Wayne State Big Data Analytics Key to Midwest Big Data Effort

Ratna Babu Chinnam

Wayne State University's College of Engineering will lead a major component of a national Big Data project announced in November by the National Science Foundation (NSF). The NSF's goal is to augment ongoing activities and ignite new Big Data public-private partnerships across the nation. The NSF has initially budgeted $5 million nationally to spur creation through four regional data hubs, each of which involves a consortium of members from academia, industry and government.

For the Midwest hub, the WSU portion of the project focuses on aiding computing for business analytics through the use of extensive data sets that allow corporations to optimize performance in areas ranging from customer service to manufacturing and supply chain management to dealership and store franchise management.

The Big Data initiative is mapped as a wheel, with the initial consortium forming the hub. The NSF is in the process of funding additional groups to form the spokes with special focus areas. Wayne State's special expertise is analytics. Its strength is the applied and practical use of Big Data tools and results to further the goals of affiliated businesses rather than generate mainly theoretical research results. Thirty companies and government entities are collaborators in the Big Data Midwest Hub, with at least half of them coming through Wayne State's programs. They include the U.S. Department of Veterans Affairs, Ford Motor Company, Domino's Pizza and Detroit-based automotive consulting company Urban Science.

"Not only will we do applied research through the hub but we also are working with the private sector," says Ratna Babu Chinnam, professor of industrial and systems engineering at Wayne State and director of the university's Center for Supply Chain Management. " We want to focus on helping organizations better manage their product development offerings and service organizations." The College of Engineering takes the lead on Big Data because of the discipline's demand for high-powered computing platforms and distributed computing methods, as well as statistical and machine-learning methods.

Big Data is popular shorthand for efforts that use extensive and high-powered computing means to analyze large sets of information for trends, patterns and logic that might defy ordinary analysis or gut feelings. For example, hospitals might claim that patients come to their emergency room because of service excellence, while Big Data might reveal that a combination of location, public transportation routes and many independent patient factors make the real difference. Chinnam points out that most entities can deal with analysis of a modest data set that might include up to five-dozen variables or factors that can be controlled or changed as part of an experiment. Big Data, however, handles situations in which thousands of variables might be applied in analysis of a particular set of information.

Such problems come up when studying networks that may include 3,000 car dealerships, 6,000 pizza franchisees or 15,000 emergency clinics, where hundreds of individual, time-stamped pieces of information are collected with every customer interaction. How do you schedule automotive service or optimize pizza delivery for a metropolitan area? What underlying trends emerge in establishing national emergency medical service? Big Data tools allow such sets of information to be analyzed and manipulated in ways that generate business intelligence rather than sometimes faulty guesswork.
"Many manufacturing and product development companies are collecting a lot of data, Chinnam explains. "There are all kinds of spurious patterns that show up just by chance. How are you able to connect the dots? You will see all kinds of correlations, but just because you have a pattern doesn't mean anything. We want to generate information that is actionable."

While the National Science Foundation's initial grant does not give any particular participant much funding, Chinnam says it is the start of a larger program that will eventually bring significant grants to successful programs. In 2016, the NSF intends to invest $9 million more to fund research in the strongest spokes of each research hub. Wayne State's business analytics spoke is a strong contender for additional funding because of the strong business interest behind it.

Meanwhile, the Big Data project at Wayne State is already achieving educational results. Cross-disciplinary programs are being created that involve tracks in analytics, business and engineering. WSU also is home to an annual Big Data symposium that is expected to draw more than 80 participating companies to campus on March 24, 2016, at the newly renovated Student Center Building.

"We have a holistic approach to business analytics and are currently rolling out a new master's program in collaboration with the Mike Ilitch School of Business," says Chinnam. "We are also launching a brand new program to basically address the creation of talent in Southeast Michigan."
At its heart, Chinnam explains, Big Data is a collaborative venture. "For the past six months, we had to make a very strong case to the hub leadership saying business analytics is a critical component. We argued for it, got it established, and we will be driving that spoke in collaboration with leading universities in the Midwest," he says.

The Midwest Data Hub is hosted by the University of Illinois-Champagne. In addition to Wayne State University, co-leads for the hub's business analytics spoke include the University of Michigan, Michigan State University, the University of Cincinnati, the University of Minnesota and Northwestern University.
##

For information on the 2016 Wayne State University Big Data Symposium, visit Bigdata.wayne.edu