ODU Research Team Wins $4.3 Million Grant Merging Engineering with Philosophy

Engineers are not likely to search for professional guidance in the writings of Plato, Kant or Tolstoy, but a trio of systems engineering researchers at Old Dominion University has done just that in pioneering a novel way to address complex problems. The research has piqued the interest of the U.S. Department of Homeland Security (DHS), which announced late last year that it was investing $4.3 million in the team's work.

Andres Sousa-Poza, the young associate professor of engineering management who leads the project, likes to refer to himself as a revolutionary and he admits that his more senior colleagues counsel him about patience. His Complex Adaptive Situation Methodology (CASM), which he pointedly calls "chasm," is geared to bring about a paradigm change in the way engineers deal with hydra-headed problems and he says he is eager to get started.

Other researchers on the project team are Charles Keating, a professor of engineering management in the Frank Batten College of Engineering and Technology and the director of the university's National Centers for Systems of Systems Engineering (NCSOSE), and Samuel Kovacic, a retired Air Force officer who is a research scientist at NCSOSE and Ph.D. student in engineering management.

The chasm that Sousa-Poza refers to is the practicality gap between today's system of systems engineering (SoSE) and a new SoSE that he sees as more useful. He believes the field, which is only about 15 years old, must adapt if it is to offer insightful guidance for border security, port security, counter-terrorism and many other problems that are formally described by engineers as "wicked" because of their complexity.

Sousa-Poza's approach requires that SoS engineers get comfortable with another word to describe these problems-"intractable." He also sprinkles his discourse with "multidiscipline," as in the multidiscipline that SoSE must become in order to adequately assess and respond to intractable problems. Engineering, he contends, offers no "golden bullet" for complex problems. "We must recognize that we can't just push a button and get a solution. We must accept intractability."

Computer experts know intractable problems as ones that can be solved, but are so messy they cannot be dealt with in a timely or efficient manner. In other words, the solution is not really a solution after all because it does nobody much good. To accept intractability, therefore, seems to be a lot like raising a white flag, admitting that our problem-solving processes are not up to the task of solving today's complex problems. This is not the sort of pronouncement you expect to hear from ODU's seven-year-old NCSOSE (pronounced "nex-sus") or from a team of researchers who have pledged to come up with $4.3 million worth of answers for homeland security strategists.

But the project team is well versed in paradoxes, as well as in epistemology (knowledge theory), dualism, empiricism, positivism and plenty other themes in Western thought going back to ancient Greece. What can we know? How do we know it? Do context and perspective matter? Does the observer affect the observed? Can math and science explain everything?

"We did a survey of the literature going back 3,000 years and found the same debate continuously," Sousa-Poza explains. The operative question, he adds, comes down to, "How do you feel about uncertainty?"

Keating laughs when he is asked if Sousa-Posa and Kovacic might be called philosophers as well as engineers. "They're far out there," he answers. "More than once, I've had to translate for them."

SoSE came into being because engineers recognized that problems in modern society cannot be reduced to system problems and addressed system by system. Take, for example, port security in Hampton Roads. When the NCSOSE was commissioned a few years back to address this issue, it identified a long list of stakeholders, each with its own culture, needs and mission. These include the Navy and other branches of the military, governments at all levels, shippers, fishermen, shipbuilders, importers, exporters, police and even waterfront communities. The ideal security system for each stakeholder could not be expected to add up to the ideal system for the port as a whole. A discrete fix might create a big problem elsewhere. Security challenges were not static-they were a moving target-and these challenges sometimes seemed to evolve and develop immunity to solutions.

NCSOSE did what it was expected to do and addressed the port security issue as a complex problem that needed a system of systems solution. The ODU engineers fashioned the typical SoSE response, which involved taking existing engineering tools off the rack, so to speak. Most of the tools are technology oriented such as computer hardware and software, and they crunch "hard" data in search of the sort of objective results that remove uncertainty. There are also existing, complementary tools that are identified as "soft" and often focus on human/social and contextual factors, as well as interpretative analysis.

The ODU researchers label this typical SoSE response as "bifurcated" and admit that they were not happy with what this two-pronged, traditional approach produced in the NCSOSE study of Hampton Roads port security. Off-the-rack tools, none of which is custom-made for use on very complex problems, are not the answer, they decided. Wicked problems with technical, social, organizational, managerial and political dimensions (think of the Katrina disaster) tend to unfold in ways that cannot be predicted by algorithms and concomitant interpretative analysis.

Explained Kovacic, "We engaged the port security research assuming that we had advanced techniques and knowledge that could be directly applied to the problem domain. We quickly realized that our most advanced methods were failing to deal with the wicked problem of port security."

"The reality quickly set in. We were inadequate at the most fundamental level," Sousa-Poza added. "We painfully learned. Our path was clear. We first had to develop the underlying theory before we could achieve the breakthroughs we were seeking. The tools and methods would follow the development of a new paradigm for SoS problems."

Keating said the port security research experience was painful. "But this initial humbling of NCSOSE researchers was a pivotal event. Any doubts that the state of existing knowledge was insufficient to deal with SoS problems was removed." Bold action followed, he added. "NCSOSE researchers resisted the temptation to follow the growing groundswell of academic, government and research entities that were simply extending failing approaches wed to paradigms of the past. This created tensions within initial industrial relationships and proved to be a first test of the NCSOSE researchers' commitment to their stance."

SoSE had gained momentum because it promised quick solutions, but Keating said NCSOSE researchers became convinced that "the direction the field was moving was simply wrong. The resulting SoS forays were sorely inadequate. Instead, NCSOSE focused on knowledge first, application second."

Perhaps it was destiny, Keating said, that about four years ago the NCSOSE mavericks crossed paths with like-minded executives of DHS. A relationship was formed, with the DHS supporting preliminary new-paradigm investigations by the NCSOSE team. And now that relationship has matured to the point that the DHS Science and Technology Directorate, Borders and Maritime Division, has awarded the $4.3 million that will provide five years of support for Sousa-Poza, Keating and Kovacic as they finish the design of CASM and test its effectiveness.

Oktay Baysal, dean of the engineering college, noted that the college established a homeland security cluster, which helped to bring about the launch of NCSOSE. "Clearly it's in all of our interest for homeland security to be a national priority and a success," he said. "We are very proud of our engineering faculty, in particular Drs. Charles Keating and Andres Sousa-Posa, who led NCSOSE to grow as an effective partner to the Department of Homeland Security's Science and Technology office."

CASM, as noted earlier, is about acknowledging and managing uncertainty. There have been eras over the centuries when great thinkers have become enamored of "eureka" tools-in the case of Newton's calculus and physics, for example-and declared that humans will be able to use these discoveries to obtain objective answers about most everything. Sousa-Poza points out that since the Enlightenment, which started late in the 17th century, the world has not been particularly comfortable with uncertainty. The prevailing opinion has been that problems are solvable if we define them, gain enough knowledge regarding them and subject this knowledge to objective analysis.

But NCSOSE researchers found that they could not engineer uncertainty out of a complex problem. The mere effort to establish the boundaries of a problem can require generalizations, and although this can lead to a quick solution, the researchers see such solutions as half-baked and not practical. Also, they say complex problems cannot be understood in a traditional sense. Together with what can be learned about a messy situation, there will always be the unknown, or non-knowledge. (Metaphysical ideas of Plato and Kant are germane here.) Besides, they add, sometimes it does not necessarily follow that the more we know about a complex problem, the better the solution will be. Finally, they note that human individuals have vastly different abilities and world views and it can be risky to predict the behavior of a group of people as a congregated "human factor."

The researchers, therefore, say that any solution for a problem such as port security that comes mostly from objective-or positivist-methods will be static, very difficult to implement and, quite possibly, counterproductive.

This is where the researchers introduce Tolstoy's War and Peace. "Of the 1,100 pages, half are about strategy and decision-making," Sousa-Poza said. "Remember Gen. Kutuzov discussing the complete absurdity of the emperor making tactical decisions while he is sitting back in St. Petersburg?" Readers of the book might also remember that Napoleon believed that he could control the course of a battle by dispatching couriers, while Kutuzov sought only to do some initial planning and then let subordinates direct the battlefield action. Typically, Napoleon would be frantically sending out orders throughout the course of a battle-orders that were often misinterpreted or made irrelevant by changing conditions-while Kutuzov would rest in his tent during the battle.

On the fourth floor of Innovation Research Park @ ODU Building One on the east side of campus, a large room well appointed for conferences and teleconferencing anchors the headquarters of NCSOSE. A roundtable section of the room, something like a command center, allows for virtual environment experiences and formal brainstorming by as many as two dozen people.

In this room, the project team proposes to put CASM into action, and, in the process, develop a multidiscipline system of systems that formalizes a way to respond to intractable problems. CASM seeks to avoid encapsulating problems within unrealistic boundaries. It refuses to ignore uncertainties, choosing literally to add "white space" to its formulations and provide guidance about dealing with the unknown. CASM will employ virtual environments, and help decision-makers who are participating in the research to get their minds around the "gestalt"-or overall form-of a complex problem. Eventually, it will put real people in the midst of real problems and test decision-making in a live environment.

There are wrong ways to engineer intractable problems, Sousa-Poza said, just as there are wrong ways laid out by Tolstoy to fight a war and wrong ways prescribed by educators to teach schoolchildren. "We need to teach people to challenge boundaries, to teach them how to think," he explained. "Faith in rigid plans that lack pragmatic robustness, such as in standards of learning (SOLs), evidence-based learning and rote memory, is misplaced."