From Himalayan forests and Tokyo research labs to LiDAR field plots in Georgia and city-scale 3D mapping in New York — the common thread is turning spatial data into measurable, useful representations of the real world.
My current work focuses on scalable 3D geospatial workflows: connecting floorplans, tax-lot geometry, imagery, LiDAR, attributes and validation into operational building representations.
Automated workflows convert georeferenced PDF floorplans and 2D polygon geometry into floor-aware, textured GLB models with unit-level tax and use attributes.
PDF floorplan → georeferencing → polygon mapping → floor/elevation logic → texture generation → GLB assembly → unit attributes → QA/QC → publishing.
Tools for preparing floorplan imagery, generating controlled grayscale texture levels, and producing consistent model-ready assets for 3D building workflows.
Research and prototyping around building-floor estimation, facade interpretation and spatial measurement using imagery, LiDAR and building geometry.
Exploring how city-scale 3D geometry, imagery, spatial visibility, tax data and derived building attributes can support assessment and operational analytics.
My doctoral and postdoctoral research worked across terrestrial, UAV and airborne scales to measure forest structure, understory vegetation and biomass.
Evaluated terrestrial laser scanning as a way to measure structurally complex understory vegetation that is difficult to observe from above-canopy sensors.
Compared statistical and variable-selection approaches for predicting forest attributes from high-dimensional LiDAR metrics.
Linked field observations and terrestrial measurements with UAV and aerial sensing to explore broader-area prediction and transferability.
University of Georgia
Athens, Georgia, USAThe University of Tokyo
Tokyo, JapanTribhuvan University, Institute of Forestry
Pokhara, NepalA selection of peer-reviewed work spanning forest structure, LiDAR modeling, long-term forest inventory and disaster resilience.
Used terrestrial laser scanning to quantify difficult-to-observe understory vegetation in managed plantation forests and evaluate TLS-derived structural measures against field observations.
Read paper ↗Compared modeling approaches for predicting forest attributes from high-dimensional LiDAR metrics, with attention to variable selection, nonlinear relationships, and model performance.
Read paper ↗Long-term forest inventory work examining how natural forests in Nepal’s Middle Hills changed over two decades.
Interdisciplinary disaster-resilience research examining post-earthquake household conditions and the role of remittances.
Read paper ↗