The Six Sigma Approach

(Published originally in 2000 on web journal of Symphony Technologies,

reproduced in Quality World in Mar 2006)

Close your eyes and try to visualize that:

• Every train arrives on time and leaves on time!
• In a company employing 100 employees, no employee is late
throughout the year!
• All the passengers get their luggage at the destination.

Are we crazy? No! These are examples to illustrate the meaning of six sigma
quality level. Six sigma is one idea that is adopted by organizations to become
world class. It is aimed at reducing the defect levels in products and processes to
a level of less than 3.4 defects per million opportunities.

Evolution of Six Sigma

The six sigma approach was first introduced and developed at Motorola in early
1990s. Later in the mid-nineties, it was adopted by General Electric and Allied
Signal. According to Jack Welch CEO of GE “Six sigma is the most challenging
and potentially rewarding strategy GE have ever undertaken” GE listed
tremendous benefits of the approach in their annual report in 1997. These
benefits included:
• 10 fold increase in the life of CT scanner x-ray tubes at Medical
Devices
• Improved yield resulting in increased capacity without additional
investments at Superabrasives- their industrial diamond business
• 62% reduction in turnaround time at the repair shops of the railcar
leasing business
• Addition of 300 million pounds of new capacity . This was equivalent to
adding new plant!

In GE, Significant portions of the bonuses were linked to the introduction of six
sigma. Six sigma was originally centered around manufacturing improvements.
The reason for this was knowledge of the statistical tools in the manufacturing
functions and the ease with which we can quantify the benefits. However these
improvements were not readily seen by the customers. The approach was
therefore broadened to all business operations. The success of these companies
with the six sigma approach caught the attention of Wall Street making it a
popular strategy that is being adopted by many organizations worldwide. Indian
organizations such as Wipro, Hero Motors, Godrej-GE have also adopted six
sigma strategy to improve their businesses (reference- Business Today, Sept 22,
1999).
What is six sigma?
Six sigma is a highly disciplined approach used to reduce the process variations
to the extent that the level of defects are drastically reduced to less than 3.4 per
million process, product or service opportunities (DPMO). The approach relies
heavily on advanced statistical tools. While these tools have been known earlier,
these were primarily limited to the statisticians and quality professionals.

Sigma (9) is Greek letter that is used to describe variability. In statistical quality
control, this means “standard deviation”. Most of us may be familiar with the
normal distribution and its properties. We are aware of the properties of normal
distribution :

• 99.73% of the area lies within mean µ ± 3σ

• 95.46% of the area lies within mean µ ± 2σ
• 68.26% of the area lies within mean µ ± σ.

When we proudly mention that our process capability Cp is 1.33, our process
spread is ± 4 σ. This would mean and estimated defect rate of 0.0063% or 63
defective parts per million (PPM). Moreover, when we deploy processes in
production, the mean of the process can shift to the extent of approximately
1.5σ. In such case the defect rate will increase to a much higher value. This
would be about 6200 PPM! If the process capability is improved to a Cp of 2.0
the PPM level will come down to 0.002. With a shift of 1.5σ, the Cpk will drop
down to 1.5 and the number of parts defective will be about 3.4 PPM. A Cp of 2.0
corresponds to the process spread of ± 6 σ. This is shown in the figures below.

1.5 σ shift

LTL UTL LTL UTL

±3 σ ±3σ

±6σ ±6σ
Cp=2, Cpk=2, PPM<1 Cp=2, Cpk=1.5, PPM=3.4

Let us consider an example of an assembly with 30 parts and 5 steps. This

means that there are 30x5 or 150 opportunities for a defect to occur. If we find
that there are 100 defects in every 100 assemblies. This means one defect per
assembly or 1miliion defects in one million assemblies. This can be converted to
1000000 x (1/150) or 6666 defects per million opportunities (DPMO). This
approximately corresponds to a sigma level of 3.97. The sigma quality level can
be approximately determined using the equation (Schmidt and Launsby 1997)

Sigma quality level = 0.8406 + 29.37 − 2.221x ln( ppm)

Refer to the following table for relation between sigma quality level and PPM.
The PPM values are calculated considering a shift of 1.5 9.

σ PPM
2 308,537
3 66,807
4 6,210
5 233
6 3.4

Another concept that is used as a metric in six sigma is Rolled Throughput

Yield (RTY). Let us assume that a part goes through ten operations. At each
stage 99% parts are good and 1% are reject. It is not very difficult to calculate
that we get good 90.43 % parts at the end of the tenth stage. This means if we
start with a batch of 1000 parts, we get 904 good parts and scrap or rework 96
part! The RTY of the process is 90.43%.

While we are talking about the statistical basis of the PPM levels, six sigma is not
only about statistical methods. The backbone of a successful six sigma approach
is strong commitment of top management. This is like any other successful
programmes. Moreover, all improvements planned through six sigma projects
must have a direct benefit that can be measured in terms of improvement in the
bottom-line.

The Six Sigma Methodology:

The projects having large impact of customer satisfaction and significant impact
on bottom-line are selected. Top management of the organization has very
important role during selection of projects and leaders. The projects are clearly
defined in terms of expected key deliverables. These are typically in terms of
DPMO levels or sigma quality levels, RTY, Quality Cost etc.
In the overall approach, the actual problem is converted in to a statistical
problem. This is done by mapping the process, defining key process input
variables (KPIVs or ‘x’s) and key process output variables (KPOVs or ‘y ’s). The
power of statistical tools is used to determine a statistical solution. This is then
converted in to a practical solution.

Practical Statistical Statistical Practical

problem Problem Solution Solution
Six sigma projects go through four phases:

The number of ‘x’ s go on getting eliminated using various statistical and other
tools. This is as if the variation is getting reduced as it passes through a funnel of
the six sigma methodology. This is sometimes called the breakthrough strategy.
Charter, Project tracker
Define
All possible Xs
Measure Process Map, C&E
Matrix, MSA,Cpk

Analyze FMEA, Multi-vari

Improve Design of Experiments

Few
‘x’s Control SPC, Fail-safing, Control Plans

The cause and effects relationship between the KPOVs (‘y’s) and KPIVs (‘x’s)
gets clearer as the project goes through the five phases. Control plans are
documented before the closure of the project so that gains are sustained. The
project leaders must demonstrate that the key deliverables of the project are
achieved and demonstrated.

Strategies for six sigma introduction:

There are three different strategies adopted by different organizations:

1. The six sigma organization

2. The six sigma engineering organization
3. Strategic selection of six sigma projects

1. The six sigma organization

In this strategy, the whole organization is trained on six sigma philosophy and
methods. The training is of varying depth for various levels. Six sigma serves as
motivational device and also as a metric. Goals are defined in terms of sigma.
While the organization can have a common language of six sigma, large
resources are required for training. All improvement ideas are likely to be
credited to six sigma regardless of the approach actually used!
2. The Six Sigma Engineering Organization:

Here, the attempt is to develop skills in the Engineering functions. The project
objectives are usually based on new products, product changes or problem
solving. One of the advantages is the relatively higher level of educational and
technical background of the individuals that enables them to learn at a faster
pace. On the other hand, individuals from other functions do not appreciate the
efforts in absence of the awareness of the techniques.

3. Strategic selection of six sigma projects

The senior management sometimes feels that the current quality processes are
generally working well to achieve the overall strategic plan. Hence six sigma
tools and concepts are used to enhance the existing quality processes and
supplement the skills of the key people thereby making breakthrough
improvements. Six sigma projects are identified considering the:
• Strategic direction of the company
• Impact on the bottom-line
• Impact on customer satisfaction

The projects having large impact requires project leaders with high degree of
competence. Full time project leaders (sometimes called the black belts) are
selected to execute the project. Selection of candidates is critical for the success.
The project leaders go through in-depth training of six sigma approach and tools
and work full time on the project. The project is expected to be completed in
about six months. Typical savings expected from a black belt project may be of
the order of Rs.100, 00000. Projects of lesser complexity may not require full
time resource. Project leaders of such projects are chosen from the same
functional area. These are sometimes called green belts. They also go through
training in the six sigma concepts and tools. Training duration is usually less than
that for black belts.

This approach requires lesser resources for training that can be customized. The
organizations adapting this approach must allocate the best people as project
leaders. Some of the potential failure modes of this approach are:
• Trained engineers tend to get isolated
• Communication barrier due to lack of common language
• Failure to develop management understanding

Criticism on six sigma:

There has been a lot of criticism on six sigma. Many feel that the tools used in
the approach have been existing and in reality there is nothing new in sigma In
short, six sigma is old wine in new bottle. The criticism is true in a way. However,
others feel that while the tools existed earlier, quantification of the metrics and
direct relation to the bottom-line makes it somewhat different. Another major
criticism is about the price tag attached to the six sigma training. Typical price tag
of a four week black belt course is $ 25000! This is prohibitive for most of the
small and medium scale organization. However, training costs may come down
in the near future. Another criticism on six sigma is it results in to short term
improvement projects. This is countered by its proponents that although one
project is of six months duration, the black belts take up next project also train
others in six sigma methodology. As more black belts are trained and large
number of projects are completed; the organization achieves long term benefits.

Concluding Remarks:
Six sigma is powerful approach achieve breakthrough improvements in
manufacturing, engineering and business processes. The approach relies heavily
on advanced statistical methods that complement the process and product
knowledge to reduce variation in processes. It is new way of doing business that
would eliminate the existing defects efficiently and would prevent defects from
occurring. Different strategies are used by organizations to introduce and deploy
six sigma approach. Each of these strategies has advantages and potential
failure modes that must be addressed and avoided.

References:
1) Forrest W Breyfogle III, Implementing Six Sigma- Smarter Solutions Using
Statistical Methods published by John Wiley and Sons, Inc.
2) Drug Sanders and Cheryl Hild, A Discussion of Strategies for six sigma
implementation, Quality Engineering, Vol 12, Number 3, 2000.
3) Gerald J.Hahn and Necip Doganaksoy, Roger Hoerl, The Evolution of six
sigma, Quality Engineering, Vol 12, Number 3, 2000.

About Author:

Hemant Urdhwareshe is Director, Institute of Quality and Reliability, Pune. He is

• Qimpro Foundation Silver Standard 2003 Award Winner
• First ASQ Certified Six Sigma Black Belt in India
• ASQ Certified Quality Engineer, Quality Manager and Reliability Engineer.
• The only person in India with 4 or more certifications from ASQ (as on Mar 2005)
• Senior Member ASQ