You are here

Big data for good

article originally appeared in The Ohio State Alumni Magazine

Global problems to solve? Data scientists say ‘bring it’

When you’re faced with a hurricane, Steven Quiring can predict if your lights will go out. 

While meteorologists tell us what the weather will be, this Ohio State climatologist can tell you whether it will knock out your electricity. More importantly, he can tell your utility company so it can make preparations ahead of time to restore it. 

How does he do this? With data — lots and lots of data.

It’s fortunate, then, that he works at Ohio State, one of only a handful of universities worldwide that are pioneering ways to use big data for the betterment of humankind.

Quiring and several collaborators create their forecasts using information from NASA’s Soil Moisture Active Passive satellite, which measures moisture in the Earth’s surface once every two to three days. You may wonder what this has to do with your loss of power in a storm. Among many things scientists have found in a decade of work is that soil moisture plays a big role in whether things like trees and telephone poles stay upright in windy conditions.

They cross-reference the particulars from NASA with data about land use, population density, average wind speeds, and duration and intensity of storms. Then, they generate a map that shows anticipated outages by region. In the case of Hurricane Matthew last fall, Quiring and his colleagues predicted five days out that 4.5 million people would lose power in Georgia, North Carolina, South Carolina and Virginia. The actual number affected: about 4.1 million in those very same states.

A few blocks away from Quiring’s lab on campus, Assistant Professor of Geospatial Engineering Rongjun Qin has developed a use for satellite data after the occurrence of natural disasters. 

Rongjun Qin Every day, satellites orbiting Earth collect images that must be measured in terabytes, or 1,000 times a gigabyte. Qin’s lab has developed algorithms to compare images taken before an earthquake to ones taken afterward in order to help first responders determine where help is needed most. By identifying structures leaning at a particular angle, for example, the algorithms can help pinpoint those in danger of collapsing or spot where impassable roads have cut people off from help. And they are able do this within days rather than the weeks it can take people on the ground to assess damage. The time saved can mean lives saved.

The power of big data

Until recent years, scientists such as Quiring and Qin had no way to deal with all the data available. Ohio State’s new Translational Data Analytics Institute is changing that.

TDAI is a $125 million effort to direct the power of data analytics expertise at the university toward a multitude of societal questions and challenges. It is a foundational asset that faculty and student researchers throughout the university are using to fuel discovery in and across fields from medicine to social work and biology to business.

The use of data to inform decisions has existed for millennia, of course. Tools dating back to 22,000 B.C., for recording information and making calculations, have been unearthed in Africa. Analytics even was the subject of a major motion picture, “Moneyball,” about the Oakland Athletics’ statistical analysis of baseball player data to manage a tiny payroll. 

What has changed is our ability to analyze multiple layers of massive amounts of data and draw understanding from immensely complex matrices of information. And there is no shortage of data to tap. In 2013, IBM estimated that 90 percent of all data in history was created in the preceding two years.

A hub for analytics

TDAI, the College of Arts and Sciences, and the College of Engineering recruited Quiring and Qin, respectively. For these and dozens of other new faculty, the university’s vigorous data analytics program is a huge draw. 

“I was very excited to discover the effort TDAI is putting into cross-disciplinary research,” Qin says. “TDAI is not only a group of faculty coming from different fields, but also a collaborative community of scientists that is excited about using translational data analytics techniques to solve problems and benefit people of this planet.”

“Translational” is the operative word, says TDAI interim faculty director Raghu Machiraju, PhD ’96, a professor of computer science and engineering in the College of Engineering and of biomedical informatics in the College of Medicine.

“It’s about making research useful for the human condition,” Machiraju says. “In the more foundational disciplines like computer science, we often invent methodologies and devices, and then applications for them are sought retroactively. In the translational context, the needs drive the creation of new methodologies.” 

Creating solutions that leverage big data requires a massive amount of coordination. While terms in this field may conjure mental images of monolithic file servers, what’s really needed is more like an ecosystem. This kind of environment requires hardware for collecting, storing and processing information; software and methodologies for mining and analyzing it; data visualization techniques to communicate meaning; and endless applications — from examining cancer patients’ DNA to improving manufacturing to, say, predicting weather-related power outages. 

And the most important element of this entire ecosystem is, of course, the people.

Gathering the experts 

Throughout higher education and science in general, it has become increasingly important to bring diverse people and perspectives together to tackle multiple dimensions of complex issues. TDAI was created to intentionally cut across disciplines and leverage the full potential of the university’s research, learning and outreach. In short, Ohio State’s breadth of expertise allows it to go beyond simply studying issues to solving them. Doing this requires collaboration, and collaboration takes work. 

Ohio State’s Translational Data Analytics Institute has 104 affiliated faculty members who represent 46 disciplines and every aspect of the ecosystem’s life cycle. An important part of TDAI’s work is bringing together those all-important components — faculty members whose research benefits from big data — and creating environments and intersections that can lead to game-changing solutions. 

In the spirit of positive collisions, TDAI hosts opportunities for researchers to discuss projects, which helps bring campus and community together. It offers seed grants for projects involving multiple colleges as well as a faculty directory that is searchable by interest and discipline. And it is creating both physical and digital data analytics hubs with shareable resources.

Into this mix, TDAI Managing Director David Mongeau is adding another key ingredient: partners from industry and the community. With a background in business and engineering, he speaks the language of both. 

“TDAI fills a need at the intersection of academia and industry,” says Mongeau, a former vice president and project lead with Battelle and Bell Labs. “We connect industry and community partners along with Ohio State experts and national funding opportunities to deliver solutions.” 

A recent example is a proposal to compete for more than $5 million to develop data analytics solutions that drive societal changes such as reducing infant mortality and opiate addiction. TDAI spearheaded the response, which cited expertise, methods and tools from more than 60 faculty in 11 colleges and units; two nonprofits organizations; and four corporate partners, including Hewlett Packard Enterprise and Teradata, an analytics company based in Ohio. 

Links with other universities are vital, too. In March, Ohio State formalized a faculty exchange program between TDAI and the Graduate School of Information Science at Nagoya University in Japan that enables data scientists from the two institutions to share knowledge and ideas.

Shuzaburo Takeda ’69 PhD, who proposed the partnership, is senior advisor to Japan’s Ministry of Education, Culture, Sports, Science and Technology. He envisioned a U.S.-Japan consortium that accelerates both countries’ expertise in data science, which he considers the most important area of innovation. 

“We are entering a new age in which data science is essential, and there are so many indicators that universities are key in advancing the field,” Takeda says. “In Ohio, there are a lot of leading industries, such as automotive and IT, and statewide, Ohio is committed to translational data analytics. Strong leadership by Ohio State and real demands coming from such leading industries can make the Translational Data Analytics Institute top-level in the world.”

Training talent

TDAI’s Mongeau recognizes that research innovations are just part of what the world needs when it comes to data science. Over and over, he encounters companies keenly interested in workforce development. In every sector, they need new talent that knows how to harness the power of big data.

One of those companies is Saama Technologies, an international firm headquartered in Silicon Valley.

“The workforce demand is — and is projected to continue — significantly outpacing supply, as often happens in a new emerging space,” says Saama board chair Ken Coleman ’65, ’72 MBA.

Now a member of Fisher College of Business Dean’s Advisory Committee, he has served Ohio State in many high-level capacities as an alumnus. Two years ago, he led Saama’s expansion in central Ohio, a move aimed, in part, at improving access to Ohio State’s intellectual capital. 

“We’re in a world where data will become increasingly important,” he says, “and Ohio State is an institution that is committed to developing data technology and research.”

To address the need for graduates with big-data know-how, TDAI is creating two new master’s degrees in translational data analytics to go along with the university’s new undergraduate major, which was the first of its kind nationwide. It’s also creating opportunities for students to interact with business people as they craft résumés and developing terminology guidelines to help employers and job applicants identify one another.

These measures and others illustrate a focus that goes beyond the mere practice of data analytics and looks to the future of the field.

“Don’t confuse our translational focus with a transactional enterprise,” Mongeau cautions. “Our aim is very strategic and forward-looking. We’re inviting think tanks, federal agencies and industry to the table with us to consider academic programming needs for data scientists, engineers and analysts. We talk about research roadmaps and long-term aspirations with industry partners, not just the immediate problems we can address with our data analytics talent and technology.”

The future is now

With such innovative research, international collaborations and leading-edge education for students, is it possible Columbus could become the Midwest version of Silicon Valley? Coleman thinks so. 

“Central Ohio is a convenient location to create a knowledge center, and there’s a commitment here to data analytics and its challenges and opportunities,” he says. “I believe the best place to put a knowledge center is around a great university, and Ohio State is one of those.”

Takeda concurs. “The science of the 21st century — digital science — is converting the real world into a virtual world and creating technologies that translate new value created in the virtual world into value in the real world,” he says. “There is a new type of infrastructure for the coming era, and we must prepare for the future. Ohio State is doing that very well.” 

 

Creators of technologies for using big data:

D.K. Panda

Professor and Distinguished Scholar of Computer Science and Engineering 

What I’m into: I design high-performance, scalable (and free for users) software libraries to manage and process tremendous volumes of data and ultra complex computations. My lab’s big data software has been downloaded more than 20,000 times, and it is used by hundreds of companies, including Baidu, IBM, Intel, Huawei and Oracle.

Why I do it: Processing huge amounts of data and making decisions within a short time is increasingly critical for a lot of uses. The novel software stacks designed and developed in my group make this possible for users and organizations working with big data. 

Scott Shearer

Professor and Chair, Department of Food, Agricultural and Biological
Engineering

What I’m into: I create sensors and sensing technologies that allow farmers to assess crop health at the individual plant level and then optimize the application of nutrients and crop protectants.

Why I do it: Connecting the agricultural field to the internet is rapidly changing our ability to feed an expanding world population and to do so in a sustainable and environmentally responsible manner.

Chris Stewart

TDAI Faculty in Residence and Assistant Professor of Computer Science and Engineering

What I’m into: I study the design and implementation of computer systems that transform unpredictable, environmentally harmful and inefficient industries. I also am creating TDAI’s Data Commons, the cyber space where faculty, students and industry partners store and share our most valuable asset: data.

Why I do it: We have seen that data — sober and true — can help us explain problems, find solutions and detail the type of technology needed to implement the solution. At TDAI, I am proud to scale these benefits beyond research silos to the whole university.

by Jenny Grabmeier