Agentic digital twin for urban building energy modeling, simulation, and decision support.

EnergyAtlas integrates urban-scale building energy simulation, geospatial data, measured energy data, and agentic AI workflows in a shared interactive environment. It supports workflows including:

• urban building energy modeling and scenario simulation;
• building-level calibration and model configuration;
• spatial exploration of buildings, loads, and energy-system characteristics;
• agent-driven simulation, analysis, reflection, and iterative model improvement.

Developed at the Environmental Systems Lab, Cornell University, Ithaca, NY.