CGBC Releases 2025-2026 Dataset Using Automated Building Data Process

The Harvard Center for Green Buildings and Cities (CGBC) has developed an automated process that enables HouseZero®—the Center’s naturally ventilated, ultra-low-energy headquarters and living laboratory—to produce high-fidelity, minute-level operational data.

The dataset contains one year of minute-by-minute data collected from 190 sensors and meters, generating millions of data points. It covers energy use, photovoltaic power generation and indoor environmental conditions, including air temperature, carbon dioxide levels and relative humidity. It also captures outdoor microclimate conditions, façade temperatures and advanced building-system operations.

The high-resolution dataset fills a gap in publicly available building-performance data and provides a resource for studying energy-efficient buildings under real-world conditions.

The automated data collection, filtering, quality-control and processing methodology follows the approach documented in the peer-reviewed paper, “Minute-Level Dataset from a Naturally Ventilated Building for Benchmarking and Learning-Based Modeling.”

To learn more, visit the HouseZero Dataset webpage.