The design tends to focus on two phases:
*Modeling: Mathematically model the relationship between output and design factors.
*Optimization: Optimization is performed through mathematical programming using a mathematical model.
However, scientific design requires good models, which means that data must be collected through experiments. And because the quality of an experiment depends on the level of experimental design, the next step is to plan it.
*Planning: Plan experiments that ensure orthogonality so that a model containing the necessary terms can be obtained with a small number of experiments.
In addition, in optimization, if a formulation is based on multiple ideas, then as many solutions as there are will be obtained, and ultimately it is necessary to select the solution to be adopted from among these. At that time, candidate solutions are comprehensively examined and evaluated, and it is not uncommon to make fine adjustments to the candidate solutions based on the actual implementation.
* Identification: Identify the solution to be adopted from the candidate solutions and make fine adjustments if necessary.
Optimization is mathematics, but specification is management based on mathematics.
The above four phases are carried out in the following order:
Planning → Modeling → Optimization → Specification Although design in the narrow sense often refers only to optimization, in this research the combination of all four phases is treated as comprehensive design. The above discussion of design will be developed within the framework of a new design methodology called hyper design.
