Speaker: Dr. Saiprasad Ravishankar, Ph.D.
Affiliation: University of Michigan
Title: Powering the Future of Imaging and Signal Processing with Data-Driven Systems
Abstract: The next generation imaging systems are expected to be increasingly data-driven. In this talk, I will present my research on efficient, scalable, and effective data-driven models and methodologies for signal processing and imaging. First, I will discuss transform learning, where interesting structures such as union-of-transforms, incoherence, rotation invariance, etc., can be considered. Transform learning-driven approaches achieve high-quality results in applications such as video denoising, and X-ray computed tomography and magnetic resonance image (MRI) reconstruction from limited or corrupted data. The convergence properties of the learning-based algorithms will be discussed. I will also present work on dictionary learning in combination with low-rank models (LASSI) and demonstrate its promise for dynamic MRI. The efficiency and effectiveness of the methods proposed in my research may benefit a variety of additional applications in imaging, computer vision, neuroscience, and other areas requiring data-driven parsimonious models. Finally, I will provide an overview of recent research and future pathways, including on physics-driven deep training of reconstruction algorithms, learning undersampling patterns in compressed sensing-type setups, online adaptive estimation of dynamic data from streaming measurements, etc.
Bio: Dr. Saiprasad Ravishankar received the B.Tech. degree in Electrical Engineering from the Indian Institute of Technology Madras in 2008, and the M.S. and Ph.D. degrees in Electrical and Computer Engineering in 2010 and 2014 respectively, from the University of Illinois at Urbana-Champaign, where he was then an Adjunct Lecturer and a Postdoctoral Research Associate. Since August 2015, he has been a postdoc at the University of Michigan. His interests include signal, image and video processing, imaging, machine learning, inverse problems, compressed sensing, and large-scale data processing and optimization. He has over 1200 Google Scholar citations and has received multiple awards including the Sri Ramasarma V Kolluri Memorial Prize from IIT Madras and the IEEE Signal Processing Society Young Author Best Paper Award for 2016.
The theme for TEDxOhioStateUniversity’s 7th annual event, Hide or Seek, will feature students, faculty, staff and alumni as speakers and performers who will inspire and challenge our concepts of science, technology, history and life.
Register and find more information at tedx.osu.edu. General admission tickets are available until February 11.
Speaker: Stephanie Stockar, Ph.D.
Affiliation: Pennsylvania State University
Title: Optimal Control of Energy Distribution Systems
Host: Professor Sandip Mazumder
Abstract: The recent growth in urban development and in the demand for personal mobility has increased the awareness towards the need for reducing the energy consumption and improving the efficiency of buildings and transportation systems. In this scenario, a renewed interest is emerging from the Industry and government agencies to advance the state of the art in the model-based optimization and control of thermo-fluid systems, which characterize the energy storage and conversion processes in propulsion systems and buildings. The application of dynamic systems and control theory to thermo-fluid systems improves the accuracy and robustness of control system design process, ultimately leading to an overall increased efficiency of the energy conversion process.
This seminar focuses on the optimal control of nonlinear dynamical systems, with application to smart buildings and connected and automated vehicles (CAVs). In particular, the seminar highlights the application of advanced optimization techniques to large-scale physical systems, which requires advancements in the theory of model order reduction for nonlinear partial differential equations as a way to improve the accuracy and physical consistency of control-oriented models. The proposed research approach is illustrated through selected case studies. First, the physics-based modeling and optimization of a smart home is presented. Then, the application of model reduction and optimal control for the optimization of the engine thermal system in a connected environment is illustrated.
Bio: Stephanie Stockar is an assistant professor of mechanical engineering at Penn State University. She conducts research is in the areas of modeling and optimization of nonlinear dynamical systems, with focus on building energy systems and advanced automotive systems. Her research approach hinges upon the multidisciplinary integration of thermo-fluid sciences with dynamic systems, modeling, optimization and control. Stockar’s work has been funded by Ford Motor Company, Fiat Chrysler Automobiles, the National Science Foundation, ARPA-E and Volvo.
Before joining the MNE Department at Penn State University, she was a research associate at The Ohio State University Center for Automotive Research (OSU CAR). She earned her bachelor and master’s in mechanical engineering from the Swiss Federal Institute of Technology (ETH), Zurich in 2007 and 2010, respectively, and obtained her PhD in mechanical engineering from The Ohio State University in 2013.
Speaker: Jeremy Renshaw
Affiliation: Program Manager, Used Fuel and High-Level Waste at the Electric Power Research Institute
Title: Robotically-Deployed Nondestructive Evaluation of Used Nuclear Fuel Storage Canisters
Abstract: As Dry Cask Storage Systems (DCSSs) that store used nuclear fuel age into the period of extended operations, nondestructive evaluation (NDE) inspections will be needed to confirm continued safe operation of these systems. NDE inspections are the cornerstone of aging management to identify and assess potential degradation of the DCSS to maintain their safety functions during the period of extended operations.
DCSSs present a challenge for inspection and delivery systems due to the high temperature, high radiation, confined entry and small annulus space. Some DCSS designs may have multiple 90° bends or other internal obstacles that must be negotiated to perform inspections. Various DCSS designs have been deployed to the field with different geometries and other substantial challenges for inspection. Despite these challenges, significant progress has been achieved across the industry in developing inspection and delivery systems for DCSS inspection.
A number of NDE techniques have been identified and either modified or refined for improved compatibility to DCSS inspections that include capabilities for high-temperature, high-radiation, and confined or occluded area inspections. While significant progress has been made, additional modifications and refinements are ongoing. Field trials have played an important role in the technology transfer process to facilitate the use of these systems by industry. To date, five field trials have been conducted using multiple NDE techniques deployed on robotic delivery systems. Future field trials are planned to further refine and develop the NDE and delivery systems.
Bio: Dr. Jeremy Renshaw is the Program Manager for the Used Fuel and High-Level Waste Group at the Electric Power Research Institute (EPRI).
He manages R&D efforts focused on all aspects of the back end of the fuel cycle including used fuel, wet and dry storage, transportation, and eventual dispositioning. These efforts include aging management and inspection of dry cask storage systems, maintaining criticality margins during wet and dry storage, understanding high burnup cladding performance, and activities related to interim and final storage options.
Prior to joining the Used Fuel and High-Level Waste group, Dr. Renshaw focused on nondestructive evaluation (NDE) of nuclear fuel assemblies, dry storage canisters, concrete structures, one-time inspections, and advanced NDE methods. He has experience with ultrasound, eddy current, thermography, terahertz, microwave/mm-wave, and acoustic emission testing, and robotic deployment methods.
Prior to joining EPRI, Dr. Renshaw worked at AREVA, Inc. as an R&D Project Manager/Senior Engineer on reactor internals inspection and other R&D efforts related to advanced NDE inspections. Previously, Dr. Renshaw worked at Sauer-Danfoss, Inc. as an engineer on the new product development team and in the materials lab.
Dr. Renshaw serves as a reviewer for multiple peer-reviewed journals, has published several peer-reviewed journal articles and conference proceedings papers, and holds one U.S. Patent.
Dr. Renshaw holds a Bachelor of Science in Mechanical Engineering, a Master’s degree in Systems Engineering, and a PhD in Materials Science and Engineering, all from Iowa State University where his research focused on thermal NDE methods.
Speaker: Justin Erwin
Advisor: Dr. Coe
Speaker: Saswata Dasgupta
Advisor: Dr. Herbert
Speaker: Tracy Preston, Communication and Marketing Manager, Corporate Engagement Office; Technology Commercialization Office
Affiliation: The Ohio State University
Speaker: Dr. David Beebe
Title: TBA (Area: microtechnology, medicine, and biology)
Affiliation: University of Wisconsin-Madison
Host: Professor Shaurya Prakash
More details will be coming soon.
Speaker: Hongcai Zhou
Affiliation: Texas A&M University
Host: Dr. Wade
Speaker: Anderson Janotti
Affiliation: University of Delaware
Title: Vacancies, Small Polarons, and Two-dimensional Electron Gases in Complex Oxides
Abstract: Progress in epitaxial growth of complex oxides have led to heterostructures with a unique set of physical phenomena, such as the formation of a high-density two-dimensional electron gas (2DEG) at the interface between two normally insulating materials—e.g. SrTiO3/LaAlO3 and SrTiO3/GdTiO3. Superconductivity and magnetic ordering have been demonstrated in these systems, sparking the interest in novel device applications. The formation of a 2DEG at the interface between SrTiO3 and Mott insulators, such as GdTiO3, have also been demonstrated, with electron densities that are over an order of magnitude higher than those realized with conventional semiconductors. Charge transport in these systems exhibit exquisite behavior, varying drastically from metal to insulator depending on the thickness of the building-block layers. Despite the intense research efforts in the last decade, the origin of the excess charge and the mechanisms that determine the density of the 2DEG are still under debate, and fundamental properties of the 2DEG are not fully understood. In this presentation, we will discuss how computer simulations can provide insights into the origin and nature of the 2DEG. Based on results of first-principles calculations we will discuss electron correlation effects and how the electronic structure of these heterostructures can be drastically altered, turning from metallic into insulating, through charge localization in ultrathin layers. Charge localization will also be addressed in the context of small polaron formation, and these concepts will be used to explain transport and optical phenomena observed in the bulk materials.