The Data and Predictive Analytics Center in RIT’s College of Science is helping industry partners harness information from their smart, interconnected devices and streamline manufacturing processes.
Director Mihail Barbosu is positioning the center as a resource for companies investing in the Industrial Internet of Things, in which connected digital networks combine big data and machine learning for increased operational efficiency.
Insights collected through large-scale data analytics can optimize manufacturing decisions, reduce down time, monitor the health of equipment, and predict timely maintenance and repairs, Barbosu said.
The interdisciplinary Data and Predictive Analytics Center, formed in September, is an affiliation of faculty and students who manipulate large-scale data, drawing upon a mix of mathematics and statistics, computer science, engineering, and other areas. The center will involve undergraduate researchers and support graduate students from the new mathematical modeling Ph.D. program, starting this fall, specializing in data analytics and simulation of complex systems.
An early partnership with ITT Gould Pumps has led to ongoing internships through the center. Undergraduate students already have gained hands-on experience developing monitoring systems that measure vibration and temperature and detect equipment failure.
“ITT wanted to know when some devices would stop working and what kind of parts and services they would need to plan for,” Barbosu said. “The larger context is the same: You get sensors, collect information, analyze data, and identify where the problem might occur and under what circumstances.”
Barbosu anticipates a growing interest in this area at RIT, especially with an international conference related to the Internet of Things planned for the fall. Nathan Cahill, associate dean for industrial partnerships and associate professor in the School of Mathematical Sciences, is organizing the event aimed at academics and industry members.
“RIT could be a leader in the Industrial Internet of Things,” Barbosu said. “We have everything here that we need: data science, mechanical engineering, computer science—and students.”
College of Science Dean Sophia Maggelakis pointed to potential collaboration with RIT’s Golisano Institute for Sustainability in its Manufacturing USA federal initiative. Earlier this year, the U.S. Department of Energy selected the institute to lead its new Reducing Embodied-energy and Decreasing Emissions (REMADE) Institute—a national coalition of universities and companies that will explore clean energy initiatives that keep U.S. manufacturing competitive.
“The ability to mine large amounts of data to discover patterns and make actionable predictions is revolutionizing manufacturing, among many other concerns,” Maggelakis said. “RIT can play a significant role in advancing how our industry partners use data analytics to inform their business decisions.”