Industrial analytics thought leadership-series.
How will industrial analytics, Industry 4.0, IIoT and digital transformation change businesses? And ultimately the world? We ask industry leaders to share their view on these technologies, how we use them, and where they will take us.
Correlations may lead you on a wrong path
Industrial analytics is all about getting meaning from data. Data is speaking and analytics is the listening device, but you need a hearing aid to distinguish correlation from causality. According to Pat Whitcomb, design of experiments (DOE) is exactly that.
“Even though you have tons of data, you still have unanswered questions. You need to find the drivers, and then use them to advance the process in the desired direction. You need to be able to see what is truly important and what is not,” says Pat Whitcomb, Stat-Ease founder and DOE expert. “Correlations between data may lead you to assume something and lead you on a wrong path. Design of experiments is about testing if a controlled change of input makes a difference in output. The method allows you to ask questions of your process and get a scientific answer. Having established a specific causality, you have a perfect point to use data, modelling and analytics to improve, secure and optimize the process.”
The boost of the pc
Pat Whitcomb is currently working three days a week and he plans to retire in a year or so. He can look back at a professional life dedicated to design of experiments (DOE). The method is rooted in pre-WWII English agriculture. Pat was introduced to it while working as a young chemical engineer in R&D at General Mills in the ‘70s. This scientific approach faced some barriers at that time.
“The concept of DOE had a big boost in the ‘80s because of the personal computer. Computational power had set the limits before. With the PC we were given much more space for experiments to understand processes. I was an early adaptor of DOE, who had felt the frustration of the lack of computer availability and software tools,” he says.
In his personal life, Pat is curious and rather fearless in deconstructing things to see how they work. He fixes bikes, cars, bath rooms, and the heater in his house. He enjoys overcoming barriers, but he never did learn to fix the automatic transmission in his old car.
Analyst out of the equation
This curiosity and the will to experiment led Pat into the chemical industry. Those same traits drove him to develop software for DOE to start Stat-Ease Inc. Pat has been integral to the spread of DOE in industry. In the ‘80s the chemical industry had the earnings and the R&D spend to be the primary adopter. Nowadays pharma is leading the pack as the heaviest user of DOE, but food, oil, aviation, and electronics are also big users.
“Our whole purpose was to put the statistical tools in the hands of the subject-matter experts in the industries. We have tried to make it as user friendly as possible for the scientists to use it and take the analysts out of the equation,” says Pat Whitcomb. “The user should be in control and understand what is going on. Software like this should not be a black box, and yet it should not be ruined by too many details. It’s all about guiding the user step-by-step and making the DOE process as straight forward as possible.”
Pat Whitcomb sees a similar culture in Camo Analytics in terms of empowering the domain specialists.
“We are kindred spirits. We are all passionate people. We share the idea of making mathematics and statistics available to other scientists.”
Domain knowledge drives innovation
Today, there is enormous hype surrounding topics such as ‘big data’ and ‘artificial intelligence’. With this hype comes the belief that data alone will drive innovation. But it’s the domain knowledge from experts, paired with DOE, that moves new ideas forward.
“The use of industrial analytics and our software help companies remain competitive. The need for innovation cannot be answered with data alone. You have to build your hypothesis, test it, and qualify it to get there. This can only be done by subject-matter experts, working and creating data on the way,” says Pat Whitcomb.
“I am not an expert in AI, but I do not think AI will come up with a new drug. I see it as very sophisticated programming, and it will excel in routine project with a lot of repetitive tasks. True innovation is too complex,” says Pat Whitcomb.
Brief history of Design of Experiments
Ronald Fisher, a distinguished British statistician, developed the method in the 1920’s. He published the seminal “Design of Experiments” in 1935. During this period, most of the DOE work occurred (literally) in the field of agriculture. A pioneering study from the University of Minnesota made use of the statistical tools developed by Fisher to evaluate different types of barley. During World War II, a more sophisticated form of DOE, called factorial design, became a big weapon for speeding up industrial development for the Allied forces.
After the war George Box described how to generate response surfaces for process optimization. From this point forward, DOE took hold in the chemical process industry, where factors such as time, temperature, pressure, concentration, flow rate, and agitation are easily manipulated. Box co-authored a textbook, Statistics for Experimenters, that formed the basis for the first DOE program by Stat-Ease. The methods could then be incorporated in a menu-driven computer program that would make DOE easy for non-statisticians.
“The use of industrial analytics and our software help companies remain competitive. The need for innovation cannot be answered with data alone. You have to build your hypothesis, test it, and qualify it to get there. This can only be done by subject-matter experts, working and creating data on the way.”
Find the right solution for your analytics needs.
Get in touch if you have questions about our products, platform, how to get started or how best to address your analytics needs.