Design of Experiments for Aerospace Products, Processes and Services 2009-01-3138
If Design of Experiments (DOE) is so good why isn't it used more? Despite its power DOE has some demanding conditions. You don't have to be a statistician to conduct a successful DOE, however.
In Aerospace we are very often faced with the design, improvement or correction of a product, process or service. The most serious task is determination of the most relevant variables or factors that affect the final result or performance of the process or product in question. Once these factors have been established, we then need to determine an optimal combination of settings or levels for each factor. There is also the complicating possibility of interaction among the factors also. Design of Experiments (DOE) is an established statistical technique used on its own or in conjunction with a Six Sigma project to determine the strongest individual factors and if there is interaction among the factors. DOE is a powerful and pragmatic tool that can efficiently and effectively help us design a piece-part or whole system as well as manufacturing or transactional process or even a service offered as a product.