What is six sigma?

Six sigma is a quality improvement methodology used to improve process capability by reducing process variation from specifications. This approach is also a business strategy to reduce defects and increase profit. The term sigma is symbolized by σ, a Latin Greek alphabet used to represent standard deviation. 

Six sigma is used to refer to philosophy, metric, or methodology. 

As a philosophy, six sigma is concerned with the continuous improvement of a process for significant defects reduction. Instead of reactive to solve the problem, six sigma is a proactive and preventive approach to problem-solving.

As a metric, six sigma strives for perfection and that defects should not exceed 3.4 defects per million opportunities (DPMO). The higher the process capability, the higher the sigma level.

Table 7. Six sigma performance level.

The normal distribution (bell curve) is a symmetric distribution of data set below or above the mean. It is used to analyze continuous data whose probability distribution forms a bell curve. The normal distribution is best used when data sets are grouped around the mean representing an equal data probability (positive and negative) from the mean.

Figure 13. Six sigma normal distribution (bell curve).

Six sigma methodologies

Six sigma uses DMAIC (define, measure, analyze, improve, and control) approach in problem-solving. DMAIC phases are built upon on the equation of;

Y = f (x)

Where Y is a;

–  Function of x
–  Process output
–  Result of process input

Where X is a;

–  Process input
–  Major driver of the problem

Define – this is the first step of the six sigma project. It defines the project, problem, goals, project team, and processes. Define phase has three steps;

1.  Define problem and project objectives

–  Define the problem
–  Identify customer requirements/critical to quality (CTQs)
–  Determine project goals

2.  Develop project charter

–  Document the project and goals
–  Scope the project
–  Identify the project team and their roles
–  Describe milestone

3.  Create a supplier, input, process, output, and customer (SIPOC) diagram

–  Identify the high-level process map documenting the input and output business process and stakeholders involved
–  SIPOC helps the team to get the overall view of processes incorporated in the project.

Table 8. Example of SIPOC diagram for corrugated box making.

Measure – this is the second phase of the project. It measures the size and scope of the problem or performance gap between the current and desired state. To determine the scope of the problem, there are steps involved to follow;

1.  Collect data

–  Understand the scope of the problem
–  Develop key questions that will help solve the problem
–  Gather only relevant information supporting project goals
–  Choose the best tool that presents data in an understandable manner

2.  Map the current process

–  Document the steps of the project from start to finish to visualize the entire process. a process map is the best tool to use

3.  Validate data

–  Ensure the measurement system used in gauging data is valid

4.  Measure the process capability

–  Utilize the collected data to measure the Y’s or the process performance to get the sigma level

Analyze – this is the third phase of the project. The analyze phase analyzes the collected data from the measure phase to identify the root causes of the problem or the X’s. The purpose of this phase is to determine key drivers of the problem (X’s impacting the performance of Y’s). Analyze phase has the following steps;

1.  Analyze the process map to determine sources of process variation

–  Review the process map for better insights about the ins and outs of the process
–  Process map helps the team brainstorm realistic ideas for improvement

2.  Identify the causes of the problem (X’s)

–  Develop a list of all possible causes of the problem
–  Use a tool like a fishbone diagram or failure mode and effect analysis (FMEA) to determine most likely key X’s
–  Choose key X’s

3.  Collect data on X’s

–  Gather enough data on selected X’s to prove its validity

4.  Determine the relationship between X’s and Y’s – Y = f (x)

–  Use a statistical tool like regression to determine the strength of correlation between variables – whether the selected problem (X) is the key factor of output performance (Y)

5.  Validate the X’s

–  Use the analysis tool such as hypothesis tests and inferential statistics to prove the key X’s
–  The goal is to get sufficient evidence to support the hypothesis
–  Interpret the result whether to reject or accept it

Improve – this is the fourth phase in the project. Its objective is to resolve X’s and find a solution to improve Y’s. Improve phase involves the following activities;

1.  Generate possible solutions to solve key X’s

–  Brainstorm potential solutions to address X’s and improve Y’s
–  Identify the operating limits of X’s
–  Analyze the cost-benefit of feasible solutions

2.  Validate solutions alternatives

–  Evaluate selected solutions using a tool like the design of experiments or pilot testing

3.  Choose the best solution and execute

–  Develop criteria metrics to assess the impact of the selected solution
–  Conduct cost-benefit analysis

Control phase – this is the last phase in the project. Control phase ensures that X’s are under control and that improvement is sustainable. Activities in this phase include;

1.  Establish and execute a control plan

–  Develop a control mechanism or system to regulate and monitor the performance
–  Ensure control plan in place to prevent failure
–  Engage process owners/operators to carry out a control plan

2.  Monitor performance

–  Establish a monitoring system to keep track of the performance
–  Act using statistical process control (SPC)
–  Document performance for record purposes

3.  Communicate business results

–  Report financial impact to stakeholders
–  Celebrate results

4.  Project sign-off

–  Disband project team members

Six sigma tools

To deliver each phase in DMAIC successfully, the following tools are needed;

Define phase

–  SIPOC diagram
–  Voice of the customer
–  Flow chart
–  Process map
–  Project charter

Measure phase

–  Measurement system analysis (MSA)
–  Data sampling
–  Process capability analysis

Analyze phase

–  Hypothesis testing
–  Inferential statistics
–  Why analysis
–  Control impact matrix

Improve phase

–  Brainstorming solutions
–  Cost-benefit analysis
–  Failure mode effect analysis (FMEA)
–  Measurement system validation
–  New process mapping

Control phase

–  Control plan
–  Statistical process control (SPC)
–  Mistake-proofing or “poka-yoke”

<< Previous topic

Next Topic >>

join the supply chain community

Get the supply chain insights straight in your inbox

You agree to our privacy policy when you sign up