Upcoming Doctoral Dissertations

"Repair and manage peripheral nerve injuries particularly those large gap defects remain as significant clinical challenges, where conventional treatments often fail to achieve full functional recovery. Scaffold-guided regeneration of neural tissues paves a promising avenue to repair such nerve damage. In recognition of the unique physicochemical and hierarchical structure of native microenvironment that nerve cells reside, various nano/microfabrication technologies have been explored to recapitulate such complexities in the scaffold design. While recognizing the importance of structural and topological guidance in directional extension of axons and neurites of neuronal cells, the current progress still lacks the combination of topological guidance with localized stimulation for better regeneration." Read more

"With the increasing penetration of wind energy into modern power systems, conventional synchronous generators are progressively displaced by converter-interfaced generation, resulting in reduced system inertia and weakened frequency response capabilities. As a prominent and rapidly growing source of wind power generation, offshore wind farms (OWFs) connected to onshore power grids through high-voltage direct-current (HVDC) transmission systems have significant potential to provide fast grid frequency support by utilizing their available kinetic energy and power reserves." Read more

"Per- and polyfluoroalkyl substances (PFAS) pose significant challenges to wastewater treatment systems due to their persistence, mobility, and complex transformation pathways. Although wastewater treatment plants (WWTPs) are designed to protect environmental and public health, conventional treatment processes are largely ineffective at removing PFAS, potentially transforming and redistributing these contaminants into receiving waters and biosolids." Read more

"The ubiquitous deployment of Machine Learning (ML) Systems faces a dual challenge of resource efficiency and data privacy, In Federated Learning (FL), communication bottlenecks and privacy leakage from shared model updates remain critical concerns. Similarly, Large Language Models (LLMs) unlock transformative reasoning capabilities at the cost of immense memory and computational overhead, alongside emerging security and privacy risks during inference." Read more