Seminar - “Learning Complex Process Systems from Data”, by Aditya Tulsyan, Process Systems Engineering Laboratory, Massachusetts Institute of Technology, Cambridge, MA, USA

Thursday, January 7, 2016 3:30 pm - 3:30 pm EST

ABSTRACT:  With the advent of the Internet of Things (IoT), smart devices and smart manufacturing, the amount of data collection has grown exponentially in a manner analogous to the Moore's law. The explosion in data availability, variety and size have enabled engineering, medicine, science and finance to heavily invest in data-based learning projects. Some of the successful learning projects, include The Human Genome Project, Google Flu Trends, Tesla Self-Driving Car Project and GE Industrial IoT.

While the learning paradigm is relatively new to chemical engineers, the process industries are well-poised to benefit from this data-science revolution. In fact, a recent white paper by the consulting company McKinsey showed that the process industry can create immense value simply by finding new ways of extracting information from the process data. Visiongain, a leading business information portal, predicts that spending by the oil & gas industry on data analytics alone will total $3.51 billion in 2016. In this talk, I discuss about a new learning platform for advanced statistical inferencing in nonlinear stochastic processes under uncertainty. A remarkable feature of the proposed learning paradigm is that it can elegantly deal with low signal-to-noise scenarios, multi-sensor failures, and multi-rate sampling - all in real-time. I also present our work on data-based learning methods to address important systems engineering problems related to soft-sensor development, safety and compliance management, adaptive state estimation, quality-by-design and data-based modeling. The application of these ideas are discussed for a variety of systems, such as digital healthcare, process and manufacturing industries, aerospace systems, pharmaceutical systems and oil & gas industries.

Bio-sketch:   Aditya Tulsyan is currently a Postdoctoral Associate in the Process Systems Engineering Laboratory at the Massachusetts Institute of Technology (USA). He received his Ph.D. in Chemical Engineering with specialization in Computer Process Control from the University of Alberta (Canada) in 2013 and B. Tech in Chemical Engineering from the Vellore Institute of Technology (India) in 2009. He has held visiting positions at the National University of Singapore (Singapore) and the University of British Columbia (Canada). He has authored numerous articles in reputed international journals and conferences. His publications have been recognized with a best paper award and a Keynote presentation. He has won several awards, including the Captain Thomas Farrell Greenhalgh Memorial Award, Stuart Preston Foster Foundation Award and Mary Louise Imrie Graduate Student Award. His research interests are in systems engineering, statistical machine learning, data-science, signal processing and Bayesian inference.