I am a Postdoctoral Researcher at Stanford University. My research focuses on AI-enabled informatics for sustainability—integrating AI with physical principles and social theories to enable large-scale mapping, monitoring, and modeling of interplay among infrastructure systems (e.g., energy infrastructure), climate threats, and human factors. The goal of my research is to advance our understanding of the climate-human-infrastructure nexus and to create generalizable, human-centered information systems that provide actionable insights for policymakers, urban planners, system operators, communities, and other stakeholders to accelerate decarbonization, improve climate resilience, and ensure equity.

My current research topics include: adoption patterns of distributed energy resources (e.g., solar PVs, EV chargers), climate resilience of infrastructure networks, representation learning for urban systems. Recently, I am also interested in developing multi-modal machine learning foundation models for large-scale mapping of energy and environmental systems to enable fast knowledge discovery for supporting sustainability-related decision making.

Previously, I obtained my PhD degree in Civil and Environmental Engineering with minor in Computer Science at Stanford, co-advised by Prof. Ram Rajagopal and Prof. Arun Majumdar. I obtained my M.S. degree in Mechanical Engineering also from Stanford, and my Bachelor’s degree in Energy, Power System and Automation from Tsinghua University.

Email: zhecheng (at) stanford (dot) edu

Recent News

• 12/10/2024 Giving a talk about electrical grid exposure to wildfires at AGU24.

• 03/28/2024 Our paper on the potential of non-residential solar has been published in Nature Energy and covered by many media outlets (e.g., Stanford News, Tech Xplore).

• 02/08/2024 Invited talk at the Solar Colloquium hosted by DOE Solar Energy Technologies Office (SETO).

• 12/19/2023 We published SkyScript, a large and semantically diverse remote sensing image-text dataset for developing and evaluating vision-language models for remote sensing imagery. Its corresponding paper is accepted by AAAI 2024.