Failure Is an Option

August 8, 2012

Failure is a successful theme for Henry Petroski. As Duke’s resident expert on design and structure, he has spent almost three decades teaching and writing about everything from bridges to toothpicks to pencils to space shuttles—including what underlies design mishaps.

In his new book, To Forgive Design: Understanding Failure, Petroski, Aleksandar S. Vesic Professor of civil engineering and professor of history, surveys some of the most infamous failures of our time, from the 2007 Minneapolis bridge collapse and the toppling of a massive Shanghai apartment building in 2009 to Boston’s prolonged Big Dig and the 2010 Gulf of Mexico oil spill. He argues those failures followed from interconnected forces—technology and culture. “Certainly some failures are attributable to design errors, but they are not the only reason that accidents occur,” he writes. “A design is a manifestation of a technological concept, but a designed thing or system can also be neglected, misused, and mishandled by its owners, managers, operators, and users.”

Design does not occur in a technological or political vacuum, he observes. “Questions relating to cost, risk, and other economic, social, and political considerations can dominate the decision-making process and push to the background technical details on which a project’s ultimate success or failure may truly depend.” 

"Certainly some failures are attributable to design errors, but they are not the only reason that accidents occur."

Although Hurricane Katrina was clearly a natural occurrence, he writes, “it was the human design and maintenance of the levees and other storm-protection systems around New Orleans that were found wanting.” In revisiting the Challenger disaster, he quotes the finding of the accident investigation board that “the NASA organizational culture had as much to do with this accident” as the insulating foam that broke away from the external fuel tank and struck the leading edge of the shuttle’s left wing. 

Petroski, in his conclusion, notes that specific failures occur “because of the coincidence of any number of factors that happen to converge at a particular place and a particular time,” much as the Titanic had its chance iceberg encounter. “But it is the nature of the human and technological condition that until incontrovertible failures do occur there is the tendency—even among designers, who should know better—to think that the technology has been mastered.”