6 Sigma is an improvement methodology developed to reduce variation in processes. Like Lean, Six Sigma is like a toolbox containing many tools. Although a small part of the tools are also used in Lean, it mainly consists of Problem Definition, Problem Solving, Data Analysis and Statistical Methods. Six Sigma is mostly used to solve complex problems of quality. For example, the diameters of some of the bolts produced are not within the desired tolerances, which can be easily solved with the the Six Sigma Method.




Variation is basically not consistently getting the same results from the output you expect from the process (quality, efficiency, etc.) There are two types of variation in a process:

  1. Centering Problem (left): When the desired parameter is above and / or below tolerances (or specification limits) it is called the centering problem. For example, the process runs slower than desired, the drill bit pierces above the set value, the fabric is darker than the color we want.

  2. Scattering Problem (right): The situation where the desired parameter is scattered over larger than desired area is called the scattering problem. For example, the cycle rate for the same product varies at different times, the color tones in different meters of the fabric are different.


There may be situations where both occur at the same time.


Variation exists in every process. The sources of variation may be due to known reasons (your journey to come home, which takes 30 minutes on average, took 45 minutes due to snow) or the natural oscillation of the process (sometimes it may take 30 minutes, sometimes 32 minutes).

For this reason, since it is known that the product desired to be produced as 30 microns in the example below cannot always be produced at this value, tolerances (specification limits) have been established.

When your process goes beyond these limits, it produces erroneous output. The more you reduce the process variability, the less likely you are to produce faulty output. This will increase customer satisfaction.




With 6 Sigma, the inputs causing these variations are detected and formulated as in the picture below.


Y = f (x) is used to define functions. First of all, the problems (Y and ys) are defined. Then, the inputs that we think affect this problem are defined. (xs) Of course, since many inputs are defined, it is necessary to prioritize them and reduce them to the numbers we can control. For this, the approach called Logic Filters is used. Logic Filters are developed to solve problems with 5 phases called DMAIC in 6 Sigma. With different tools used in each phase, many input (xs) mentioned above are eliminated and reduced to the controllable number that has the most effect on the output.




DMAIC is the name given to the unique problem solving management phases of 6 Sigma. These phases are called Define, Measure, Analyze, Improve, Control in English. DMAIC is the abbreviation of these phase names. Each phase has been designed to determine the most important inputs (xs) that affect the output (Ys) by using tools within itself. Many Operational Excellence problems can be solved if the steps are followed correctly.

We interpret the phases in as follows:


DEFINE: It is the phase in which the output and potential inputs of the problem are defined, the scope is determined and the project charter is created.

MEASURE: It is the phase in which the defined outputs and inputs are measured with accurate and precise measurement systems, the analysis of measurement systems (MSA) is performed and the accuracy and precision of the measurement results are ensured.

ANALYSE: It is the phase in which important factors that create variability in the output are determined by graphical and statistical analyses.

IMPROVE: It is the phase in which important factors affecting the output are improved and optimized.

CONTROL : It is the phase in which actions are taken to ensure the continuity of the improved process.