Amount Awarded: $25,000
Balancing the social, economic, and environmental dimensions of sustainable development remains a persistent challenge in urban planning. While infrastructure solutions often dominate the conversation, one overlooked yet critical factor is access to sky view. Sky openness—quantified by the Sky View Factor (SVF)—influences a wide range of urban processes, from mental and physical health to solar energy efficiency, microclimate regulation, and air quality. Despite its importance, current methods for estimating SVF are either data-intensive or limited in spatial coverage, hindering their integration into city-scale planning.
This project proposes a novel, cost-effective framework for high-resolution SVF estimation by integrating LiDAR-based 3D urban modeling with deep learning-based sky segmentation from street-level imagery. Focusing on two climatically and morphologically distinct cities—San Francisco and Trondheim—we aim to (1) develop an end-to-end pipeline that combines multimodal data for SVF estimation, (2) benchmark and validate results using LiDAR models, deep learning predictions, and traditional fisheye photography, and (3) analyze spatial variability in SVF to assess its socio-economic and environmental impacts across different urban forms.
By advancing both methodology and application, this research contributes a scalable tool for SVF mapping and provides new insights into how urban form mediates sustainability outcomes. Our approach supports planners and policymakers in making informed decisions that reconcile competing priorities in the built environment.