By Kathryn ONeill, Managing Editor, SDM Pulse
As President Obama’s choice to become a member of the US Nuclear Regulatory Commission (NRC), MIT Professor George Apostolakis is no stranger to systems challenges.
At the NRC, “90 percent of problems are systems problems,” Apostolakis said. The work is multidisciplinary, involving not only “hard” disciplines such as nuclear reactor physics and materials science, but also the inevitable challenges of managing people and preventing human error.
Nevertheless, Apostolakis said he’s found traditional engineers find it very difficult to accept that human issues are so important. “They consider it ‘soft,’” he said.
MIT’s System Design and Management Program (SDM) is valuable precisely because it zeroes in on these kinds of problems, said Apostolakis, who has taught SDM students since the program’s inception. “You can’t be the leader of a company and be an equation kind of guy. You need a broader view,” he said.
“Not to diminish detailed engineering work, but this by itself is not enough,” he said. “A lot of difficult and interesting problems that have a societal element to them are really systems problems.”
The KEPCO professor of nuclear science and engineering as well as a professor of engineering systems, Apostolakis teaches a course in engineering risk benefit analysis (sometimes referred to as ERBA). The class introduces SDM students to risk assessment, decision analysis, and cost-benefit analysis—three major approaches to solving problems through the use of probability theory. The goal is to expose students to reallife problems that are systems-based and to give them tools with which to address them.
In every module, Apostolakis invites guest speakers to provide students with examples from their experience. He’s had representatives from NASA explain how they assess risk before a space launch, for example, and how they conduct tradeoff studies to explore the value of alternative designs.
Members of MIT’s Department of Facilities have also come to class to explain how they prioritize the infrastructure renewal process on campus by applying decision analysis, a methodology that helps managers to analyze and compare choices.
Facilities might consider whether delaying a project will pose any hazard to humans, how much it will cost, how much disruption the work with cause, and how long it will take. Once various factors have been weighed and the methodology applied, the result is a clear and logical ranking of options, Apostolakis said.
“It’s a very practical application,” said Apostolakis, noting that the decision analysis module is universally popular with students.
In addition to teaching ERBA, Apostolakis has also served as thesis advisor to a number of SDM students over the years. In particular, he remembers one who evaluated the water supply of a city to determine its vulnerability to terrorist attack. The city found the material useful in its decision-making, he said, which is not always the case with academic work.
Another SDM student considered what NASA could learn from its experience with past technologies to estimate the reliability of new systems (the so-called “heritage problem”). “I was very pleased” with that project, Apostolakis said.
“SDM students are usually more mature (than other ERBA students) and more willing to deal with systems issues,” Apostolakis said. “A program like SDM—it’s a great thing.”