Shainin Tools PDF
Shainin Tools PDF
Shainin Tools PDF
Research Article
USE OF SHAININ TOOLS FOR SIMPLIFYING SIX
SIGMA IMPLEMENTATION IN QMS/ISO CERTIFIED
ENVIRONMENT– AN INDIAN SME CASE STUDY
Anand K. Bewoor*, Maruti S. Pawar
Address for Correspondence
*1Mechanical Engineering Dept.,Vishwakarama Institute of Information Tech.,Kondhwa
(Bk), Pune 411048, Maharashtra, India
2
Professor and Vice-Principal, B. M. I. T., Solapur University, Solapur Maharashtra, India.
E-mail: bewooranand@yahoo.com, drmspbmit@rediffmail.com
ABSTRACT
Six sigma for small- and medium-sized enterprises (SMEs) is an emerging topic among many
academics and Six Sigma practitioners over the last two to three years. Very few studies have been
reported about the successful applications of Six Sigma in SMEs. Main objective of this paper is to
examine the extent of usefulness of a simpler but not very frequently used methodology known as the
Shainin methodology for simplifying the implementing Six Sigma. To confirm whether Six Sigma
implementation is simplified, this paper highlights the comparison of three DOE approaches viz.
Classical, Taguchi and Shainin methodology.
A case study based research work done in ISO certified Indian SME, concludes that, Six Sigma
implementation process can be simplified by using Shainin tools and proper use company’s ISO/QMS.
KEYWORDS Six Sigma, Shainin Tools, QMS, Indian SMEs.
1. INTRODUCTION and new product and service development.
In recent past, academicians, practitioners Six Sigma relies on statistical methods and
and organizational researchers have the scientific method to make dramatic
recognized that the Quality Management reductions in customer defined defect
System (QMS) process and the Six-Sigma rates’’ (Linderman et al., 2003). Since its
process are disciplines that have a initiation at Motorola in the 1980s, many
powerful potential to affect an companies including GE, Honeywell,
organization’s ability to compete within Sony, Caterpillar, Johnson Controls etc.
an increasingly global and dynamic have adopted Six Sigma and obtained
marketplace (Falshaw et al., 2006). QMS substantial benefits (Pande et al., 2000).
certification (such as ISO 9000, TS Spectacular development of an
16949) demonstrates the capability of an organizational performance due to Six
industry to control the processes that Sigma implementation many companies
determine the acceptability of the product are reported in the published literature.
or service being produced & sold. These, Antony and Banuelas (2002) presented the
traditional QMS are having some key ingredients for the effective
limitations like methodological assistance introduction and implementation of Six-
etc. (Bewoor and Pawar, 2008). But new Sigma in manufacturing and services
QM methods continue to grow (Xingxing organizations as: Management commit-
Zu et. al., 2008) for example, Six Sigma, ment and involvement, Understanding of
which is ‘‘an organized and systematic Six Sigma methodology, tools, and
method for strategic process improvement techniques, Linking Six Sigma to business
of full factorial designs and fractional problems into three X’s, viz., the Red X,
factorial designs. A full factorial allows us the Pink X- the second most important
to test all possible combinations of factors cause(s), and the Pale Pink X – the third
affecting output in order to identify which most important cause(s). According to
ones are more dominant. A fractional him, these three Xs together account for
factorial tests just a fraction of the over 80 per cent of the variation that is
possible combinations. Though a very allowed within the specification limit and
popular tool, many engineers and quality when captured, reduced, and controlled,
practitioners find design of experiments these can eliminate this variation. Shainin
difficult primarily because of the developed techniques (Shainin and
complexity of having to create the Shainin, 1990; 1992a; 1992b; 1993a;
conditions for conducting the experiments 1993b; Shainin, Shainin and Nelson,
in an industrial environment where 1997) to track down the dominant source
interrupting production lines and changing through a process of elimination (Shainin,
machine settings may be sometimes 1993b), called progressive search. These
difficult and unproductive. techniques, also referred to as the Shainin
2.2 Shainin DOE Approach System for quality improvement,
An alternative to the Classical and developed over a period of over 40 years,
Taguchi experimental design is the lesser- are simple but at the same time powerful
known but much simpler Shainin DOE and easier to interpret and implement in an
approach developed and perfected by industrial environment. In a way, these
Dorian Shainin (Bhote and Bhote, 2000), may be considered as the non-parametric
consultant and advisor to over 750 equivalent of Taguchi’s DOE as they do
companies in America and Europe. not make any restrictive assumptions
Shainin’s philosophy has been, “Don’t let about population parameters. The Shainin
the engineers do the guessing; let the parts techniques are primarily known to
do the talking.” Shainin recognized the produce breakthrough improvements in
value of empirical data in solving real- eliminating chronic quality problems.
world problems. He introduced the These are highly effective in pinpointing
concept of Red X, the dominant source of towards the root cause and validating it.
variation, among the many sources of No statistical software was needed to
variation of a problem that inevitably analyze the data. In fact, Shainin DOE
accounts for nearly all the unwanted does not even require knowledge of
effect. difficult statistical tools. Simple operation
In fact, Shainin (Shainin, 1995; 1993b) like counts, additions, subtractions, etc.,
had classified all causes of chronic quality makes calculations relatively easy. In
addition, the success of the projects can an industrial operation. Applications of the
lead to a very positive effect on the morale Classical and Taguchi methods in various
of the employees in terms of convincing fields have been extensively researched. In
them that Six Sigma can be implemented contrast, the Shainin system has not been
without complex statistics and big jargons. extensively reviewed, academically, and
The subject of the Shainin methods is very very limited studies have been carried out
vast and this paper highlights the in this area.
applicability of only a few of the Shainin 3.1 Studies about comparison of
tools. However, there is a lot of scope for DOE approaches
more research on this methodology Bhote (2000) compared Shainin
particularly comparative research of some techniques with Design of Experiments
of the Shainin methods like Paired and Taguchi methods, in the context of the
Comparison and B Vs C Analysis vis-à- electronics industry and concluded that the
vis the more popular statistical tools like Shainin techniques are simpler, less
factorial designs and non-parametric costly, and statistically more powerful
testing. Although these methods are not than the other two. Logothetis (1990) also
necessarily the best, according to Steiner evaluated the Shainin techniques in
et al. (2008), the guiding principles of the relation to the Taguchi methods and
Shainin tools are powerful, and at least, in statistical process control methods.
combination, unique. Also, these tools are Verma, et al (2004) used a slightly
best suited for batch to high volume different approach to compare the
production. methods. In their study, three cases of
3. FINDINGS FROM VARIOUS Taguchi experiments were picked up from
CASE STUDIES ABOUT DOE the available literature and the Shainin
APPROACHES method was then re-applied to find out
Bhote and Bhote (2000) described these whether it had an edge over the other DOE
tools in their books, but there have been techniques. A comparison between
many criticisms regarding their claims and Taguchi and Shainin techniques in an
the tools described. Though, Nelson aerospace environment was offered by
(1991) and Moore (1993) criticized the Thomas and Anthony (2005). A few other
Shainin System as unsubstantiated and authors who have studied these techniques
exaggerated, Steiner, et al (2008), are of are Ledolter and Swersey (1997), De
the opinion that some of the ideas behind Mast, et al. (2000) and Steiner and
the Shainin System are genuinely useful. MacKay (2005). The Classical DOE,
Goodman and Wyld (2001) offered a case Taguchi DOE, and Shainin DOE are
study involving the use of Shainin DOE in compared with each other in Table 2.
Without taking educated guesses as to the as; BigY (Response i.e. Defective welding)
factors of real importance, authors have = f [X (Sources of variations i.e. CO2
suggested to adopt the Shainin Welding process)]. Therefore, new SSVs
Techniques. The Shainin’s Techniques are now related to CO2-Welding process
been employed to identify whether the are listed viz. Voltage, Current, Gas Flow
primary cause of shabby/defective and Wire Feed Rate. To check whether
welding lay within the process itself or any relationship exists within the
within the components used. This allowed identified parameters or not; data related
for a first stage filter to be employed that to all these parameters are collected (refer
cut down the factors to a manageable table 3), regression analysis is carried out
number. Key stages, in which Shainin and Graphs are plotted. Graph of Wire
tools were applied, are explained below. Feed Rate vs Current clearly shows the
Initial tool selected for measuring and positive relationship (refer figure no. 3).
analyzing the response was Product Hence, new SSVs identified parameters
Process Search, as of variations in the related to CO2-Welding process are now
identified suspectable sources of limited to: Voltage, Wire Feed Rate and
variations (SSV) i.e. input material Gas Flow.
parameter (as compared with their As the identified parameters were design
standard specification) viz. SSV-1. parameters of process and number of
Material Thickness (Specifications – 2.0 parameters are equal to 3 hence, it has
mm +/- 0.18), gets changed during been decided that, process characterization
processing. Data was collected for 100 analysis i.e. Full Factorial Analysis tool is
samples. to be used. All stages of full factorial
Observation 1 – It has been observed that, method are explained as follows,
minimum and maximum value of sheet Stage 0: As the response is attribute in
metal (raw material) thickness as an nature, consider current setting as the ‘–’
important input to production process setting and identify ‘+’ setting on the basis
belongs to same category of response. of experience on domain expert for each
Therefore, as per Product Process Search parameters (refer table 4).
method the end-count is zero. Hence, it Stage 1: To find out whether the
has been concluded that, SSV-1: Input parameters and the levels identified in
material parameter (i.e. Thickness) is not stage 0 are correct or not. Then, we have
creating problem. Next another to produced 3 batches in ‘–’ setting and 3
brainstorming session has concluded for batches in ‘+’ setting. Calculate D/d ratio,
characterization of CO2-Welding process if D/d ratio is >=1.25 and <3 then the
as process itself is yielding in to settings identified in Stage # 0 are correct
variations, which is required to be and we can go for Stage # 2. Accordingly
analysed. Hence, relation can be written trials are conducted; the results are
tabulated in table 5. D/d ratio is 0.4, which objective is lower the better. Using above
indicates that, the levels identified in stage equation, offline iterations are done.
0 are not correct. Hence, new parameters While doing iterations ‘+’ve settings are
levels are identified by considering earlier refereed as ‘1’, ‘-’ settings are referred as
‘+’ ve setting as new ‘-’ setting and new ‘-1’. Values some of the offline iterations
‘+’ ve settings for all parameters are and its calculated responses are tabulated
identified and set (refer table 6). Again in table 9. Then, experiments are carried
new trials are conducted and the results out using the levels of the parameters for
are tabulated in table 7. D/d ratio is 10, which responses are zero or less than zero
which indicates that the levels identified in and physical outputs are analyzed.
2nd settings are acceptable for further Response for setting in case of experiment
consideration. no. 9 (shown in same table) resulted in to
Stage 2: Construct factorial table and proper welding (considered as an optimum
collect the data for each combination in output).
the factorial table and quantify the Improvement Phase: Conclusions of
contributions of the interactions. earlier phase (identified optimum levels of
Table 8 shows factorial design and plan the parameters as shown in table 10) are
of experimentation. Accordingly used as an input to this phase. Once
experiments were performed, which optimum settings are set then, it is
resulted in to following important necessary to validate it. This was done, by
conclusions. using the Shainin B vs. C analysis, which
Parameter- A: As if we change from + is a confirmation tool to verify whether
level to - level then response increases by the actions taken have actually improved
2.5 points. the process (Bhote and Bhote, 2000). In
Parameter- B: As if we change from + this case, 6B vs. 6C, i.e., 6 batches (10
level to - level then response decreases by units per batch) with modification and 6
1.5 points. (10 units per batch) without modification
Parameter- C: As if we change from + (B – with modification and C – without
level to - level then response decreases by modification) was analyzed to validate the
5 points. improvement action, i.e., the modification
Stage 3: Make a simple mathematical of CO2 machine operating parameters
equation based on the contribution of (table 11).
significant parameters and arrive at the The data in table 12 exhibited the
optimal setting. responses with B and C conditions. As per
Y= 84.875 –3.125 A + 14.162B + 4.875C rule of this technique, the final analysis is
+2.625 AB – 4.375 BC – 7.125 CA + done based on the ‘end-count scheme’. In
7.625ABC this case, end count is 8 (greater than 6),
As response ‘Y’ considered is which confirms that identified root causes
shabby/defective welding hence, our are correct.
Further, the result clearly validates the mentioned in table 10. New specifications
improvement against the criteria not only helped to improve the quality
mentioned in table 13. The data has level but also productivity by reducing
exhibited no overlaps of the responses defect/rework rate and optimizing the use
with B condition and C condition. The of resource and time (e.g. Wire Feed Rate
conclusion being that the process has been from 10 Min/min to 6.5 Min/min and Gas
improved by changing the CO2 welding Flow from15 Lit/min to 14 Lit/min).
machine operational specifications as
Table 3: Data related to all these interactions among identified parameters
Sr. No. wire feed voltage current
1 50 27 40
2 55 13 90
3 55 16 100
4 55 18 80
5 55 20 100
6 55 22 110
7 55 22 110
7 55 25 100
9 55 28 90
10 55 30 90
11 65 17 100
12 65 19 100
13 65 23 100
14 75 30 160
15 80 20 150
16 80 27 140
17 100 26 190
The improvements identified were also • Procedure has been developed for
used to set the action plan for other periodic monitoring of CO2 welding
varieties of such components for machine operational specifications w. r. to
horizontal deployment. quality level of output.
Control Phase: • Implemented controls to make sure
The focus of the control phase is to sustain that the actions taken in Phase-III are done
the gains of the improvement phase. This forever.
is usually achieved by documentation and • All these modifications have been
standardization of the control measures. included as a part of Company-QMS
For controlling the process at Six Sigma procedure to ensure the reliability of Six
level, following actions were suggested. Sigma level quality of the process.
• Appropriate modifications have been The operational framework developed and
done in CO2 welding machine operating used in this research-work is described in
and training manuals. figure 4.
It clearly shows the major stages in the delivering certain objectives set by ISO
process integration and implementation. It such as: prevention of defects at all stages
shows initially the sequential nature of the from design through servicing; techniques
stages whereby the Six Sigma phases are required for establishing, controlling and
using appropriate imputes from company verifying process capability and product
QMS database to continently execute the characterization; investigation of the cause
project. The operational framework also of defects relating to product, process and
shows the stages in sequence whereby the quality system; continuous improvement
six sigma DMAIC phases are using of the quality of products/services.
accurately Shainin quality tools. From the results of case study based
5. DISCUSSION AND CONCLUSIONS research work we draw following
The aim of this project is to defeat the conclusions,
biggest “excuses” cited by SMEs as the i. The key phase of the DMAIC
reasons Six Sigma is not feasible, incurs methodology is the measure and
high costs and involve complexity of analysis phase. The tools and
implementation. In addition, it helps to techniques used in this phase
break down so many of the barriers that determine the success or failure of
stand in the way of individuals using the project to a large extent. In
statistical and/or unfamiliar problem both the projects, the Shainin tools
solving methods by acting as a step-by- have been very effectively used to
step guide. This research work focus on pinpoint the root causes and
use of Shainin tools specifically, as they validate the improvement actions.
are easy to understand, involves simple ii. No statistical software was needed
mathematical calculations (so that bottom- to be used to analyse the data. In
line people can also understand it very fact, Shainin DOE does not even
easily) and time required for training is require knowledge of difficult
also less, which is one of the important statistical tools. Simple operation
requirements of SMEs. During this case like counts, additions, subtractions
study, during use of Shainin tools, small etc., makes calculations relatively
samples of BOB and WOW pieces were easy. Therefore the training
sufficient to analyse the data as reported required for application of Shainin
earlier. A very important factor is that data tools is simple and requires less
collection was done for the project time (1-2 days).
undertaken online without disturbing the iii. In addition, the success of the
regular production. projects had a very positive effect
Thus in short, we can understand that, use on the morale of the employees in
of Shainin tools for simplifying Six terms of convincing them that Six
Sigma implementation can provides an Sigma works without complex
appropriate methodology for SMEs for statistics and big jargons.