gtfierro/mortar

Dataset Card for Dataset Name

Mortar is a public dataset of anonymized Brick models and timeseries data from real buildings.

Brick models for these buildings are available at https://huggingface.co/datasets/gtfierro/mortargraphs

Dataset Details

The Mortar dataset is a dataset of real-world building metadata and timeseries data. This extensive collection provides rich timeseries data, such as temperature, occupancy, luminance, and HVAC sensor readings, across various commercial and institutional buildings. Each building's data is described with the Brick schema, a metadata standard for building operations that structures relationships among building assets, such as HVAC systems, lighting, and sensors.

• Temporal Coverage: Up to several years per building, mostly at 15-minute intervals
• Annotations: Brick model metadata for each building
• Primary Use Cases: Energy usage analysis, fault detection, demand response, occupancy analysis, and predictive maintenance.

BibTeX:

@article{fierro2019mortar,
  title={Mortar: an open testbed for portable building analytics},
  author={Fierro, Gabe and Pritoni, Marco and AbdelBaky, Moustafa and Lengyel, Daniel and Leyden, John and Prakash, Anand and Gupta, Pranav and Raftery, Paul and Peffer, Therese and Thomson, Greg and others},
  journal={ACM Transactions on Sensor Networks (TOSN)},
  volume={16},
  number={1},
  pages={1--31},
  year={2019},
  publisher={ACM New York, NY, USA}
}

Dataset Card Contact

Dr. Gabe Fierro (email found at https://gtf.fyi)