Decision Making in Engineering Design
“Engineers do design” — a factual statement made by many both inside and outside of the engineering community. The statement has been the basis of speeches by William Wulf, President of the National Academy of Engineering; by John Brighton, Assistant Director for Engineering at the National Science Foundation (NSF); and by industry leaders commenting on the current outsourcing of manufacturing and, now increasingly, some engineering design jobs overseas. Design permeates the activities of all engineering disciplines: Civil Engineers working on large-scale infrastructure systems in transportation; bioengineers creating new sensors for human health monitoring; mechanical engineers developing new alternative energy sources and power trains for the hydrogen economy; and electrical engineers linking information and communications networks through new advances in photonics. So if all engineers are already doing design, why do we need a program that supports design theory research? Given that engineering design crosses all the disciplinary domains in engineering, our challenge is to focus on creating the new knowledge, advancing the support tools, and building the necessary principles and foundations into a domain-neutral framework that enables engineers to meet the future needs of society. As a research community, a design research program is needed to continue our work to establish the set of principles that underlie all design, such as:
Design requires a clearly stated objective function.
Design must address the uncertainties within all aspects of the system to better inform the decision-making.
Over the past three decades, design theory research has taken several twists and turns, as computational tools became the standard for how engineers of all disciplines “did design.” In an early NSF Workshop report, Design Theory '88 , research was categorized into topical areas focused on the design process that included the computational modeling; the cognitive and social aspects; the representations and environments; the analysis tools including optimization and the design “for” such as “for manufacturing.” At that time, the NSF program was called Design Theory and Methodology and consisted of three components that essentially captured these five topical areas: The first, Scientifically Sound Theories of Design, established a home for proposals that were directed at creating the scientific basis for the design process. The second, Foundations for Design Environments, was aimed at advancing the understanding of fundamental generic principles that could be used and understood across engineering domains. The third, Design Processes, was focused on the how and why of the design process, including early work on life-cycle concepts and concurrent design.
At this point, you may ask, “So what is new?” The tools certainly have advanced over the years, from early computer-aided design (CAD) through solid modeling capability. The introduction of virtual reality, computer integration engineering, and collaborative and distributed design processes created demands upon the community to focus on how decisions were made, under what conditions and to what purpose. Decision-based design became a major thrust for the research community, with the issues of uncertainty and predictive modeling capability becoming the foci. As with any science, the theories must be put forward, tested for consistency and completeness, and then incorporated (or not) into the framework of the science. This is true, too, for engineering design, if it is to become more than just an ad hoc, intuitive process that is domain-specific. In response, the Open Workshops on Decision-Based Design , a series of face-to-face and website workshops, addressed the spectrum of issues that were raised.