PhD Student (Year 1)
EIES-Lab · The Hong Kong Polytechnic University
Kechuan Dong joined EIES-Lab as a PhD student in September 2026. His research sits at the intersection of artificial intelligence, geospatial analysis and urban energy systems, with a focus on spatiotemporal deep learning for large-scale rooftop and facade solar potential estimation. He has also worked on 3D human pose estimation for prosthesis users through multi-source data fusion, and is interested in interpretable, physics-informed AI for real-world sustainability problems.
2026 – Present
PhD in Building Environment and Energy Engineering
The Hong Kong Polytechnic University
2023 – 2026
MS in Information and Communication Engineering
The University of Tokyo
2018 – 2022
BS in Electronic Information Science and Technology
Lanzhou University
Solar energyCo-first author
Combines 3D building footprints with meteorological data to estimate building-integrated PV potential in 120 cities worldwide; facade potential averages 68.2% of rooftop potential.
· Nexus
Solar energyCo-first author
Integrates AIoT, GIS and meteorological data into a facade BIPV potential model that accounts for 3D occlusion; facade element parsing and shading turn out to be key to accurate estimates.
· Advances in Applied Energy
Solar energy
A deep learning framework that encodes spatiotemporal data to predict shadow casting and rooftop PV potential (annual MAPE 2.85%, shading mIoU 89.23%), about 158 times faster than simulation.
· Applied Energy
2
3
Spatiotemporal deep learning that predicts shading and rooftop PV potential about 158 times faster than simulation.
Building-level facade and rooftop PV potential for 120 cities, from 3D footprints and meteorology.
Street-view facade element parsing (window, wall) coupled to 3D shading for facade PV.
2026
The Hong Kong Polytechnic University
