Research for Data Sciences and Engineering Thrust

(Mohamed Abdel-Mottaleb, Xiaodong Cai, Mei-Ling Shyu, Kamal Premaratne, Manohar Murthi, Miroslav Kubat, Jie Xu, Ramin Moghaddass, Cheng-Bang Chen)

Data analytics and artificial intelligence (AI) drive creative solutions for many societal issues in areas such as biomedicine, health care delivery, security, cybersecurity, transportation, and smart grids. In order to understand and extract meaning from diverse real-world data, and to engineer new systems based on the insights from such data, there is a need for research on fundamental algorithms, and systems for handling and harnessing big data. ECE is currently focusing on the algorithmic aspects of big data and informatics, namely the design of AI, machine learning (ML), statistical signal processing, data mining, data fusion, image processing, game theory, and computer vision methods for processing various types of signals and data with different levels of quality and volumes. ISE is also developing advanced analytical and data mining models that can be used to transform large-scale datasets into actionable insights and decision-making intelligence in real time. Another active area of research at ISE is to develop a new generation of analytical models that can employ heterogeneous data collected from large-scale network/graph structures to monitor and control such systems in an optimal manner. To enable the widespread adoption of these algorithms by human specialists without AI expertise and in real-world deployment scenarios, new hires are needed in areas including automated ML, interpretable ML, trustworthy and robust ML, ML acceleration, and ML privacy and fairness.

(Jie Xu, Hammam Alsafrjalani)

Given the trend to connect devices and systems of all manners and scales, such as medical devices, manufacturing plants, or sensors in smart buildings to the Internet or other networks, there is a need for fundamental inquiry into the design of hardware, software, and algorithms in such cyber-physical systems. In particular, cyber-physical systems are generating an unprecedented amount of data and execute many mission-critical tasks essential to contemporary society, and so the performance issues of security, self-healing and adaptation, and robustness are of paramount importance. The ECE department has research projects in security at various level of system design to address these challenges. These projects include developing models and algorithms to optimize and secure computing and cyberphysical systems including cloud/edge computing, mobile networks and Internet of Things; and AI technologies to make sense of their operational data and secure the cyber space. As cyberphysical systems generate, transport and store a tremendous amount of data, supporting and enabling subsequent artificial intelligence derivation, many new hires are needed in various topics of cyberphysical systems, including cloud and AI security, distributed and edge AI systems, future networks, and cyberphysical hardware.

(Sakhrat Khizroev, Sung Jin Kim, Michael Wang, Onur Tigli, Pratim Biswas)

Fundamental research on devices, circuits, and systems at different scales is needed to engineer solutions to challenges in medicine and health care, energy harvesting and storage, sensing the environment, and in other domains. In particular, all the emerging computing paradigms, e.g., quantum computing, neuromorphic computing, and in-memory computing, are inherently based on nanotechnology, and nanotechnology is becoming that indispensable tool required to bridge multiple disciplines and thus has a significant impact on other vital areas including AI and big data, cybersecurity, and power electronics. ECE faculty are active in application areas of nanotechnology such as biomedical applications, energy harvesting, information processing and storage, and have strengths in four main themes of nanotechnology including nanophotonics, microelectromechanical systems (MEMS), nanoelectronics and nanomagnetics/spintronics.

View Article

(Xiaodong Cai, Mansur Kabuka)

With the advances in technologies such as next generation sequencing, enormous amount of genomic, transcriptomic, and other omcis data have been generated from various tissues of diferent organisms under various conditions. More recently, omics data have been collected at single-cell level. The availability of massive data in bioinformatics provides unprecedented opportunities for knowledge discovey, development of new diagnosis tools and therapeutic methods for various diseases, but also raises new challenges for data mining and analysis. ECE faculty is active on the development of Innovative mahine learning algorithms and AI methods for analyzing heterogeneous, multimodal, and high-dimensional data in bioinformatics including DNA seuqnces, gene expression, DNA methylation and other clinical data, in collaboration with faculty members in the medical school.

(Suhrud Rajguru, Abhishek Prasad, Andrew Dykstra, Ozcan Ozdamar)

Neural engineering applies engineering methods, including data analytics and artificial intelligence, to study neural physiology and to assess, modulate and restore neural function with the goal of understanding and treating neural diseases. BME is focusing on the assessment of vision, hearing and balance; monitoring neural function during surgery; and the development of neural interfaces, neuro-prosthetic devices and rehabilitative approaches to study and restore neural function.

(Mohamed Abdel-Mottaleb, Weizhao Zhao, Fabrice Manns, Noel Ziebarth, Mansur Kabuka)

Medical imaging develops and applies imaging technology and image processing to study biological and physiological phenomena on all scales, from the molecular to the organ level, to detect disease, to plan and monitor treatment, and to guide surgery. BME is focusing on the development of technologies and approaches for detection and treatment of cancer and for image-guided surgery, including Magnetic Resonance Imaging (MRI), nuclear imaging, X-ray CT and ultrasound; the development of optical coherence tomography, atomic force microscopy and optical technologies for high resolution imaging of tissues and organs; and medical applications of lasers and other forms of energy for treatment, diagnosis, and stimulation; the characterization of physiological function in the ear, eye, brain and other organs.

(Nurcin Celik, Seok Gi Lee, Murat Erkoc, Ramin Moghaddass)

The field of energy and environment research spans several domains including grid design and optimization; natural hazard mitigation; carbon mitigation; climate dynamics; energy efficiency; energy storage; environmental chemistry; combustion; fuels; nuclear fusion; ocean waves and ocean- 5 atmosphere dynamics; public policy; solar energy; sustainable and resilient cities; water resources and hydrology; and wind energy. The work combines mathematical and computational modeling, experimental work, and extensive engagement with industry. ISE faculty are working on data-driven modeling and decision making for large-scale complex and dynamic systems such as distributed electricity grids, energy-aware manufacturing scheduling, electrical material handling vehicle planning, transportation planning, energy demand management, diffusion of solar energy adoption, and data analytics foundations for systems operating under uncertainty and dynamic environments.

(Gang Wang, Esber Andiroglu, Matthew Trussoni, Pratim Biswas)

Effectively addressing the U.S. and world’s energy needs in the near and long-term has important design implications for the building sector. CAE faculty maintains research activity in areas of building envelope optimization in context of energy consumption, environmental comfort and water/wastewater system management. Current research activities include development and accuracy improvements of virtual fan air flow and virtual pump water flow meters using the fan/pump motor system characteristics.

In addition, the development of a variety of sensors allows for low cost methodologies for distributed monitoring. COE faculty are deeply engaged in development of PM sensors (through work in the Aerosol Science and Technology Initiative). In addition, collaborative work with researchers who rely on satellite remote sensing, allows for an integrated system to be developed and deployed. Due to the high volumes of data generated, concepts of data science and engineering are being utilized to enhance environmental informatics. Machine learning and artificial intelligence modalities are being developed and deployed.