

Wherever direct measurable data was not available, it was decided to validate such causes

by process observation or GEMBA analysis (Womack, 2011). The validation plans in

Tables 2.16 and 2.17 present the type of data available and possible analysis to be

performed for validation of all potential causes in the cause and effect diagram. Based on


61

these validation plans, each potential cause was validated and conclusions were made

regarding whether the cause is a root cause or not. The details of validation of the causes

are presented in the remaining part of the analyze phase. The final conclusions regarding

these causes are also included in the last column of Tables 2.16 and 2.17.

Table 2.16: Cause validation details for foot thickness

Sl.

No Cause

Observation / Validation

method Remarks

1 Wrong setting master Only one master used Not a Root cause

2 Wrong reading taken GEMBA observation Not a Root cause

3 Dressing frequency not OK Trend chart Root cause

4 Unwanted correction GEMBA validation Root cause

5 Untrained operator Only trained operator used Not a Root cause

6 Machine not taking

correction

GEMBA observation,

Backlash in feed mechanism

Root cause

7 Jerky movement of spindle Observation and data

collection

Not a Root cause

8 Component not getting

clamped properly in fixture

GEMBA observation,

Repeatability study.

Not a Root cause

9 Gauge GR&R not OK GR&R study Not a Root cause

10 Less stock for grinding Data collection Not a Root cause

11 Wrong correction for depth Data collection, GEMBA

validation

Root cause

Chapter: 2

62

Table 2.17: Cause validation details for plunger taper

Sl.

No Cause

Observation /

Validation method Remarks

1 Grinding stock Regression analysis Root cause

2 Pre-part material taper not OK Regression analysis Root cause

3 Bent plunger GEMBA validation, data

collection

Not a root cause

4 Grinding feed DOE Root cause

5 Grinding speed DOE Root cause

6 Alignment between headstock

and tailstock not OK

GEMBA validation,

observation

Not a root cause

7 Machine not taken the proper

correction

GEMBA validation, data

collection

Not a root cause

8 Unnecessary correction GEMBA validation Not a root cause

9 Dressing speed DOE Root cause

10 Dressing feed DOE Root cause

11 Dressing depth DOE Root cause

12 Wrong taper correction GEMBA validation Not a root cause

13 Wrong input by operator GEMBA validation Not a root cause

14 Gauge R & R not OK Gauge R & R study Not a root cause

15 Centre hole not OK in plunger GEMBA validation Not a root cause

16 Dressing frequency not OK DOE Root cause


63

As per the cause validation plan, causes related to input raw material characteristics

like grinding stock, pre-part taper and pre-part runout were validated by regression

analysis. For this purpose, 116 components in a batch were selected and data were

collected on grinding stock, pre-part taper and pre-part runout. As all these variables are

continuous, the effect of these input dimensional characteristics on the plunger taper needs

to be validated by a multiple regression analysis. If the regressors are linearly related, the

inference based on a regression model can be misleading or erroneous (Draper and Smith,

2003; Montgomery and Peck, 1982). When there are near linear dependencies between the

regressors, the problem of multicollinearity is said to exist. Hence, before performing the

multiple regression analysis, the variables were tested for multicollinearity.

Multicollinearity can be studied through the variance inflation factor (VIF). From the VIF

of the regression analysis (Table 2.18), it is evident that multicollinearity is not present in

the data. In the regression analysis, the p-value for ANOVA and for the regressors

‘Grinding stock’ and ‘Barrel roundness’ were found to be less than 0.05 confirming that

these two has a significant impact on plunger taper.

Table 2.18: Output of regression analysis

Predictor Coefficient SE of

coefficient t – statistic p – value VIF

Constant 0.0011716 0.0001156 10.14 0.000 -

Grinding stock 0.007544 0.002225 3.39 0.001 1.0

Barrel roundness 0.04107 0.01424 2.88 0.005 1.1

Pre-part runout -0.008743 0.008053 -1.09 0.280 1.0

Chapter: 2

64

In the cause and effect diagram, there were few causes related to machine/ process

parameters. During the team discussion about validation of these causes, it was understood

that at the time of installation of the machine, the machine manufacturer has only provided

an operating range for the machine/process parameters for the machine. Hence during the

installation of the machine, all the parameters were set by trial and error method. So far, no

scientific approach was adopted to optimize these parameters. Hence it was decided to

conduct a DOE in the improvement phase to identify the optimum levels for the machine/

process parameters.

(iv) The improve phase

Now, as per the plan in the analyze phase, it was decided to conduct a DOE with the

machine / process parameters. After a detailed discussion with the technical personnel of

the process, the parameters (factors) identified for experimentation were dressing

frequency, dressing feed rate, dressing depth, grinding feed and grinding stock. It was also

felt that the interaction of dressing frequency with grinding stock and dressing depth are

important to be estimated. As in the previous case, it was decided to experiment all the

selected factors at three levels. One of the levels for the selected factors was identified as

the existing level and the remaining two levels were selected based on operational

feasibility and cost considerations. The factors and respective levels are presented in Table

2.19.


65

Table 2.19: Factors and their levels for experimentation - foot face grinding process

Sl. No. Factor Level

1 2 3

1 Dressing frequency 30* 40 50

2 Dressing feed rate 80 100* 110

3 Dressing depth 30 35* 40

4 Grinding feed 1 2* 3

5 Grinding stock 35 45* 55 *Existing levels

For conducting a factorial experiment with five factors each at three levels require

quite a large number of experiments to be conducted, which would be very costly and time

consuming exercise. Also, the interest was only to estimate two interactions; it was decided

to use fractional factorial experimentation using orthogonal arrays for conducting this

experiment (Wu and Hamada, 2011). Estimating the main effects of five factors and two

interactions require 27 experiments to be conducted using the orthogonal array L27(313).

The design layout for experimentation was prepared by allocating the factors and levels to

the L27(313) orthogonal array.

It was decided to replicate each experiment two times. As per the design layout,

experiments were conducted in a random sequence and data were recorded. These data

were analyzed using Taguchi’s S/N ratio method. Since the response selected is of foot

thickness, nominal-the-best type of S/N ratio was selected for analysis. These S/N ratio

values were calculated for all the 27 experiments and further analysis was carried out. The

S/N ratio values and the raw data were analyzed separately to identify the important factors

of the process. ANOVA was carried out for the S/N ratio values and the significance of all

Chapter: 2

66

factors and interactions were tested. From the ANOVA table presented in Table 2.20, it is

clear that the factor grinding feed and the selected interactions are significant at 5 percent

level. From the main effect and interaction plots of S/N ratio, the optimum factor level

combination was identified and is presented in Table 2.21.

Table 2.20: ANOVA table for S/N ratios - foot face grinding process

Source DF Seq. SS Adj. SS Adj. MS F p - value

Dressing frequency 2 21.269 21.269 10.635 2.9624 0.109

Dressing feed rate 2 12.637 12.637 6.319 1.7602 0.233

Dressing depth 2 20.702 20.702 10.351 2.8832 0.114

Grinding feed 2 58.64 58.64 29.32 8.1671 0.012*

Grinding stock 2 5.173 5.173 2.587 0.7206 0.516 Dressing frequency x Dressing depth 4 84.962 84.962 21.241 5.9167 0.016*

Dressing frequency x Grinding stock 4 100.528 100.528 25.132 7.0006 0.010*

Error 8 28.72 28.72 3.59

Total 26 332.631 * Significant at 5 percent level.

Table 2.21: Optimum combination from the experiment– foot face grinding process

Sl.No Parameters Optimum level after DOE 1 Grinding feed - in microns 3

2 Dressing frequency - in numbers 50

3 Dressing depth - in microns 35

4 Grinding stock - in microns 35

5 Dressing feed rate - in mm/min 70


67

The optimum factor level combination identified was considered as the solutions to

the causes related to machine/ process parameters. For the remaining root causes, a

brainstorming session was performed with all the people working on the process and

solutions were identified. After identifying the solutions, a risk analysis was performed to

study the possible risk associated with the selected solutions. From the risk analysis, it was

concluded that there is no risk associated with any of the selected solutions. Hence an

implementation plan was prepared for all the solutions with details of responsibility and

target date for implementation. All the solutions were implemented as per the plan and

results were observed.

(v) The control phase

During the control phase of a Six Sigma project, the mechanisms for sustainability

of the achieved results are introduced in the process. Due to tool wear present in the

process, even after the improvement actions, the collected data plotted against time showed

an increasing trend in case of foot face thickness. Because of this upward trend in

dimension, the operators tend to give corrections in the process for avoiding components

produced outside the specification limits. Since tool wear cannot be eliminated from the

process, the process adjustments by the operators are necessary for the process. The

unscientific way of incorporating corrections in the process was resulting in high variation

in the process. Hence the team decided to standardize the process adjustments through

scientifically established techniques. Two questions need to be answered during any

process correction is that, when to give correction and how much correction to be given.

These two questions can be answered by the nested ANOVA analysis (Montgomery and

Runger, 2007) and Beta correction technique (Taguchi, 1981). Hence it was decided to

Chapter: 2

68

apply the beta correction technique for standardizing the corrections. For this purpose, 256

consecutive components from the process were selected without any adjustment and the

foot thickness was measured. A nested ANOVA was performed on this data to identify

when a significant change happens to the process. From the nested ANOVA, it was found

that the p-value between 8 components is significant and hence corrections are required

after 4 components. Now, the corrections/ adjustments to be given in the process have to

be calculated based on beta correction technique. In beta correction method, it is suggested

that instead of giving the full correction, a fraction of the correction to be given after taking

measurement on the machined component (Taguchi, 1981). Hence, as per this method, if

‘X’ is the measured dimension of the component and ‘T’ is the centre of the specification,

then

Correction , ) where

0,

1 , , (2.1)

Traditionally, for implementing the beta correction method, a ready reckoner used

to be prepared and displayed near the machine as per the above formula. Since these

operations were done on a high precision and advanced technology machine, in this case,

these process adjustment details were incorporated in the machining program itself. As per

this program, corrections were made to the process, whenever necessary. This has helped

the process to maintain the components with less variation in the process. After

implementation of this solution in the process, a sample of size 1000 units was collected

over a period of two months time and the process capability was evaluated for the foot

thickness. From this analysis, the observed PPM total was found to be 5000. Similarly the


69

data for taper was evaluated and the observed performance was found to be 5128 PPM. As

discussed in section 2.2, all the measures were introduced for standardizing and

maintaining the results. Table 2.22 summarizes the improvements achieved from this

study. Figure 2.13 represents the comparison plunger taper variation before and after the

study.

Table 2.22: Comparison of results - before and after the study

CTQ Measure Before After

Foot thickness PPM 35333 5000

Sigma Level 3.31 4.08

Plunger taper PPM 51333 5128

Sigma Level 3.13 4.07

Figure 2.13: Plunger taper - before and after

Data3.52.82.11.40.70.0-0.7

AFTER.

BEFORE.

Each symbol represents up to 12 observations.

Chapter: 2

70

2.3.2 Observations

This study discusses the successful implementation of Six Sigma methodology

integrated with Taguchi’s beta correction technique in an automotive supplier company. As

a result of this study, the first pass yield improved significantly in the process. A cost-

benefit analysis was performed with the help of the finance department for this study. The

annualized savings due to reduction in repair, scrap and tool cost was estimated to be

around US$87,000.

These results of the effect of implementation of Beta correction technique with Six

Sigma initiatives make a novel contribution to the growing body of literature. This study

contributes uniquely by elucidating the synergistic impact of Beta correction for greater

effectiveness of Six Sigma programmes in engineering industry. There has been little, if

any, research so far about the effects of simultaneous implementation of Six Sigma and

Beta correction. This is highly relevant to machining processes of engineering and

manufacturing organizations where tool wear is a major issue. If the process is influenced

by tool wear, it necessitates corrections in the process for compensating for tool wear. This

is a case study where Taguchi’s Beta correction technique was integrated in the control

phase of Six Sigma methodology to have proper monitoring of the process with tool wear.

This has helped the process to reduce the variability.

Similar case studies were tried in reducing rejection and rework in a honing process

of a manufacturing company, reducing patient waiting time in pathology and outpatient

departments of a super-specialty hospital and improving the quality of road construction in

a wind mill company. All these four studies gave significant results in the respective

processes. Some of these studies demonstrate the application of Six Sigma in very


71

uncommon areas. The details of these studies are discussed in Gijo and Antony (2013),

Gijo, Antony, Hernandez and Scaria (2013), Gijo and Sarkar (2013) and Gijo and Scaria

(2010).

During these studies, data and its analysis gave confidence to the people and the top

management of the organization for taking decisions about the process. Six Sigma projects

were identified with respect to the needs of the business/customer, and the problems

addressed were of highest priority to the organization. Hence these improvement projects

were leading towards significant results.

2.4 Shortcomings in Six Sigma implementation

Quite a few articles were published in various aspects of problems of Six Sigma

Implementation (Breyfogle, 2005; Hahn, 2005). Some of these publications discussed the

critical factors for successful projects (Gupta, 2005; Hariharan, 2006). This study is aimed

at identifying the potential failures during Six Sigma implementation in various

organizations so that the lessons learned from these failures can be beneficial for future

practitioners.

During this study, implementation of Six Sigma methodology was observed in

various organizations. The organizations considered for this study were from various

sectors of business including Automotive, Engineering, Chemical, Non-Conventional

Energy, Software, Information Technology (IT), Information Technology Enabled

Services (ITES), Business Process Outsourcing (BPO), Healthcare Services, and

Hospitality Services etc. Information was collected from 1320 projects across 82

organizations. These include public and private sector organizations, covering small,

Chapter: 2

72

medium and large enterprises. The number of projects executed in these organizations

varies from one to 145. The researcher interacted with the Six Sigma project teams in

various capacities viz, trainer, consultant/ guide or evaluator for these projects. A detailed

study of these projects as well as interaction with the project teams helped to identify the

weak areas in project selection, execution and further implementation as detailed below.

2.4.1 Project selection, CTQ identification and target setting

Six Sigma projects are linked directly to business metrics and bottom-line of the

organization. But, quite often, projects are selected from isolated pockets of some

departments, which no one considers as important to the business/ customer. Such projects

fail to get support of the people working in the process, leading to failure of the project. In

quite a few projects, the Critical to Quality (CTQ) tree diagram gives large number of

CTQs. In the name of prioritization of CTQs, most of the teams pick up the easiest CTQs

to make improvements in the project. As a thumb rule, many a times, project teams

identify the target as 50 percent reduction in defects for projects. In most of these projects,

after achieving the 50 percent improvement, the customer/ business requirements were not

met; still the project was added to the list of success stories in the organization.

2.4.2 Having numerical quota for project and trained manpower

In quite a few organizations, the top management decides target for number of

projects to be completed, BBs/GBs to be trained and the total savings to be achieved

through Six Sigma initiatives in a year. All these unrealistic targets put the people in the

organization to unnecessary hardships leads to diluted implementation of Six Sigma

methodology.


73

2.4.3 Defect definition

Defining a defect can be difficult for any business process that isn’t a standardized

manufacturing process, especially for innovative business processes (Harry and Schroeder,

2000). A defect in service industry resulting in customer dissatisfaction can range from a

simple misunderstanding of terminology to complete disrespect by the employee towards

the customer; these two outcomes receive the same value in Six Sigma methods because

they both are defined as a single defect.

2.4.4 Variation in training and certification of GB/BB

Since Six Sigma is only a methodology, organizations customize the concept during

implementation. Hoerl (2001) discusses the structure and curriculum for Black Belt

training. But, quite a few times, in the name of customization, training/ consulting

organizations dilute the curriculum and duration of training for Six Sigma BBs and GBs.

The GB training available in India ranges from lecture based three day training without

project to seven-day sessions requiring completion of projects. The BB training ranges

from eight days to 20 days class room training followed by one to three projects. Hence the

depth of knowledge and understanding of Six Sigma concepts varies significantly among

the certified BBs/ GBs.

2.4.5 Training on statistics to non-statisticians

Six Sigma courses on BB/GB are generally stuffed with high-level statistical

analyses that require deeper understanding of statistics. Lot of statistics is generally taught

to the BBs/ GBs through Minitab software. Even without understanding the meaning of

output generated through the software, the BBs/ GBs complete the projects successfully

and make attractive presentations.

Chapter: 2

74

2.4.6 High attrition rates for BBs/ GBs

As more and more organizations implementing Six Sigma methodology, there is lot

of demand for qualified BBs/ GBs in the job market. Hence, maintaining the certified

BBs/GBs is always a challenge for organizations. In these organizations, even the Six

Sigma movement takes a natural death.

2.4.7 Validation of cause - the GEMBA way

Most of the Six Sigma projects, the team carryout extensive brainstorming sessions

and identify quite a few potential causes that can create variation in CTQ. These potential

causes are to be validated based on data collected from the process. Due to difficulty in

collection of data and its analysis, many teams try to use GEMBA as the validation method

for the causes. This is totally against the philosophy of ‘mile deep’ problem solving

through Six Sigma methodology.

2.4.8 Accuracy of collected data

Most of the organizations use some software like SAP for capturing the information

of their operations. In such situations, the data corresponding to all activities in the

organization are generally recorded in the computer system. If the data available in the

system is not reliable, the actions taken based on its analysis also may not be valid. This

can lead to improper selection of root causes and wrong action in the process.

2.4.9 Application of tools and techniques

During the Six Sigma BB training, quite a few statistical techniques are discussed as

a part of the curriculum. Once the training is completed, most of the BBs will be searching

for application of these high level statistical tools in their projects, because many of the


75

BBs are having a misconception that, for the project to be ‘good’ or ‘acceptable’ there need

to be plenty of statistical analysis. In this way, statistical techniques are often

indiscriminately used in the project.

2.4.10 Sustainability of results achieved

The Six Sigma project team works hard to achieve the results and ultimately the

project is closed and the concerned people are certified as BB or GB. In quite a few

organizations, there are no mechanisms available even to monitor the sustainability of

achieved results for the completed projects.

2.4.11 Estimation of financial benefits

Once the Six Sigma project is completed, most of the project teams closely monitor

the result of the project for few weeks or a maximum of two months before closing the

project. Based on the results obtained for two months, the teams estimate the financial

savings for a year and close the project. Once the project is closed, there may not be any

system for capturing the savings from the project and finally all the estimated financial

savings will not reflect in the balance sheet of the company.

2.4.12 Six Sigma or Lean which one to embrace?

When Six Sigma started its implementation almost 30 years before, there used to be

a general confusion among industries and individuals regarding which concept to be

implemented in their organizations - Six Sigma, TQM or something else (Marash et al,

2004). After 30 years down the line, now the debate is going on between Six Sigma or

Lean or Lean Six Sigma. Especially in service/ financial sector, Six Sigma is giving way to

Lean and Lean Six Sigma.

Chapter: 2

76

2.5 Conclusion

Even though these case studies gave wonderful results, completing them was not an

easy task. Like any other initiative, there were a few people who were initially opposing

the movement in the organization, but later got convinced about the methodology after

observing the results. The projects were also affected by the attrition problem in the

organisation. Quite often in-depth data collection from the processes was extremely

difficult. Sometimes even people are hesitant to give detailed data due to the fear of

adverse effect on individuals, if something wrong was identified.

We hope, these studies will encourage industries to use the Six Sigma method to

deal with difficult problems, especially where causes are not obvious. Successfully

avoiding the common mistakes/ shortcomings will yield long-term benefits for an

organization and accelerate its march towards becoming best in its class. The key to

success is to identify these challenges early and take robust corrective actions to nip the

problems before they become an issue (Marash et al, 2004). A high level of technical

ability is required for the benefits to be gained, but the success of such projects also

depends on the correct aims being set, the correct team being selected, and the correct

atmosphere being created for the project; thus there are both technical and management

challenges to ensure Six Sigma project success. With responsible professionals’ conscious

efforts to avert potential shortcomings, Six Sigma has all the ingredients to continue to be

successful for many years to come.


77

References

1. AIAG (2002). Measurement systems analysis, reference manual, 3rd edn., Automotive

Industry Action Group, Southfield, Michigan.

2. Banuelas R., Antony J. and Brace M. (2005). An application of Six Sigma to reduce

waste, Quality and Reliability Engineering International, 21(6), 553-570.

3. Brady M. and Allen T.T. (2006). Six Sigma literature: A review and agenda for future

research, Quality and Reliability Engineering International, 22(3), 335-367.

4. Braglia, M., Carmignani, G. and Zammori, F. (2006). A new value stream mapping

approach for complex production systems, International Journal of Production

Research, 44(18), 3929-3952.

5. Breyfogle III F.W. (2003). Implementing Six Sigma: smarter solutions using statistical

methods, John Wiley, New York.

6. Breyfogle, F. W. (2005). 21 common problems (and what to do about them), Six Sigma

Forum Magazine, 4(4), 35-37.

7. Bryman, A. and Bell, E. (2006). Business research methods, Oxford University Press,

New Delhi.

8. Cooper, D.R. and Schindler, P.S. (2006). Business research methods, Tata-McGraw

Hill, New Delhi.

9. Draper, N. R. and Smith, H. (2003). Applied regression analysis, 3rd edn., John Wiley,

New York.

10. Firka, D. (2010). Six Sigma: an evolutionary analysis through case studies, The TQM

Journal, 22(4), 423-434.

Chapter: 2

78

11. Gijo, E.V. (2011). 11 ways to sink your Six Sigma project, Six Sigma Forum

Magazine, 11(1), 27-29.

12. Gijo, E.V. (2012). Shortcomings in Six Sigma Implementation: an experience in Indian

industries, Science and Society, 10(2), 109-116.

13. Gijo, E.V. and Antony, J. (2013). Reducing patient waiting time in outpatient

department using Lean Six Sigma methodology, Quality and Reliability Engineering

International, (wileyonlinelibrary.com) DOI: 10.1002/qre.1552.

14. Gijo, E.V., Antony, J., Hernandez, J. and Scaria, J. (2013). Reducing patient waiting

time in a pathology department using the Six Sigma methodology, Leadership in

Health Services, (in press)

15. Gijo, E.V. and Sarkar, A. (2013). Application of Six Sigma to improve the quality of

the road for wind turbine installation, The TQM Journal, 25(3), 244-258.

16. Gijo, E.V. and Scaria, J. (2010). Reducing rejection and rework by application of Six

Sigma methodology in manufacturing process, International Journal of Six Sigma and

Competitive Advantage, 6(1/2), 77-90.

17. Gijo, E.V. and Scaria, J. (2013b). Simultaneous implementation of Six Sigma

methodology with Beta correction technique in automotive supplier process,

International Journal of Advanced Manufacturing Technology, (Under revision).

18. Gijo, E.V., Scaria, J. and Antony, J. (2011). Application of Six Sigma methodology to

reduce defects of a grinding process, Quality and Reliability Engineering International,

27(8), 1221-1234.

19. Grant, E. L. and Leavenworth, R. S. (2000). Statistical quality control, 7th edn., Tata

McGraw-Hill, New Delhi.


79

20. Gupta, P. (2005). Innovation: the key to a successful project, Six Sigma Forum

Magazine, 4(4), 13-17.

21. Hahn, G.J. (2005). Six Sigma: 20 key lessons learned, Quality and Reliability

Engineering International, 21(3), 225-233.

22. Halliday, S. (2001). So what exactly is Six Sigma? Works Management, 15(1), 15.

23. Hariharan, A. (2006). CEO’s guide to Six Sigma success, Six Sigma Forum Magazine,

5(3), 16-25.

24. Harry, M. and Schroeder, R. (2000). Six Sigma: The breakthrough management

strategy revolutionizing the world’s top corporations, Doubleday, New York.

25. Hoerl, R. W. (2001). Six Sigma Black Belts: what do they need to know? Journal of

Quality Technology, 33(4), 391-406.

26. Hoerl, R.W. (2004). One perspective on the future of Six Sigma, International Journal

of Six Sigma Competitive Advantage, 1(1), 112-119.

27. Ishikawa, K. and Lu, D.J. (1985). What is total quality control? The Japanese way,

Prentice Hall, Englewood Cliffs, NJ.

28. Jiménez, E., Tejeda, A., Pérez, M., Blanco, J. and Martínez, E. (2012). Applicability of

lean production with VSM to the Rioja wine sector, International Journal of

Production Research, 50(7), 1890-1904.

29. Kaushik, P., Khanduja, D., Mittal, K. and Jaglan, P. (2012). A case study: application

of Six Sigma methodology in a small and medium-sized manufacturing enterprise, The

TQM Journal, 24(1), 4-16.

30. Landis, J.R. and Koch, G.G. (1977). The measurement of observer agreement for

categorical data, Biometrics, 33, 159-174.

31. Lee, T. W. (1999). Using qualitative methods in organizational research, Sage, CA.

Chapter: 2

80

32. Mahanti, R. and Antony, J. (2009). Six Sigma in the Indian software industry: some

observations and results from a pilot survey, The TQM Journal, 21(6), 549-564.

33. Marash, S.A., Berman, P. and Flynn, M. (2004). Fusion management: harnessing the

power of Six Sigma, Lean, ISO 9001:2000, Malcolm Baldrige, TQM and other quality

breakthroughs of the past century, QSU Publishing Company, Virginia.

34. Montgomery, D.C. (2002) Introduction to statistical quality control, 4th edn., John

Wiley, New York.

35. Montgomery, D. C. and Peck. E. A. (1982). Introduction to linear regression analysis.

John Wiley, New York.

36. Montgomery, D.C and Runger, G.C. (2007). Applied statistics and probability for

engineers, John Wiley, New York.

37. Montgomery, D.C. and Woodall, W.H. (2008). An overview of Six Sigma,

International Statistical Review, 76(3), 329 - 346

38. Pande, P., Neuman, R. and Cavanagh, R. (2003). The Six Sigma way team field book:

an implementation guide for process improvement teams, Tata McGraw-Hill, New

Delhi.

39. Phadke, M.S. (1989). Quality Engineering using robust design, Prentice Hall,

Englewood Cliffs, New Jersey.

40. Stewart, R.A and Spencer, C.A. (2006). Six Sigma as a strategy for process

improvement on construction projects: a case study, Construction Management and

Economics, 24(4), 339-348.

41. Taguchi, G. (1981). On-line quality control during production, Japanese Standards

Association, Tokyo.


81

42. Taguchi, G. (1988). Systems of experimental design, Vols. 1, 2, UNIPUB and

American Supplier Institute, New York.

43. Womack, J. (2011). GEMBA walk, Lean Enterprise Institute, USA.

44. Wu, C.F.J. and Hamada, M. (2011). Experiments – planning, analysis, and parameter

design optimization, John Wiley, New York.

45. Yin, R. K. (2003) Case study research: design and methods, 3rd edn., Sage, CA.

chapter 2 - inflibnetshodhganga.inflibnet.ac.in/bitstream/10603/35278/10/10_chapter2.pdf ·...

Documents