DeepSolar-3M / README.md
rajanie's picture
Update README.md
500b14f verified
metadata
license: mit
tags:
  - rooftop-solar
  - solar-energy

DeepSolar-3M

🔨 Repo under active construction

📄 Paper: DeepSolar-3M: An AI-Enabled Solar PV Database Documenting 3 Million Systems Across the US
📍 Conference: ICLR 2025 - Tackling Climate Change with Machine Learning Workshop


Overview

DeepSolar-3M provides fast, high-resolution mapping of rooftop photovoltaic (PV) systems across the United States.
This repository contains county-level and blockgroup-level aggregated data from our AI pipeline.

Key features:

  • Scalable detection of PV installations from aerial imagery
  • Blockgroup-level and county-level aggregation of PV system statistics
  • Detailed breakdowns by system type (residential, commercial, utility-scale, solar heating)

📊 County-Level Dataset

Each entry corresponds to a U.S. county (identified by FIPS code) and includes:

  • Total PV system count
  • Total PV area (in square meters)
  • Median PV area (m²)
  • Average PV area (m²)

Breakdown by system type (% of systems):

  • Residential systems
  • Commercial systems
  • Utility-scale systems
  • Solar heating systems

🗺️ Block Group-Level Dataset

Each entry corresponds to a U.S. Census block group (identified by GEOID/Block Group FIPS) and includes all the features listed above.


📬 Citation

If you find this resource useful, please cite:

@inproceedings{prabha2025deepsolar3m,
  title={DeepSolar-3M: An AI-Enabled Solar PV Database Documenting 3 Million Systems Across the US},
  author={Prabha, Rajanie and Wang, Zhecheng and Zanocco, Chad and Flora, June and Rajagopal, Ram },
  booktitle={ICLR 2025 Workshop on Tackling Climate Change with Machine Learning},
  url={https://www.climatechange.ai/papers/iclr2025/55},
  year={2025}
}

Contact

Feel free to reach out in case you have any questions - rajanie@stanford.edu