For many of you Design of Experiments (DOE) will be a new concept. And some of you who have done some reading or research into it might have been put off trying it, as it often comes across as complicated.
But it really isn’t! Put simply, DOE is a statistical methodology to help you decide what combination of experiments are best to run to understand how the inputs to your system affect the output that you care about. Rather than running every single combination of input at every single setting, DOE will help you pick the subset of those combinations that will answer the main questions you have about your system. For example, “does input A need to be high or low to maximise my signal?”, or “does input A and input B both need to be high at the same time for a robust signal?”, but equally important “does input C diminish my signal?”.
Ultimately DOE can be used to find the optimal set of inputs and their levels to achieve the output you are after, whether that is to maximise your yield, minimise your noise, or find a robust signal to variations in your experimental conditions.
In the following documents you will learn the background and fundamentals of what DOE is - a perfect place to start!
Rather than starting with what DOE is, we would like to take you on a journey of what it isn’t and dispel some assumptions on the way! Dive in here to learn more about the basics and fundamentals of DOE.
Doing a DOE does not mean you have to run 1000s of experiments - in fact quite the opposite. DOE is a campaign of iterative experiments each subsequent design being informed by data from the previous iteration. Each iteration can have different flavours, as you learn more about your system. To find out more about the different types of experiments you can do with DOE continue here.
Running an experiment really is a means to an end. What we all really care about is the data we generate at the end and the insights we gain. DOE and data modelling go hand in hand, and before you get worried about the maths, let Synthace take care of this bit for you. Read on to learn about the fundamentals of modelling and how it can really be quite simple.