A two-stage probabilistic framework that reconstructs physically meaningful variables missing or uncertain during monitoring and propagates their uncertainty to the final prediction, producing reliable and well calibrated predictive intervals.
A forecasting framework that combines weather models with data driven and physics based approaches to predict wind power generation for grid operators across North America, from the regional level down to individual wind farms.
Study of wake interactions in offshore wind farms and their impact on energy loss, focusing on turbulence, yaw misalignment, and environmental factors that shape turbine performance and farm wide efficiency.
A computer vision based system for detecting and tracking passenger origin destination pairs in resource limited transit systems, combining onboard video analytics with GPS data for accurate and scalable mobility sensing.
Application of multi camera photogrammetry for accurate 3D structural displacement measurement during dynamic testing, using synchronized multi vision triangulation and advanced image based tracking.
Use of UAV-based lidar to detect water bodies hidden beneath dense forest canopies, improving watershed delineation and hydrologic analysis in complex environments.
Study of the seismic and thermal performance of integral, semi-integral, and conventional bridges, highlighting how soil-structure interaction and connection details influence resilience.
