Making the most of solar power in Masdar City, Abu Dhabi, solving sewage and water issues in Pittsburgh, maximizing traffic flow in Istanbul—these case studies illustrate how dynamic models and hands-on insights can come together in smart and wannabe smart cities, observed IBM’s Jurij Paraszczak in a lecture on March 8 at Isenberg. Paraszczak, a former director of IBM Research Industry Solutions and Smarter Cities Program, was the guest of UMass Amherst’s student chapter of INFORMS (Institute for Operations Research and the Management Sciences), which hosts an ambitious series with speakers from industry and academe.
By definition, “a smart city is optimized around a set of goals,” he said, entailing the systemic coordination between a city’s challenges—energy, transportation, waste disposal, and many other activities. If we don’t account for that interactivity, “we’ve failed,” he insisted.
Data are grist for the analysis of smart cities, observed Paraszczak. Simple models that connect “physics” with everyday activities and systems dynamics models, which allow for a broad view, require less data than statistical models that predict the future from the past. The latter employ machine learning to flesh out patterns.
It is impossible to build sufficient storage to capture all of the data within our grasp, Paraszczak continued. To that end, informed judgment is critical. “Data have to be dynamic, not static,” he emphasized. Another challenge with data gathering, he said, involves privacy. Witness recent high-profile surveillance activities via social media. Paraszczak recalled that when staying at UMass Amherst’s hotel, the parking garage would not allow him to reenter gratis after leaving. “They must have photographed my license plate. You have to be very careful with what the world is doing,” he cautioned.
Wanted: Hands-on experience and expertise
It is also critical, he added, to incorporate insights from hands-on professionals (like department of public works employees) into your models. But many professionals typically work in silos. DPW workers, for instance, have little to do with a city’s tax department. “The methods of talking together do not exist and there is competition among departments for resources. Also, people [i.e., citizens] want to be heard—most cities don’t do that well,” he remarked. The result: A city’s quality of services suffers. In contrast, smart cities deploy informed, data-driven models but also resolve who decides, what resources are deployable, and who will make compromises for those scarce resources.
Paraszczak declared his smartest of smart cities to be Minneapolis, which he characterized as “unique and well-managed.” The city, he said, leverages diverse community perspectives in data-rich, predictive models. “Minneapolis, in fact, devotes 20 percent of its revenues to IT,” he noted.
Data are so numerous and rich that you cannot do without IT experts in the mix, Paraszczak told his largely student audience. “Ask a mathematician and you’ll miss layers of underlying data. The upshot: “As you go out into the world as IT professionals, you’ll find millions of opportunities like these.”