Chapter 2
Decisions in Engineering Design
Abstract
Engineering design is the process by which engineers’ intellect, creativity, and knowledge are translated into useful engineering products that satisfy particular functional requirements and meet engineering specifications while complying with all constraints. The traditional design approach has been one of deterministic problem-solving, typically involving efforts to meet functional requirements subject to various technical specifications and economic constraints, among others. In general, engineering design is a loosely structured, open-ended activity that includes problem definition, representation, performance evaluation, and decision making.
Keywords
Attitude toward risk; Decision theory; Decision tree; Design decision making; Game theory; Utility theoryEngineering design is the process by which engineers' intellect, creativity, and knowledge are translated into useful engineering products that satisfy particular functional requirements and meet engineering specifications while complying with all constraints. The traditional design approach has been one of deterministic problem-solving, typically involving efforts to meet functional requirements subject to various technical specifications and economic constraints, among others. In general, engineering design is a loosely structured, open-ended activity that includes problem definition, representation, performance evaluation, and decision making.
A number of approaches have been proposed to organize, guide, and facilitate the design process. The main objective is seeking a logical and rigorous means to aid in developing a satisfactory design, or one that is acceptable to the customer or user of the product. All approaches heavily involve decision making, which is integral to the engineering design process and is an important element in nearly all phases of design. In fact, it is fair to state that the center of all approaches is decision making.
In this chapter, we define decision making as the process of identifying and choosing alternatives from the set of possible alternatives. Alternatives are developed based on certain criteria and requirements. In the meantime, the preferences of the decision maker are incorporated to sufficiently reduce uncertainty about alternatives, which helps to achieve the desired goals and ensure a high-quality decision.
Various methods are commonly used to aid designers in decision making, such as a decision matrix (Voland 2004), a decision tree (Eatas and Jones 1996), quality function deployment (Akao 2004), and so forth. These methods are generally ad hoc and incorporate relatively high levels of subjective judgment. An additional set of methods address variability, quality, and uncertainty in the design process, such as the Taguchi method (Lochner and Matar 1990), Six Sigma (Park and Antony 2008), Design for X (Huang 1996), and so on. These tools are more analytical and are typically coupled to the processes used to produce products. Design theories also exist, such as Suh's axiomatic design (Suh 1990), which are less widely used but offer more rigorous analytical bases. Finally, certain other methods are used primarily in the fields of management science and economics, such as utility and game theory, which are being explored in the current research for feasibility and applicability to support decision making in engineering design.
In the late 1990s, the National Science Foundation (NSF) initiated a series of studies to determine research priorities in engineering design by examining industry and education needs and to formulate recommendations for the NSF's Engineering Design Program (NSF 1996). The NSF funds an online decision-based design open workshop to engage design theory researchers in a dialogue to establish a common foundation for research and educational endeavors. The NSF also sponsored Gordon Research Conferences in 1998 and 2000 on theoretical foundations to examine theories and techniques for decision making under conditions of risk, uncertainty, and conflicting human values. Such studies promoted research in design theory and led to the development of decision theories and decision-based design, including the application of utility theory and game theory to support design decision making. For those who are interested in pursuing research topics in design theory or decision-based design, please refer to Lewis et al. (2006) for excellent discussions.
This chapter is essentially a prelude to the broad subject of design theory and methods, in which we view engineering design as a decision-making process and recognize the substantial role that decision theory can play in design. Therefore, we start by discussing basic decision methods and theory in this chapter to provide readers with an overview and broad understanding of design decision making. With such an understanding, we extend the discussion in later chapters to more practical methods and tools, such as design optimization, that are widely employed by engineers for the support of design decision making. More specifically, this chapter aims to (1) introduce basic decision theory and methods that aid readers in applying them for general decision making, (2) provide basic concepts of utility and game theories that were explored in recently to aid engineering designs, and (3) use simple design examples to illustrate the concepts and methods for applying the theories to support practical engineering design problems.
2.1. Introduction
Design is a process involving constant decision making. In an engineering design context, the role of decision making can be defined in several ways. The decision process is influenced by sets of conditions or contexts; some are controllable, such as business context, and some are uncontrollable, such as market and economy conditions. The business context represents the long-term view of the company and is in general largely in the control of the company. Decisions such as capital investments, product lines and upgrades, and product marketing strategy are determined by the company. However, some aspects of business contexts, such as market share (which is influenced by competing products), are somewhat uncontrollable. Also, the state of the economy and market demands are not controlled by the company. Correctly assessing the context for making a decision is important because it dictates the level of effort and long-term impact. Decisions with long-term impacts often are irreversible after implementation; therefore, the decision maker must seriously analyze the context and impact of alternatives before arriving at a decision. A large number of short-term incremental decisions can, however, be made relatively risk free in general.
Whether the conditions are controllable or not, there is always uncertainty involved in decision making. Narrowing the focus to product design, the uncertainty largely comes from the inputs, such as the completeness of and variation in product requirements and constraints established by the customers. Closing with the customer is an iterative process, in which reconciling the customer's needs with the developer's design capabilities require collaboration and experience with the product. Decisions made in earlier stages must be re-evaluated from time to time and adjustments need to be made in response to factors or events of high uncertainty that were not predictable or controllable throughout the design process.
In addition to the uncertainty involved in decision making, the designer's preference in choosing one alternative over another plays an important role in design, especially in dealing with multiobjective design problems, in which a designer is juggling competing objectives. In some cases, design decisions are made by design groups of a product development team, in which decisions are made to maximize their respective objectives that could be mutually competing or even conflicting.
In general in product development, decisions are made at different levels under different kinds of scenarios. At a high level, decisions are made for scenarios such as team organization, product cost, work breakdown, and suppliers. At mid-level, a decision involves issues such as design requirements, material selection, subsystems and components, and the manufacturing process. At a low level, a designer determines design objectives, geometric shape and dimensions of the individual components, and so forth. There are many methods that aid in decision making. Some of these methods developed decades ago are more ad hoc and incorporated relatively high levels of subjective judgment, such as decision matrices, in which weighting factors that significantly impact the decision are assigned by the designer. When these methods are used, they are generally applied to support more significant project decision making at a higher level. Methods developed more recently involve rigorous theory and mathematical frameworks in decision making, such as using utility theory.
In this chapter, we intend to provide a basic introduction on both conventional methods and rigorous decision theory. We start by introducing conventional methods that are commonly employed and are easily found in design textbooks, including the decision matrix and decision tree, in Section 2.2. These methods are deterministic and rely on subjective judgment. Although we start with conventional methods, one of our focuses in this chapter is to discuss rigorous decision theory. In Section 2.3, we introduce decision theory, in which decision making is formulated and solved mathematically. In Sections 2.4 and 2.5, we discuss utility theory and game theory, respectively, which have been explored and adapted to support engineering design recently. In Section 2.6, we use simple design examples to illustrate the concept and steps in applying utility and game theories to solve simple design examples. These examples demonstrate the application of modern decision theories to support engineering design.
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.