Six Sigma is a disciplined process, a scientific methodology, a business strategy and a set of techniques
to increase the performance of business processes, by eradicating defects and decreasing variation.
Although major elements of the statistical and logical toolset that we use in quality management were developed
between the 18th and 20th century (by Euler, Gauss, Franklin, Pareto, Gosset, Taylor, Shewhart, Neumann, Deming, Juran, and many others), the Six Sigma
framework and principles were introduced by Bill Smith at Motorola in 1986. In the following decades, especially in the 1990’s, many multinational
companies started to use it, in order to improve the performance of their business processes, and thus increasing prestige and saving billions of
The name Sigma comes from the Greek letter Ϭ (sigma), which means the standard deviation of a statistical
population in mathematics. The intention of Six Sigma is to improve quality and business processes, and to eliminate defects, by decreasing
this standard deviation (variation of processes).
To drive results, Six Sigma not simply means the usage of mathematics, but the forming of focus projects
with management commitment and the devotion of project members. Conducting and closing a successful Six Sigma project brings many benefits for
various fields of the company, or even for the complete supply chain:
- Reduction of process variation.
- Decreasing the number of defects, increases prestige.
- Improvement of business performance (better deliverables, quality, speed, response time, etc.).
- Increased customer satisfaction.
To complete the task, the sigma principle always transfers the practical problem into a mathematical
problem, then solves the mathematical problem, and transfers this solution into a practical solution.
The basic principle of 6 Sigma (Source: qMindset.com; Tibor Kliment - Master Black Belt)
By using various tools from the Six Sigma toolset, the project team is working on to find:
- The interaction between the practical (pragmatic) and the mathematical problem.
- The transfer function between the mathematical problem and the mathematical solution.
- The translation of the mathematical solution into a practical one.
During the analysis of the possible transfer functions (correlation, regression, etc.), the team tries to identify the key inputs that affect the output on the largest scale.
6 Sigma fundamental shift (Source: qMindset.com; six-sigma-material.com)
The major intention of Six Sigma is to reduce variation of outputs (what the customer detects, notices)
by focusing on the inputs (that we are able to directly influence). By identifying and improving key input factors, the variation of outputs
decrease, leading to better performance and improved customer satisfaction. A process is Six Sigma capable, if it has so low standard deviation,
that it fits 12 times (6 + 6) into the specified tolerance range (USL – LSL), and being properly centered, meaning a Cpk value of 2.0. If this is
the real capability of the process, it causes "only" 3.4 defects out of one million.
The correlation of various sigma levels, Cp, yield and reject ppm (Source: qMindset.com)
In order to properly accomplish a project, the Six Sigma methodology has a pre-defined framework, which is
the so called DMAIC that stands for Define, Measure, Analyze, Improve and Control. Remark: the DMAIC is used for process improvement, while
there is Design for Six Sigma (DFSS), which uses the DMADV (Define, Measure, Analyze, Design, and Verify) framework.
In the following section, we focus on the major steps of the DMAIC framework, and the toolset we use in
|The DMAIC approach of 6 Sigma
||Project charter / contract
||Problem statement (identification of "pain")
Definition of project limits (in the scope
and out of the scope) and schedule
Goals and objectives (what we want, and what we expect from the project)
Forming the team
Management commitment (sign-off)
|Selection of CTQ (Critical to Quality) points
||Customer requirements - VOC (Voice of Customer), QFD (Quality Function Deployment)
characteristics of product / process
Identification of project metrics (Epsilons – Y's) in connection to CTQs
levels of Epsilons (USL, LSL)
SIPOC (Suppliers, Inputs, Processes, Outputs, Customers) analysis, or its inverse
Analysis of internal stakeholders (and suppliers if needed)
Estimation and consideration of project risks
Clear expectations from the project
Experts in the team are
necessary (statistics simply does not supplement expertise!)
Management agreement and focus
Goals must be achievable
||MSA (Measurement System Analysis)
||Conducting MSA to gather trustful information during the measurement (Resolution, Bias, Linearity,
Improvement of measurement system (if required based on the MSA results)
||Detailed process mapping and VS (Value Stream) mapping
Data collection plan to measure key Y
characteristics (outputs) and the X process parameters (inputs)
Measurement of performance metrics: Yields, OEE,
Process capability (Cp, Cpk), process sigma, etc. on the actual baseline, before improvement
Cause-effect diagrams, FMEA for potential effect of X's on Y's, Interaction of KPIV (Key Process Input Variables) and KPOV (Key Process
MSA study is necessary to verify if the gathered data is accurate
Systematic data collection plan
||Testing the hypothesis of cause-effect (X's and Y's)
Finalization of cause-effect diagrams,
Ishikawa, 5 Whys
Correlation, regression, ANOVA
(Analysis of Variance), Line plots, Box plots, Scatter plots, Interaction plots, 3D surface plots, Main effects plots, Contour plots, etc.
||Isolation of major causes (vital process inputs)
List of major causes to focus on (during
Understand major causes
The effect of major causes must
be proven instead of speculation
||DoE (Design of Experiment) and regression analysis for improvement and optimization of the most
vital input factors (X’s).
FMEA for the potential risks in connection to the implementation of modifications
||Implementation plan of change with timing, and responsible associates
Pilot (implement and
test it in small scale before spreading it on large scale)
Implementation of actions, counter-measures, parameter-changes with
Auditing the implemented changes for short-term assessment (tool, process change, training, etc.)
To eliminate major causes that are responsible for defects,
variation or delay
Actions must be as much error-proofed as possible (latter control does not supplement error-proofing)
To optimize vital input factors for the best output
Have we reduced the root cause on a measurable level?
||New process map and process monitoring
Definition of test frequency: 100% or
SPC (Statistical Process Control)
Design and documentation of controls (methods, technology): modified
CP (Control Plan), FMEA (Failure Mode and Effects Analysis) and operation sheets
||Start of modified control, selected indicators are put in place
Second MSA study (to verify
is the improved variation is measurable)
Finalized project documentation and project report towards the management
To implement the right control methods in the right place
The process owner has to maintain the control in the long-term
As soon as the project is closed, the management (and the process owner) signs it off. The
documentation of the project should be archived for later use (e.g. for lessons learned purposes).
But who is able to conduct a Six Sigma project? Here comes the relevance of various Six Sigma belts and
roles, summarized in the following table:
|Roles and belts of 6 Sigma
||Tasks / Responsibilities / Competency
||The owner of the targeted process to be improved by the project (e.g. manufacturing dep. leader,
financial officer, etc.)
|Sponsor & Leader
||Senior executive, who supports the initiative and the implementation of Six Sigma
||Supports a given Six Sigma project, by making management level decisions, providing resources,
||Low or basic knowledge of Six Sigma principles
Provides expertise in the project (e.g. expert
of welding, accounting supervisor, etc.)
|White Belt (rarely used designation)
||Very basic knowledge of Six Sigma principles, mostly gained on an online course
Usually being impacted by improvement projects, not being part of a problem solving team
||Basic knowledge of Six Sigma principles
Support projects, being part of a problem solving team
||Wide knowledge of Six Sigma principles
Leads green belt Six Sigma
||Higher knowledge of Six Sigma principles
Can be full-time professional
Leads black belt
level Six Sigma projects
Coaches green belts
|Master Black Belt
||Master of Six Sigma principles
Highly experienced full-time professional
level Six Sigma projects, and defines strategic initiatives for business improvement (usually at multinational companies)
and coaches black belts and green belts
Advisor of top level projects
After the 6 Sigma business strategy was spreading at Motorola and GE, many organizations started to offer various green and black belt certifications, but there is no one and only
official certifier. If you intend to make your business process improvement project, and want to get certified, search for companies or universities that provide trustworthy training services.
Source: qMindset.com; iassc.org; sixsigma-institute.org; asq.org
Keeping a business process on the optimal level, being near to the mean with the least possible variance
is essential, as customers feel the variance, not the mean!
This list contains a superb DMAIC process definition by Minitab (based on the introductory video of Companion by Minitab)
- Define: Define and Scope Project
- Measure: MSA and Project Baseline
- Analyze: Develop Y = f(X) Relationship
- Improve: Implement Proposed Improvements
- Control: Implement Control Strategy
A Six Sigma project always need to be SMART:
- Time bound
How to select from multiple Six Sigma projects which one to focus on? There are subjective and objective
answers. First of all, do not start a Six Sigma project for "low hanging fruits". If the solution is simple and evident, just do it. Of
course there are situations, when such a decision is not easy. Then comes the objective evaluation. Select these topics / issue to focus your
Six Sigma resources on:
- A project that brings high improvement of quality and prestige.
- A project that has very high P / L (Profit / Loss) factor.
- A project that improves the response speed of a process (e.g. faster product ordering, etc.)
- A project, which is achievable and real (choosing an improvement that is not real only takes time, but brings no benefits).
Don’t forget! Six Sigma is not only deemed a methodology of quality improvement. It’s intended to improve
any kind of business process from manufacturing to development, from finance to IT, and so on.
During the project definition phase one crucial point is to choose the right team, with experts in the
affected field. Six Sigma itself does not supplement expertise!
Source: qMindset.com; minitab.com/en-us/products/companion/