In this section we cover the groups of tools that will help you improve your processes through modelling it through the tools of Design of Experiments DoE
DoE is a tool which helps you see how a process works, which things act together (interact) and which are independent and how they align to create the output you see,
In 5 Videos we explain :
The overview of DoE
What are interactions
How to set the input levels in your experiment
How to control your experiment
How to interpret you results
and what you will get as an output of the experiment.
We are holding a workshop on this event in January 2020, please register you interest at CottonInnovations@gmail.com if you are interested in knowing more
"Brad helped me a few years ago on DoE where we used the techniques to thrift the number of factors and then run the main experiment. It was great to see the outcome but even better to see the engineers turn from sceptics to converts. A powerful technique to solve the tough problems helped by a great coach in Brad!"
"Great to see you doing videos Brad. Interesting and engaging as always"
Can I run a DoE without a numeric input and output as I do not have anything to measure?
Technically no. There is ways around this please reach out and speak to me if you would like to know more as to how we run these experiments
Can I run a fractional design in place of a full Factor DoE what will be the problems?
Yes you can but factional designs will have limitations in what they can tell you about the process
Can I run a factional design in place of a full design if I do not understand the process?
Yes but this can be very misleading. we never advise this practice. As fractional designs do not have the power to tell you the real issues, they can get confusing it plan and interpret and they cannot see all the interactions.
I run a machine which produced many parts at the same time. Can I do a DoE on this machine?
Yes, please reach out to us as these take careful planning and can take a long time to conduct the analysis if you do not set the machine up properly
I work in an office, Can DoE be used in an administration process?
Yes, these are tricky to run, but you can run DoE's in these environments
In this section of Videos we discuss tools we can use to define a process. this section covers DoE and how we can use it.
Designing DoE (Design of Experiment) is a tool for understand how things interact and contribute to the final output.
In this series of 5 Videos we discuss
what is a DoE,
how to create one,
what needs to be considered and
how we interpret the results.
The DoE shows the strength of the main effects (the factors we chose to adjust) and the factors which interact. In this video we discuss using a cup of team what is are interactions
Setting the levels in an experiment are as important as running the experiment. Set them to large and the you may lose control. Not enough and the noises effects of the experiment may confound your results
Every DoE is open to large opportunities for failure due to experimental poor disciple, the effect of these are discussed in this video
Having completed the DoE how you interpret the results is key to understanding that you have controlled the experiment and are able to find better setting for the process
In this section of Videos we discuss tools we can use to define a process. this section covers tools we can use to see a process and what us affecting it through mapping.
In this video we discuss how a process can be mapped using the CNX tool and its power to see the relationship between noises and Controls and Inputs which when combined create our outputs
In this video we discuss how a process can be mapped using the fishbone (Ishikowa or Cause and Effect) tool and its power to see the relationship between the 3 factors we can control and the 3 factors which are more of a constraint in and process.
In this video we discuss how a process can be mapped using the f5 Why and compound 5 Why tool to take you from the effect we see in failure to the root cause of the failure.
Fix root cause and the problem is solved, fix the effect - nothing is solved