By Eric Smalley
November 21, 2012
If Professor Richard de Neufville were to have a motto, it might well be "Expect the unexpected." The Professor of Engineering Systems and Civil and Environmental Engineering has been working to get designers of complex systems to embrace flexibility. Technological systems will be more effective in the long run if the people who build them give up trying to divine the future and instead imagine –– and prepare for –– many possibilities, even those that are unexpected.
"Right from the start you ought to recognize that things might be different than anticipated and build in the capability to react," said de Neufville, who holds a dual appointment within the MIT Department of Civil and Environmental Engineering and the MIT Engineering Systems Division. De Neufville is also the founder of ESD’s Technology and Policy Program.
Traditional systems design involves creating specifications that guide how the system is built. The specifications are derived from a set of requirements that outline what the system is supposed to do. The problem is that the longer the system’s expected lifetime, the further system designers have to peer into the future to identify those needs.
"The problem with that is the forecast of what is needed inevitably is wrong — not because people are doing a bad job, but simply because things happen," said de Neufville. "You get an optimum solution for a clientele, market or situation that turns out not to exist."
Inevitably new technologies arise, new competitors appear on the scene, or the client’s needs change. And it’s generally difficult to reconfigure a design that was optimized for a particular set of requirements rather than made adaptable, said de Neufville. Enabling a system’s managers to react to unanticipated events can significantly increases the system’s performance, he said.
The Blue Cross Blue Shield Tower in Chicago is a good example of designing flexibility into a system, said de Neufville. The tower, which serves as headquarters for Health Care Service Corporation, was built 33 stories tall in 1997. The building’s designers engineered the tower so that additional stories could be added later, and the building was extended 24 stories from 2007 to 2010. The health insurance company was able to wait until the need for more space was definitive rather than building a larger building initially and having unused extra capacity, he said.
Designing for flexibility is an evolution of the engineering process. The initial approach was optimization –– how to make things better, said de Neufville. The next approach focused on decision analysis, which involves making the best choices given an uncertain environment, he said. Designing for flexibility is about "dealing the cards to yourself," he said. De Neufville co-authored a book on the subject, titled "Flexibility in Engineering Design", that was published by the MIT Press in 2011.
De Neufville’s MIT career has focused on system analysis and system design, he said. "It’s essential to take a larger encompassing view, otherwise you can optimize a piece of a system but that piece might not fit with the rest of the system."
De Neufville is a member of the MIT faculty team that is developing the new Singapore University of Technology and Design. That work has made him think about what’s special about MIT. His conclusion: it’s a bottoms up organization and it’s self-organizing. "People see a job to be done and they figure out how they’re going to do it," he said.
Richard de Neufville