Application of the process optimisation

What is the objective?

Six Sigma is an approved management method based on statistical studies to systematically analyse your process. The goal is the improvement of the process capability and the elimination of variances of performance variables. Process solutions, which can have much higher savings on average can be developed and installed with relatively low expenditures. For this reason, the application of Six Sigma is very effective and efficient because the “jump-effect” is optimally tied together by improvement projects and the process of continuous improvement.

What are my tools?

For your statistical data analysis, I use the tools of the Six Sigma toolbox in order to reach the demanded process- and product quality optimisation. It’s a matter of:

• statistical methods to optimise processes for technical, medical or economical applications
• statistical analysis based on different classes of measurement data
  – like numerical measurement data
  – or attributive measurement data (e.g. text like product names or different machine settings)

What are typical applications?

Hypothesis tests

Using statistical methods, it can be determined to what extend process changes would have a real significance for the result. (E.g. evaluation of the influence of paper machine clothing at trials or evaluation of modifications etc.)

Identification of root causes

Which process changes lead to the variations of the performance- and quality properties, e.g. in case of disturbances?

With the help of statistic, “cause and effect” relations between process input variables and performance- and quality variables can be found. Just shifts in dewatering in the Former- and in the Press section have a significant influence on the process capability as well as the paper quality at paper production. These changes often run stealthy and over several years. However, when analysing historical measurement data, these shifts will be discovered. That said, the reasons often still lie in the dark. Two-dimensional analysis with only one potential input variable, which are often based on graphs, can only describe this insufficiently. With the help of the statistical data analysis many potential process input variables can be analysed together and the key performance input variables can be investigated according to their strength of influence and their direction of influence (positive or negative correlation).

With this knowledge, valid process optimisation measures can be implemented. In the future, key performance variables can be regulated and controlled. As a consequence, the process can be optimised and stabilised sustainably.
So the solution to optimise your process can probably be found in your database!

Statistical designed trials (Design Of Experiments DOE)

DOE are to be used in trials, where the strength of influences and the direction of influences shall be found for several input variables as well as for the best settings if required. Advantages:

• statistically examined results despite of a minimum number of trials
• iidentification of interactions between the process input variables

Training and Coaching

Training sessions at or outside your facility to teach statistical methods for process optimisation, for example:

• custom-tailored training sessions to use Six Sigma tools for Statistical Process Optimisation
• Six Sigma Belt trainings (with certificate, if required)
• Coaching of Six Sigma projects

More examples for applications and publications