The name Six Sigma is a management program based on the quality control of variance (indicated by the Greek letter Sigma), which aims to bring the quality of a product or a service to a certain level, which is particularly favourable for the consumer.

First introduced by Motorola in the second half of the 80's by Bob Galvin and Bill Smith, it spread to other major companies such as General Electric, Toyota, Honeywell and Microsoft.

The objective of the methodology is to achieve such a control of the process to have only 3.4 defective parts per million, which leads to very restrictive limits on the variability of the production process.
This variability is to be so restricted that initially the common opinion was that it was impossible to reach and many believed that a three sigma would be acceptable. However, some industries have shown that such a goal is attainable. The Six Sigma methodology aims to the elimination of defects and waste, rather than simply improving the average performance. This ensures a direct impact on the customer, defined as a user of the process or product. This marks a step forward compared to a way of thinking was widespread before the introduction of the method, as explained by the following example.

Methodology

Method used to improve existing processes and in operation involves the following steps:
Defined. At this stage the working group should identify the process or product to improve, translate customer needs into requirements (known in English CTQ - Critical to Quality), identify the participants required for the project and develop a high-level planning. At the end usually requires the approval of the Directorate to proceed.

The main instruments used in this phase are:

• Flowchart
• Gantt diagram
• inside or outside the frame (out of box)

Measure. This phase consists in evaluating the performance level of the process or product in relation to the requirements identified. We must first identify and validate a proper system of measurement, then make the measurement. Finally, to compare the results with the required standards, we can calculate the current level of sigma.

Typically in use:

• Descriptive statistics
• Sampling
• Test-Repeatability Reproducibility (Gage R; R)

Analyze. Based on the measured data, apply statistical techniques to identify the causes of defects and to quantify the extent to which each case (or the whole of the cases) influences the condition studied and its variance.
The instruments used are numerous, some of the most common are listed here:

• Analysis of variance (ANOVA)
• Pareto histogram
• Regression
• Correlation
• Stratification

Improve. This is the stage where we propose and implement improvements to: until you understand fully the causes of defects. It is a prerogative of Six Sigma to avoid a jump immediately to solutions. We need to create a range of possible solutions by acting directly on the most important cases, analysis of costs and benefits, assess the need and modalities of any experiments or pilot tests.

To select the best solution, are used frequently:

• FMEA
• DOE (Design of Experiments - Progettazione degli Esperiment)
• Cost-benefit analysis

Check. In this phase watches the process in order to standardize and stabilize. It quantifies the extent of improvement and takes action to support the wording of the final procedures and training of staff. You can use the following tools:

• Chart Control
• Plan for quality control

From the initial stages, this method is indicated by the acronym DMAIC.
To these were subsequently added:

• Recognize: development of the strategy of intervention, (zero of DMAIC);
• Standardize: definition and coding best practices, (level six of DMAIC);
• Integrate: complementary actions to support the philosophy adopted (level seven DMAIC).