ABSTRACT
Designers and decision makers in general, seek to maximize their expected utility from a decision as we have discussed already in this book. A utility function maps the attribute level to an abstract notion of satisfaction, thereby facilitating decision making under uncertainty. So far, we have considered only one underlying attribute, which can be a significant simplification. In most decision problems, more than one attribute is involved. As an example, let us consider the design of a car seat. In addition to the cost of the seat, the designer might want to consider the weight, material and physical dimensions of the seat. Howdoes the designermeasure the overall worth of the seat when it is clearly measured in terms of multiple attributes and involves tradeoffs? More succinctly how can decision analysis helpmake decisions under uncertainty when multiple attributes are involved? Table 8.1 shows examples of decision problems that involve multiple attributes. Thurston (1991) emphasized that evaluation of designs using multiattribute utility
theory is essential to selecting the best design that satisfies all the objectives. Over the years, decision-based-design as a field has seen a considerable development, Lewis et al. (2006). Proper modeling of a decision maker’s preferences over multiple attributes is complementary to design optimization, i.e. only when we have a mathematical formulation that models tradeoffs between attributes can we maximize a design’s overall worth. In this chapter, we present methodologies and formulations that address this important issue.
