TGCC Chapter Program July 2020
Tires: a Test that Answered the Question Efficiently
Ken Johnson, Statistical Engineer, NASA Engineering and Safety Center
Design of experiments (DOE) is a set of tools used by many firms and organizations to help answer engineering, scientific, marketing, even political questions through physical or computer experiments and tests. The tool set is aimed at efficiency. The method starts with rigorous framing of the questions to be answered. It helps guide choice of input parameters and output responses and considers constraints such as cost, schedule and impractical combinations of inputs. The focus on planning means the number of test runs necessary to achieve test goals can be calculated in advance and that the resulting test data can be analyzable using standard, engineer-friendly software.
The speaker will use a real example to show how a seemingly intractable problem can be succinctly framed, how engineering knowledge can be used in estimates of sample size and how efficient design of experiments tools help shape a test matrix. Graphical and mathematical representations of the model will show how the engineering question was solved and firmly grounded compared to results from a common testing approach. And the speaker will discuss how you can start using the design of experiments DOE tool kit yourself.
The talk will be geared toward anyone who has interest in making their process less variable, their design more robust, or their rigs more reliable. No statistical knowledge or math facility is assumed. You will leave with a basic understanding of what makes DOE tools valuable, roughly how they work, and how you might learn more.