The Power Of DOE: Producing High- Quality Plastic Bottles Using Minitab .

The Power of DOE: Producing High-Quality Plastic Bottles Using Minitab Gil Farias Slide1 of 50 2019 Minitab, Inc.

A Case Study of Design of Experiments (DOE) by MINITAB Applied on definition of Color in High Quality Plastic Bottles Slide 2 of 50 2019 Minitab, Inc.

Title comes here Some instructions about our slides from now on. It is important to understand and memorize All Slides have the following structure: 1- Title: All slides have a Title and a subtitle 2- Eventually we have hyperlinks on red dot CONTENT 3- Content: All slides have a content in the middle 4 – Take away: All slides have a take away, important message about the content 5- Return Key: All slides have a Return Key, to return to Summary Slide (presentation mode) Always the Take away considers some key words and important things to remember. Key Word TAKE AWAY ALAWAYS HAS TIPS. Slide 3 of 50 2019 Minitab, Inc.

Summary 1 – Insights about DOE; 5 2 – What is in for me; 8 4 – How we learn and improve; 9 5 – Practicing observation; 10 6 – Methods of experimentation; 16 7 – DOE Case Study 18 8- DOE Case Study Conclusions; 44 Presentation mode, Click on the circle to go to the specific topic. This is the summary for our DOE Case Study. The next slides will cover each topic not deep, but in a superficial level. Slide 4 of 50 2019 Minitab, Inc.

Some insights about DOE. It is important to learn benefits of DOE methodology To know differences between DOE and trial and error or one factor at a time approaches to experimentation Learn basic DOE terminology Distinguish between the concepts of full and fractional factorial designs Use Minitab to run and analyze a DOE This Case Study is the basic of DOE with MINITAB, this is enough to better understand how practical is DOE by MINITAB. Slide 5 of 50 2019 Minitab, Inc.

History of DOE. Sir Ronald Fisher was the innovator in the use of statistical methods in experimental design. The first to develop and use the analysis of variance technique for analyzing data. Many of the early applications of experimental design techniques were in the area of agriculture and the biological sciences. Modern experimental design methods are now widely used in industrial, commercial, transaction and financial applications. 1920’s - introduction of statistical methods in agriculture by Fisher 1950’s - introduction in chemical engineering (Box, .) 1980’s - introduction in Western industry of Japanese approach 1990’s - combinatorial chemistry, high throughput processing We Can say that DOE is nearly 100 years old. It has been used and improved since then. MINITAB was a huge improvement on use of DOE. Slide 6 of 50 2019 Minitab, Inc.

History of DOE. Duncan Hines used designed experiments in the 40 s to 50’s on their cake mixes. Their goal was a robust design for the most consistent product. Duncan Hines (1880 –1959) was an American pioneer of restaurant for travelers. He is known today for the brand of food products that bears his name. Slide 7 of 50 2019 Minitab, Inc.

After this case study. What is in for me? Learn how we use experimentation to improve our processes; Learn how we can experiment more effectively to improve our processes with fewer resources in a shorter time period; Learn how to test if some of the inputs are working together to influence the response; Learn the basics of DOE in Minitab; DOE gives you a framework for how to set up the experiment and how to analyze the data when you’re done on data collection. Slide 8 of 50 2019 Minitab, Inc.

How do we learn and improve? Products and processes are continually providing data that could lead to their improvement. So what has been missing? There are several possibilities: We are not collecting and analyzing the data provided; We are not proactive in data collection; We are unable to translate the data into information; A significant event has not occurred; 1. Significant Event. 2. Somebody Sees It. Observer 3. Research. Study How we can learn more efficiently “In order to learn, two things must occur simultaneously: something must happen (significant event) and someone must see it happen (perceptive observer).” George Box Slide 9 of 50 2019 Minitab, Inc.

Let s practice how to learn by observing. Practicing During World War II, the US Commander of American Air Bases, wanted to finding a way to increase aircraft survivability of American heavy bombers flying in the campaign over Europe, without compromising its flight range or maneuverability, idea was to cover the aircraft in heavy armor plating in some parts. The US Commander came up with the idea of contact America’s Statistical Research Group (SRG). Abraham Wald was the PIC for this task and quickly he decided it is a simple problem to solve, Wald and his team just started by gathering data. They looked at returning B-29 bombers and made note of where they had taken fire. They observed and noted as bombers continued to limp back to air bases, fuselages riddled with holes, with some even described as “Swiss cheese.” Wald found himself tasked with a simple problem: If you can only apply armor to certain parts of the aircraft, where do you apply it? Let s see the real data collected by Wald and his team during World War II on American heavy bombers. Let s see next page. Slide 10 of 50 2019 Minitab, Inc.

Let s practice how to learn by observing After data collection, it is easy to answer: Where do you reinforce the airplanes? What are your recommendations? Real data plotted on airplanes draws about fire holes when returning from Europe Campaign on World War II. Slide 11 of 50 2019 Minitab, Inc.

Mr. Wald Conclusion. Missing Holes. Let s see now: What were Abraham Wald's recommendations? The planes that received the most shots in the highlighted areas were able to fly back. But those planes hit in unmarked areas have not returned. No one was analyzing the bullet marks on the non-returning planes. This case illustrates the Survival Bias, which is quite common when analyzing data to test a hypothesis: If we use the only available information as sufficient, we will largely ignore the causes of these problems. Since the amount of missing information is always infinitely larger than the available information, you have to ask the right questions. The planes that not returned were shot exactly on the areas that returned planes didn t get any fire hole. Let s see more about “SURVIVAL BIAS”. Slide 12 of 50 2019 Minitab, Inc.

Explanation about Missing Holes Theory Mr. Wald explained to the Defense Department that the proper thing to do was to apply the additional armor to the places bullet holes weren’t, rather than the places a high number of bullet holes were recorded. Why? Because the data had been compiled by survivor aircraft, and therefore represented the opposite end of the data pool, in terms of the problem at hand. ATTENTION! The Survivor Bias can distort the Reality or what you see. Slide 13 of 50 2019 Minitab, Inc.

How do we improve? By creating significant events and observing them, we can obtain knowledge faster . That is basically what occurs in a designed experiment. Let’s look at an example of these two things occurring (significant event and perceptive observer) simultaneously. DOE can help us to understand better significant events by observing them, collecting data and understanding the results. Slide 14 of 50 2019 Minitab, Inc.

What is Process? Controllable Factors PROCESS. Inputs A controlled blending of inputs which generates corresponding measurable outputs. Outputs From Understanding Industrial Designed Experiments, Schmidt & Launsby Uncontrollable Factors Process Inputs interacting with Controllable and Uncontrollable Factors and producing Outputs (Responses). DOE works with Controllable Factors. Slide 15 of 50 2019 Minitab, Inc.

Methods of experimentation. Experimentation has been used for a long time. Some experiments have been good, some not so good Our early experiments can be grouped into the following general categories: 1.Trial and Error 2.One-Factor-at-a-Time (OFAT) 3.Full Factorial 4.Fractional Factorial 5.Others Those are common methods of experimentation. Slide 16 of 50 2019 Minitab, Inc.

Why did we use DOE? The structured methodology provides a directed approach. Avoid time wasted with “hunt and peck.” Don’t need 30 years of experience to design the tests. The designed experiment gives a mathematical model relating the variables and responses. No more experiments where you can’t draw conclusions. The model is easily optimized, so you know when you’re done. The statistical significance of the results is known, so there is much greater confidence in the results. Can determine how multiple input variables interact to affect results. A DOE analysis requires less manual input from engineers. Setting up a DOE analysis with MINITAB is time consuming. Often we used a trial and error approach to testing or just changed one variable at a time. Now we know why a statistically designed experiment is better. Slide 17 of 50 2019 Minitab, Inc.

DOE Case study. Real Life with MINITAB. From now onwards we are going to see a practical use of DOE. By using MINITAB on how to predict Color on High Quality Plastic Bottles using the CIE color spaces. What the hell is CIE and Color definition on High Quality Plastic Bottles? From now on we will be using real data. Because of some confidentiality restrictions names, locations and positions were omitted from this Case Study. Slide 18 of 50 2019 Minitab, Inc.

A bit explanation about Color. The CIE 1931 RGB and CIE 1931 XYZ color spaces were created by the Commission Internationale de l'Eclairage (CIE) in 1931. They were the first defined quantitative links between distributions of wavelengths in the electromagnetic visible spectrum and physiologically perceived colors in human vision. The mathematical relationships that define these color spaces are essential tools for color management, when dealing with inks, plastic products, illuminated displays and recording devices such as digital cameras. This is how all colors are measured, described and identified by numbers. With no doubt on identification. Slide 19 of 50 2019 Minitab, Inc.

Back to DOE Case study. Real Life with MINITAB. BACKGROUND ABOUT THIS CASE STUDY: Information about Company, specific location and period of time of this Case Study will be hidden due to confidentiality requirements. This Case Study was developed by me, when I was working as Operational Excellence Consultant in China. At that time I was consulting for a company that produces Extra High Quality of Plastic bottles, specially used for Medicine, Wine, Whisky, High Purity water etc. This Case Study is related to a Extra High Quality bottles producer located in China, mainland. Those bottles are very specific for each customer. Slide 20 of 50 2019 Minitab, Inc.

DOE Case study. Real Life with MINITAB. BACKGROUND: A lot of work usually had been done into developing the formulation of the ingredients XP2 %, Blue XP % and Red XX %, those are some of the ingredients used by that company for the Color Bottle setting. We need to better understand the color variation and tolerances when producing color bottles. We need to define a predictive model for each of those three additives in the solution. With this it will be able to predict the X*, c*, d*, RT and the Green Index GI those responses (CIE 1931 XYZ ) specify the color level and each percentage of the three additives to determine the final bottles solution. Examples of preform plastic samples produced to check color The bottles production process is basically the same as all other normal bottles on the world, injection preform and blow mold bottles. Slide 21 of 50 2019 Minitab, Inc.