the Connector - Munther Dahleh

September 4, 2014

Munther Dahleh, the Connector at MIT

“If you want to understand bank runs in financial systems, or congestion in the air transportation system, you have to understand the interaction between social and institutional behavior, and physical and engineered systems,” Professor Munther Dahleh says. “This means integrating education and research across campus, including the business school, humanities and social sciences, sciences, and engineering.” Read about Dahleh's many ways of connecting —to postiviely impact people at MIT and worldwide.

See the Sept. 4, 2014 MIT News Office article by School of Engineering writer Leda Zimmerman titled "The connector - Munther Dahleh’s expertise in mathematical modeling gives him profound insight into people, and MIT," also posted below.

There are very few things Munther Dahleh hasn’t done at MIT. With an international reputation in electrical engineering, a research emphasis on challenging multidisciplinary problems, a host of academic leadership positions, and a deep engagement with students in and out of the classroom, he's wholly, and cheerfully, committed to living and working at MIT. “I truly enjoy immersing myself in the institution and everything that comes with it,” says Dahleh, the newly appointed William A. Coolidge Professor of Electrical Engineering and Computer Science. “Academics or student life — I find it all fascinating.”

Whatever the assignment, as a researcher, faculty resident of a dormitory, lab director, or umpire in tricky game of inside baseball, Dahleh thrives on “interactions with peers and with students.”

Dahleh’s passion for connection proved essential in a significant venture he embarked on in December 2013. He was asked to lead what he calls “a reinvention” of the Engineering Systems Division (ESD), where he currently serves as acting director. Founded in 1998, and headquartered in the School of Engineering, ESD was envisioned as a home for tackling complex and large-scale engineering and socio-technical problems. “If you want to understand bank runs in financial systems, or congestion in the air transportation system (for example, delays all around the country because Atlanta had one inch of snow), you have to understand the interaction between social and institutional behavior, and physical and engineered systems,” Dahleh says. “This means integrating education and research across campus, including the business school, humanities and social sciences, sciences, and engineering.”

But in the 15 years following ESD’s inception, Dahleh says, a new world of data collection and high-speed computation has emerged, demanding a paradigm shift in MIT’s existing academic and research structures. For Dahleh, it became clear that this presented an opportunity to create “a new overarching entity that belongs to all of MIT,” he says, addressing complex 21st-century societal challenges and drawing on firepower from all of MIT’s schools.

In a frenzied five months, Dahleh captained the effort to shape this new entity, which involved shepherding 39 faculty members on four different committees toward a common conclusion. “Our faculty members are smart, and they push for their ideas. You need to argue and arrive at some sort of meeting of the minds,” he says. “It was a hard job.”

“Munzer is one of the very few people who can find an exciting and compelling vision that brings together intellectually diverse groups,” says committee member Karen Willcox, a professor of aeronautics and astronautics. “He is somebody who has the respect of different communities and intellectual cultures.” Dahleh had a head start reaching out to colleagues outside his home in the School of Engineering; he has collaborated with faculty in all five schools. Throughout this potentially volatile enterprise, he knew how to connect with these colleagues. “I was most impressed by how much time he spent between meetings engaging people one on one,” Willcox says. “Throughout the process, he made such an effort to understand everybody’s perspective. I couldn’t believe how much patience he had. I would have exploded!”

The outcome of this effort, a 75-page report, makes the case for creating a new organization at MIT devoted to understanding and managing complex societal and technical systems, and to new programs in 21st-century statistics. The core ambition of this new entity is to address critical challenges in energy, transportation, healthcare, financial, and social systems through the multidisciplinary study of human and institutional behavior, statistics, and information and decision sciences. The director-designate of this new organization: Munther Dahleh.

This appointment trades not only on Dahleh’s collaborative instincts but the scope and scale of his own research interests. For his PhD in electrical engineering at Rice University, Dahleh was motivated by the potential for new approaches to control systems for such complex engineering feats as the Boeing 777 and the International Space Station. He created a computational framework for designing feedback control systems in the presence of uncertainty, a mathematical approach that allowed for a much simpler and systematic way to address complex performance requirements without sacrificing robustness or increasing risk. Because his computational formulas streamlined design processes, it became integral in many areas of manufacturing.

His approach has led to a range of research partnerships: He has worked with NASA on attitude control of the space station; with Brigham and Women’s Hospital on prostate cancer treatment; with Ford on ways of using GPS data to improve the efficiency of hybrid car batteries; and with neuroscientists from MIT on the brain and human motor control. “I am driven by real-world problems,” Dahleh says. “But my contribution is providing a mathematical foundation for thinking about them.”

Dahleh also studies the areas of systemic risk and optimization. With colleagues from MIT’s Laboratory for Information and Decisions Systems (LIDS), where he was also recently named interim director, he has been thinking about the stability of the power grid, especially as it begins to incorporate renewable energy sources. “My work is to understand electricity supply and demand in real time, and to understand exactly the mechanism of failure — how outages occur — so we can monitor them and prevent them from occurring.” In collaborations with MIT economists and others, he also works on problems of traffic congestion and resilience in transportation networks and financial systems.

More recently he has turned to analyzing social networks. He worries that “we don’t yet have the paradigms to think carefully about questions of social behavior,” even as reams of data come pouring in from such sources as Google. He would like to model the way information propagates in systems, and learn whether networked communities embrace good ideas, or arrive at a consensus that does not make sense. “As we interact on Facebook, and have all sorts of noisy interactions through multiple pathways, do we learn the truth from each other?” Dahleh asks.

Here Dahleh spots another opportunity to bring people together. “Social scientists have spent years thinking about questions like these. But they are not used to this gigantic amount of data and its connectivity,” Dahleh says. “That’s what engineers and computer scientists are very good at. So here is a merger of fields I’d like to define and understand.” The new MIT entity might be just the place to spark such a fruitful merger.

While coaxing this entity into existence, running ESD and LIDS, and maintaining his research commitments, Dahleh will juggle a few other balls, as well. As the housemaster in MacGregor, he tends to 350 undergraduates. (“I don’t describe it as a job. It’s a way of life. Different people belong to different communities. We belong to MacGregor,” he says.) And he is also chair of MIT’s Committee on Discipline, which he describes as a way of helping “great kids understand their mistakes.”

The unifying, foundational theme for Dahleh remains what it always has been: teaching and mentoring students. “What I most enjoy is seeing the transformation from the day they walk in to the day they leave. If I extract that talent and see them successful out there, this is a great accomplishment for me.” One advisee, Spyros Zoumpoulis PhD ’14, attests to the success of Dahleh’s methods. “Munzer showed me how to set up a research problem and present the research results as a relevant and rigorous story with a concrete message. I learned from him what research should aspire to do," Zoumpoulis says.

Dahleh offers freedom to explore, but points the way, Zoumpoulis adds: “He was as much a friend as collaborator.”

For Dahleh, personal and professional satisfaction comes from “having a positive impact on the lives of as many people as possible.” He even views his research as “the food I use to interact with other people.” Any legacy he might leave at MIT, he says, “is not about a result, or anything I do that makes it big in the world. I don’t dream about that. It’s more about the people.”