Journal of Mechanical Design

companion website

GUEST EDITORIAL: DESIGN OF COMPLEX ENGINEERED SYSTEMS

12/1/2010 Author(s): Bloebaum, C.L, McGowan, A.M.R., 
Volume: 132(12) – December, 2010

Many past editorials have discussed definitions of design, pointing out the wide variation across individuals and industries. We adhere to a definition that recognizes design to be a matter of making rational decisions regarding available alternatives in order to achieve one’s stated preference. In an upcoming special issue of this journal, research articles on the topic of “designing complex engineered systems” are being invited. This begs the question: What is a complex engineered system? Further, what are the unique design challenges of such systems?

We define complex systems to be those for which tightly coupled interacting phenomena yield a collective behavior that cannot be derived by the simple summation of the behavior of the parts. In essence, these are highly interdisciplinary systems in which the existence of inherent couplings potentially leads to irrational results. Complex systems may be biological (human body), natural (rain forests), or engineered (aerospace, naval architectures, drilling platforms, and medical devices).

Growing human needs in national defense, environmental sustainability, medical advancements, and human prosperity, in general, have been drivers in the increased complexity seen in engineered systems. However, whether one is designing a new material for a particular behavior across scales or a large-scale transportation system, involving numerous interacting disciplines, the inherent couplings wreak havoc with the practice of imposing a traditional structured hierarchical design process. Further, there is little room today for an antiquated organizational approach rooted in single discipline superiority.

In general, present efforts to address complexity in engineering design tend toward managing the complexity through more processes rather than attempting to rigorously understand it through theory or even exploit it to improve system performance. While the reliance on traditional methodologies and processes tied to a particular organizational structure can have detrimental results at the smaller scale, the failures are magnified substantially for large-scale complex products. Today, the most common system engineering approaches involve using a hierarchical decomposition within a requirements-driven framework. Even the most meticulous implementation of this approach when used for large-scale complex systems (such as military aircraft) can still lead to mind-boggling cost overruns in the hundreds of millions of dollars and substantial time delays that set projects back by years (or even lead to project cancellation). We can (and we must) reverse this pattern. The widespread and rapidly growing prevalence of complex engineered systems means that the critical gaps in design theory and methodology have pervasive and damaging impacts from small to large scales.

Now, more than ever, our engineering design community must rise to the challenge of addressing these issues. However, what exactly are the issues? What makes them so difficult? Most importantly, what opportunities are there for new research, given the challenge of the complexities that exist?

Design of complex engineered systems has the incredible challenge of mitigating and minimizing the inevitable irrationality that is the hallmark of any activity involving multiple individuals or disciplines. It is impossible for any person or even group of people to understand and manage all the couplings that exist in such a system. Hence, the ability to make rational decisions in such an environment is seriously compromised. Consider that in just over 100 yrs of air-powered flight, we have gone from a design team of two people to design teams of over 2000, with costs for large systems in the hundreds of millions (if not billions) of dollars. The time for adding ever more elaborate processes, where theoretical foundation is absent, is past. It is now time for a design science firmly rooted in mathematical rigor that encompasses organizational and other complex social realities, particularly in light of the challenges of designing complex engineered systems.

We highlight two (of many) broad areas of research needed in the area of designing complex engineered systems: the need for a rigorous design theory and the often overlooked, but invaluable, input from the social sciences.

To date, the engineering design community has produced critical results toward establishing a design science based on mathematical rigor. These include research in decision-making with uncertainty, multidisciplinary design optimization (MDO) for coupled disciplines, as well as advances in information technologies to support decision-making. The issue of appropriately modeling and reflecting uncertainties across inherently coupled disciplines, scales, and components remains a very challenging problem. Challenges in MDO still exist in appropriately understanding and modeling inherent couplings (interacting phenomena), incorporating uncertainty modeling, achieving large-scale rigorous optimization, and avoiding approaches, in both analysis and optimization, which violate the basic tenets of aggregation. In considering some of the needed design tools, there are significant opportunities for visualization, virtual reality, and data driven approaches that can enable improved decision-making in the complex design environment. Further, a tremendous challenge exists in developing true interoperability for design tools that are used in the design process.

These are only a small number of the research issues that are particularly critical for designing complex engineered products. However, another key area of research pertains to achieving rationality (or minimizing irrationality) in large-scale designs, wherein large numbers of individuals, subsystems, and companies (contractors) are involved. Social scientists from psychology, cognitive science, sociology, economics, and business can yield substantial dividends to engineering design, particularly where complexity is involved. For complex engineered systems, the final design is a function of the organization that created the system. While it may be impossible to achieve rationality, there are unique opportunities for reducing the chaos that presently exists such as those approaches being researched in the emerging field of mechanism design (not to be confused with the design of physical mechanisms).

In the upcoming special issue of the journal on designing complex engineered systems, we hope to see papers that address these and many other engineering, social, political, and economic issues that arise in designing complex engineered systems, with an emphasis on instilling a necessary rigor throughout the design enterprise. We also call attention to the comprehensive presence of design in all aspects of engineering: from cradle to grave. Upon careful examination, we see design permeating the entire engineering spectrum from the earliest conceptualization in basic research and development to the environmental considerations in product disposal. That is, research in the science of design that includes decision-making, teaming, accessing trades, optimization, and human behavior can improve the work of the laboratory researcher, as well as the operational coordinator. Additionally, it is our hope that experts within traditional design disciplines will accept the challenge of partnering with experts in nontraditional disciplines (i.e., the social sciences) to tackle the most difficult of research problems, those residing at the interfaces, which demand a legitimately interdisciplinary approach. It is our feeling that truly paradigm-shifting research can result from such partnerships.

It is an advantageous time to bring to light the design science of complex engineered systems in a special issue.