What is DOE?

In general usage, design of experiments (DOE) or experimental design is the design of any information-gathering exercises where variation is present, whether under the full control of the experimenter or not. However, in statistics, these terms are usually used for controlled experiments. Design of experiments (DOE) is a series of tests in which purposeful changes are made to the input variables of a process so that we may observe and identify corresponding changes in the output response.

It is a valuable tool to optimise product and process designs, to accelerate the development cycle, to reduce development costs, to improve the transition of products from research and development to manufacturing and to effectively trouble shoot manufacturing problems. Today, Design of Experiments is viewed as a quality technology to achieve product excellence at lowest possible overall cost.

 

Purpose of Experimentation

  • Comparing Alternatives. It helps us to compare two or more alternatives and choose the best one. For instance, choose the best vendor who supplies the raw material.
  • Determining which input variables are most influential on the output or response variable. For example, what are the significant factors amongst raw material, temperature, process time, operator and etc.?
  • Determining where to set the influential controllable input variables so that output is near the nominal requirement. For instance, If temperature is the most influential factors, what is the temperature which results the best quality product.
  • Determining where to set the influential input variables so that variability in the response variable is small.
  • Determining where to set the influential input variables so that the effects of the input variables which are out of our control are minimised.

 

Benefits of DOE

A typical experiment which is usually used to show the advantage of DEO is called Catapult experiment. The main questions in this experiment are that how far does the catapult throw the ball? How can we adjust the catapult to hit a target? Which one is more important? Stop angle, Rubber band tension, hook, …?

 

DOE

 

As a result of a simple DOE, it is reported that it took about 200 man hours and the experimental results showed that predicting the ball landing site was accurate within 15 inches, whereas the design of experiment approach requires 6 man hours and it was accurate within 3 inches. It shows how DOE can save resources to hit the target sooner, more precise, and more efficient.

 

Implementation

The main steps to implement an experimental design are as follows.

  1. Define problem statement and experimental objectives. For example, an aircraft manufacturer is interested to maximise paint adhesion to aluminum surface of aircrafts. Different primer paints and painting methods are considered to find the best solution.
  2. Select the output (Response) variable. As the output variable, a finished paint was applied and the adhesion force was measured.
  3. Select the input variables (factors) and their levels. For instance, Painting Method with levels “dipping” and “spraying” and primer type with types 1, 2and 3.
  4. Select the experimental design such as Full factorial design, fractional factorial design, Taguchi design by considering sample size (number of replicates), the selection of a suitable run order for the experimental trials, criticality of errors and etc.
  5. Execute the experiment and collect data. It is important to use a data collection form to collect the data and make notes of observations, other factor settings, etc. during the experiment.
  6. Analyse the data and draw statistical and practical conclusion using different statistical software such as Minitab
  7. Validate the statistical conclusion in practical situation by setting the process inputs at the practical results and see if the process output responds as expected.
  8. Document and summarize the results and conclusions, in tabular and graphical form, for the report or presentation on the study.

 

How CBIS can help you

Please contact us if you need more details on how our expert team can assists you in training and implementation of DOE and lean six sigma projects.