Author's Choice

Six Sigma- Definition, Principles and Methodoloy

Six Sigma
Six Sigma is a business process that allows companies to drastically improve their bottom line by designing, monitoring everyday business activities in ways that minimize waste and resources while increasing customer satisfaction. Six Sigma is defined as a philosophy, methodology and a breakthrough strategy to solve problems. According to Breyfogle III (2001), the classical definition of Six Sigma as defined by Fontenot, Behara & Gresham (1994) is that it is a way to measure the probability that companies can produce any given unit of product (or service) with only 3.4 defects per million. It means catching defects 99.9997 percent of the time (Blakeslee, 1999). Sigma is a symbol in the Greek alphabet and is used to indicate the variation about the average of any process. When sigma is applied to a business process the sigma value is a results metric indicating how well that process operates. The goal of Six Sigma is not to achieve six sigma levels of quality but the most visible impact of Six Sigma is improved profitability and immediate jump in profit margins.

SIGNIFICANCE OF SIX SIGMA STANDARDS
Companies gain a competitive edge by reducing defects in their industrial and commercial processes. A defect is anything that blocks or inhibits a process. For example, when a machine operator fails to change a gear during maintenance, it affects the process and not the product, but yet it is a defect in the overall picture of the company’s profit. Six Sigma applies to products and services and not to the companies. It creates specific improvement goals for every process and forces organizations to reexamine the way in which work gets done. According to Harry, 1998, companies must implement Six Sigma in everything that they do. Each sigma value creates an exponential reduction in defects. This results in the reduction of cycle time, reduction in testing and inspection times and customers are more satisfied. As stated in Breyfogle, Cupello and Meadows (2000), one Polaroid executive says that “Six Sigma is not only about reducing the number of defects per million opportunities, but it is a systematic approach to reduce variability in a process through assimilation and organization of information”. Even today, four sigma companies can produce six sigma products through huge amounts of rework, but they can’t raise their prices to recapture these costs since they must price their products competitively .

SOME SIX SIGMA PRINCIPLES AND PERFORMANCE METRICS
The choice of what to measure is crucial to the success of Six Sigma implementation. Metrics generally when used in the right context give a direction for the most important activity in the given situation. As adopted from Prasad (1998), the 5 most critical metrics are:

1) Cost to quality- There are two contributory elements that affect the cost of quality:
cost-to-ensure-quality (c-t-e-q) and (b) cost-to-correct-quality (c-t-c-q). Cost-to-ensure-quality is the cost of doing things right, the cost of doing right things, and the cost of preventing mistakes. Typical examples include the cost of training, establishing procedures, insurance, preventive or contract maintenance, planning activities and analyses of performance data, etc. The cost-to-correct-quality is the cost incurred because of doing things wrong, the cost of doing wrong things, and the cost of inspections to discover mistakes committed earlier.
C-t-q Effectiveness = ({cost-to-ensure-quality} / {cost-to-quality}) *100
If the c-t-q effectiveness number is close to 100, the company is doing things more right than wrong. The goal of the cost of quality system is to reduce the cost of quality to the lowest practical level.

2) DPMO- Defects per million opportunities gives additional insight into a process by including the number of opportunities for failure within the calculations. It avoids giving focus to the rate of defects only at the end of the process. A DPMO metric enables a uniform measurement for the process and not only the product.

3) Rolled Throughput Yield - Reworks within an operation have no value, but rolled throughput yield measurement can give visibility to those process steps that have high defect rates and/or rework needs. The multiplication of the process operation step yields gives an estimate of the overall process throughput yield. It can be calculated from the number of defects per unit by
Yrt= e-DPU where DPU=defects per unit

4) Capability Index- This index is the ratio of the specification width to the distribution width of the process
Cp = USL-LSL/ 6 USL = Upper Specification Limit
LSL = Lower Specification Limit

SIX SIGMA ASSUMPTIONS
According to Blakeslee, 1999, there are several assumptions in six-sigma quality measurement.

a) The normal distribution of data is most common in real-world applications too.

b) The process mean shifts by exactly1.5 (sigma)

.c) The process mean and the standard deviation are known and that the process capability parameters Cp and Cpk are point values

d) Defects are randomly distributed throughout the units and that there is independence between part/process steps.

These pitfalls and assumptions do not invalidate six-sigma quality programs, though managers should be aware of them before designing a quality process.

THE SIX SIGMA CHANGE METHODOLOGY
Every project has a process or a design problem in search of a solution. The Six Sigma Methodology tries to find these solutions and increase the bottom line.

The “Breakthrough Methodology”
This concept takes executives through the maze of business, technology, manufacturing and quality, production and delivery system issues. Six Sigma is a disciplined method of using very rigorous data gathering and statistical analysis to pinpoint sources of errors and ways of eliminating them. Specific tools are used to reduce costs, improve capacity and margins and process transactions in a shorter period of time with much fewer mistakes. As adopted from Harry & Schroeder (2000), there are eight fundamental steps or stages in applying the “Breakthrough Methodology” to a company’s business. They are Recognize, Define, Measure, Analyze, Improve, Control, Standardize, and Integrate. The 4 core phases can be detailed as follows:

a) Measure Phase – Here, companies try to understand the nature and properties of data collection with analysis tools and measurement systems to pinpoint process defects and capability.

b) Analyze Phase – This phase converts the practical business problems into focused statistical problems. Information critical to explaining defects is isolated.

c) Improve Phase – Using the Design for Six Sigma (DFSS), this phase takes the products or services from the beginning and configures them to produce six sigma results. E.g. Motorola designed a process to produce a virtually defect-free pager.

d) Control Phase – This phase ensures that mistakes do not get repeated, by continuous monitoring of processes in the system.
It is very important for companies to note that they should have a culture that looks at the data with the right skills, which is taught in the Strategy.

SIX SIGMA QUALITY PROGRAM – A SIX SIGMA PROJECT IMPLEMENTATION
In 1988, Motorola, Inc. was the first company to receive the Malcolm Baldrige National Quality Award, for which it had vigorously pursued a quality management program called “Six Sigma”. Involved in their efforts was an improvement process that is used by all employees to improve the quality of products, services and processes. This process can be customized to the specific organization, but it follows a generic pattern –
1) Define products and services – Define products and services that are provided to both internal and external customers including information, consulting and follow-up.

2) Identify customer requirements – Identify and determine expectations of customers for each product and service, in measurable terms

3) Compare the product with customer requirements –Identify gaps between what the customer is receiving and what he/she actually gets.

4) Describe the process – Provide a detailed description of each process. Flowcharts and process charts can be used in this step.

5) Improve the process – Evaluate each process in terms of its value and correlation to other processes. Some processes can be eliminated and others added.

6) Measure quality and productivity – Attempt benchmarking “best-in-class” companies and organizations for targets for quality and productivity improvements.

MOST SIGNIFICANT ISSUES AND TRENDS IN SIX SIGMA
Although Six Sigma has been one of the most effective management tool for a host of companies in different sectors of the industry in recent times, companies need to place the Six Sigma concept in an appropriate context to reap the maximum benefits from it. This section deals with the most significant issues related to Six Sigma.
1. Hype versus reality
• Long term PPM (parts per million)= 3.4- There is nothing outside 3 sigma (Gauss photo on Ten Deutsche Marks bill, Shewhart 1931, Deming 1982, Neave 1990). Gauss model of normal distribution approximates 3 sigma limits to match actuality 99.73% of the time

• Process variation is half that of specification range or Cp=2- Once Cp=1 is surpassed, it is almost impossible to achieve Cp=2, in presence of instability (Shewhart 1931) and target problems

• 1.5 sigma shift is assumed to arrive at 3.4 PPM- Amount and type of shift is a matter of reality and not of assumption.

• Six Sigma can handle any problem situation- There is a noticeable absence of reliability, multivariate and robustness methods.

• Six Sigma is a philosophy, methods and strategies- It is basically a statistical methodology. It is to be combined with other engineering strategies like effective communication, the forgiving principle (Bajaria and Copp, 1001)

• Six Sigma emphasizes the teachings of statistics as a science of conformance- Problems can be equally effectively solved if it is used as a science of indication.
The usefulness of the six sigma philosophy and breakthrough strategy can be limited if we are not conscious of the above limitations. Hence, the best solution is to adapt six sigma to the company’s particular problem-solving methods.

2. Kano Model of customer satisfaction
According to Thomaz Pzydek (2001), Noritaki Kano developed the model of the relationship between customer satisfaction and quality.

The obstacle of and the confusion within companies about the true measurement of customer satisfaction and loyalty within the six sigma framework can be explained by the Kano model of customer satisfaction. This model shows that there is a basic level of quality that customers assume that the product will have. Note that the entire basic quality curve lies in the lower half of the chart, representing a feeling of dissatisfaction in the customers. The expected quality line graphs expectations that are actually considered by customers as a measure of quality of service provided. The customers are thus not satisfied if their expectations are not met. Excited quality is something more than expected. E.g. In my project on online website deployment at Mahindra British Telecom, India, I incorporated additional features like free web space available for buying something above the rate of $5 on the site. This increased the number of hits many fold.
It is not enough to simply track competitor progress, but also customer satisfaction levels need to be judged, since expectations are influenced by outside factors as well. For e.g. the manufacturing revolution has sparked a revolution in the service and the computer industry about six sigma implementation.

3. Right question, right activity and right metric
The forcing of Six Sigma activities in the company can sometimes lead to the creation and mandatory use of metrics that have no long-term or short-term use. All six sigma metrics are beneficial for tracking, but they are not valuable for every situation. There can be a lot of waste of time and effort in some cases. E.g. Rolled throughput yield is not something that should be measured when there is a troublesome product or process. It takes a lot of time and money to collect the data necessary to estimate rolled throughput yield on a process of even most modest size. Six sigma quality level metrics can be very deceptive since they bear an inverse, non-linear relationship to defect rates (see Fig.6). Higher sigma quality does mean lower defect rates, but the relationship is nothing near to linear. As stated by Breyfogle III, 1999, an improvement from 3 to 4 sigma is not the same as one from 4 to 5 sigma, which means that a unit change in sigma does not correctly reflect unit change in the process capability. Hence a company needs to be cautious in using quality level metrics like PPM, Cp and Cpk to measure improvement activities in their processes.

4. Six Sigma beyond manufacturing
In the evolving, highly competitive global marketplace, consistent and recurrent customer satisfaction is essential for long-term survival. Degree of competitiveness of most products, whether they are consumer goods or goods for the defense industry, largely depends on exceeding the customer’s expectations. Quality to the customer means improved fits and clearances, no defective parts, shiny paint, reduced number of parts, improved quality comfort and superb performance, all at the same time.

4.1 Six Sigma in customer satisfaction
Customer satisfaction is achieved not through a single act, but a coordinated array of actions, each contributing a useful and interesting dimension towards an artifact’s overall performance (Behara, Fontenot & Gresham, 1994). Customer satisfaction is a multi-stage process rather than a single-stage process. Customer satisfaction means meeting the customers’ needs, at the right time, and in the quantity, price, and performance they want.
It is even more difficult to reach a level of Six Sigma in customer satisfaction. As adopted from a case study in Behara, Fontenot & Gresham, 1994, in independent customer satisfaction surveys conducted in 1991 and 1992, researchers at Service Strategies International, a Dallas-based customer satisfaction research company measured the satisfaction levels of customers with various high-tech manufacturing firms. By providing ratings on 32 attributes related to quality of service, product performance, field service representatives, and company image, customers indicated what they expected of the ideal high-tech manufacturing company. Then, customers rated their satisfaction with the client company and/or one other competitor considered to be “best in class” on the same 32 attributes.
Based on the findings of this six sigma analysis on customer satisfaction, the client company appeared to be in accordance with an improved level of average sigma ratings. Although the client company continued to improve its customer satisfaction ratings from year 1991 to 1992, six sigma levels were difficult to achieve as customers’ expectations continually changed (usually increasing). While the client company makes products at near zero-defects level, its customers’ satisfaction is also a function of support services such as product/ engineering support. Process simplification is essential to reduce the number of defects and thereby increase customer satisfaction. Current sigma levels in these areas are well below six sigma levels. The value of a loyal customer and the cost of a lost customer are two strong reasons to pursue the goal of zero customer defections.

4.2 Design for Total Quality
Design for quality (DFQ) programs are commonly directed towards achieving 100 per cent prevention and 0 per cent inspection goals. DFQ means building quality control into the product (using error-proofing techniques, or statistical means such as six-sigma) when the product is first conceived and designed. “Product quality is governed by the teams’ choice of 7Ts (talent of the work force, tasks, teamwork, techniques, technology, time, and tools)”. Designing for total value (DFTV) is a powerful technique that allows work-groups to determine systematically the total value of the product over its lifetime in conjunction with appropriate analysis tools like Six Sigma. There is no question that significant, linear improvements to any relatively new process are readily achievable. The question is how to discover a different process to obtain increased productivity in a specific operation. According to Prasad, 1996, there is a three-step procedure for selecting the right process-

1) Establish criteria, set minimum productivity goals, and identify benefits.

3) Adapt the concepts posed by the outside company operations.

In the event of new business processes, there is a great opportunity to achieve massive gains in productivity by starting with a tabula rasa—borrowing the best pieces, parts, processes, and technologies from all over and integrating them in entirely new ways of doing things (Prasad, 1996). The objective of such an exercise is increasing productivity—primarily by reducing “cycle time”—an unexpected additional benefit is often not just a reduction of the product defect rate but also a consistent improvement in the overall quality of the base product or service that is supplied to the customer.

Popularity: 1% [?]