Design of Experiments, or short “DOE” is a controlled experiment in which the experimenter is interested in both quantifying and optimising the effect of a number of input variables on one or more output variables.
Optimizing is the process of finding the optimal setting of each input variable which gives a required output.
Input variables are causing the variation, are also called factors.
The output variables being measured are called the responses.
Whoa. Now what does that mean?
In practical terms.
Let’s say that we have a process that has following inputs:
Material Temperature 100-150 deg – 50 choices
Tool temperature 10-28 deg – 18 choices
Process speed 100 – 275rpm – 175
Knob 1 position 0-100% -100
Knob 2 position 0-100% – 100
Cool time 12-25 min – 13
Hold time 20-50 min – 30
We have 7 factors here to work with.
Doesn’t sound like much right?
How many choices do we have here actually?
Not for setting that work (your technician on the floor might figure it out by playing with inputs and looking at outputs). We want the best possible settings.
We have 50*18*175*100*100*13*30 choices.
That is 614 250 000 000 choices. It is too complicated for technical guys to find the best choices.
The answer to this is DOE.
