Smart* Data Set for Sustainability

The goal of the Smart* project is to optimize home energy consumption. Available here is a wide variety of data collected from three real homes, including electrical (usage and generation), environmental (e.g., temperature and humidity), and operational (e.g., wall switch events). Also available is minute-level electricity usage data from 400+ anonymous homes. Please see the Smart* home page for general information about the project, or the Smart* Tools download page for software that was used in the collection of this data.

Questions? All questions regarding these traces should be directed to Erik Risinger.

Traces

All traces are in CSV format, with data fields named in the included FORMAT files. Please see our dataset paper for further details. Data collection and processing is ongoing, so check this page for future updates.

UMass Dataset - 2019

DeepRoof dataset

The deeproof dataset contains satellite images of roofs along with the planar roof segments of each roof. The folder also contains source code to visualize the dataset.

• DeepRoof dataset :deeproof.tar.gz (675 MB) SHA256:a51c3f5e63ada5c8ba7ff9dbbc209bb34d97c18b73a9ece09116cfeb1f628f21

UMass Smart* Dataset - 2017 release

Apartment dataset

The apartment dataset contains data for 114 single-family apartments for the period 2014-2016.

• Electrical dataset (aggregated):apartment-electrical.tar.gz (354 MB)
• Weather dataset apartment-weather.tar.gz (642 kB)

Home dataset

The traces are made available for multiple years in CSV format and includes a SUMMARY for each file.

• HomeA
  • Electrical dataset (aggregated and individual circuits):HomeA-electrical.tar.gz (24 MB)
  • Weather dataset HomeA-weather.tar.gz (651 kB)
• HomeB
  • Electrical dataset (aggregated and individual circuits):HomeB-electrical.tar.gz (24 MB)
  • Weather dataset HomeB-weather.tar.gz (643 kB)
• HomeC
  • Electrical dataset (aggregated and individual circuits):HomeC-electrical.tar.gz (28 MB)
  • Weather dataset HomeC-weather.tar.gz (642 kB)
• HomeD
  • Electrical dataset (aggregated and individual circuits):HomeD-electrical.tar.gz (65 MB)
  • Weather dataset HomeD-weather.tar.gz (435 kB)
• HomeF
  • Electrical dataset (aggregated and individual circuits):HomeF-electrical.tar.gz (47 MB)
  • Weather dataset HomeF-weather.tar.gz (642 kB)
• HomeG
  • Electrical dataset (aggregated and individual circuits):HomeG-electrical.tar.gz (42 MB)
  • Weather dataset HomeG-weather.tar.gz (433 kB)
• HomeH
  • Electrical dataset (aggregated and individual circuits):HomeH-electrical.tar.gz (8.9 MB)
  • Weather dataset HomeH-weather.tar.gz (216 kB)

NIOM occupancy dataset

This dataset contains 3 weeks minute level aggregated energy data, and the ground truth occupancy status data for the same periods. This dataset is used to evaluate NIOM algorithm.

• Home A
  • Spring- Aggregated electrical data: Home-A-spring-energy.tar.gz (80 kB)
  • Spring- Ground truth occupancy status data (two occupants): Home-A-spring-occupancy.tar.gz (99 kB)
  • Summer-Aggregated electrical data: Home-A-summer_energy.tar.gz (82 kB)
  • Summer-Ground truth Occupancy status data (two occupants): Home-A-summer-occupancy.tar.gz (99 kB)
• Home B
  • Summer - Aggregated electrical data: Home-B-summer_energy.tar.gz (81 kB)
  • Summer - Occupancy status data (two occupants): Home-B-summer.tar.gz (81 kB)
• All of the above: occupancy-niom.tar.gz (542 kB)

Solar panel dataset

This dataset contains 1 minute level solar generation data for 50 rooftop solar panels. This dataset is used in SunSpot paper to evaluate the accuracy of the SunSpot system.

• Solar dataset: solar-panels.tar.gz

SunDance dataset

This dataset includes hourly energy data (net meter, solar generation) and weather data (weather condition data from public weather stations and apis) for 100 solar sites in North America from 2015/1/1 to 2016/1/1 used in SunDance paper.

• Sundance dataset: sundance-data-release.tar.gz (58 MB)

Physical-blackbox Model Dataset

This dataset includes the weather and the normalized solar generation data to learn the physical blackbox model (BuildSys'18). Also the code to model the shading effects is included. Code and dataset: physicalmodel-data-release.tar.gz (163 MB)

UMass Smart* Dataset - 2013 release

 Update (Apr 2013): The initial dataset release contained incorrect environmental data for homes A and B. 
 Corrected environmental traces have been posted.

Home dataset

• Home A
  • Electrical data from circuits (and aggregate): homeA-circuit.tar.gz (161 MB)
  • Electrical data from individual energy meters: homeA-meter.tar.gz (3.2 MB)
  • Electrical data from dimmable and non-dimmable switches: homeA-switch.tar.gz (145 KB)
  • Voltage and frequency data on both electrical phases: homeA-phase.tar.gz (79 MB)
  • Environmental data (indoor and outdoor): homeA-environmental.tar.gz (577 KB)
  • Furnace on/off data: homeA-furnace.tar.gz (2.2 KB)
  • Door open/close data: homeA-door.tar.gz (43 KB)
  • Motion detector data: homeA-motion.tar.gz (560 KB)
  • All of the above: homeA-all.tar.gz (246 MB)
• Home B
  • Aggregate electrical data: homeB-power.tar.gz (25 MB)
  • Environmental data: homeB-environmental.tar.gz (446 KB)
  • All of the above: homeB-all.tar.gz (25 MB)
• Home C
  • Aggregate electrical data: homeC-power.tar.gz (27 MB)
  • Environmental data: homeC-environmental.tar.gz (1.2 MB)
  • Generation data (solar, wind, and battery voltage): homeC-generation.tar.gz (8.7 MB)
  • All of the above: homeC-all.tar.gz (36 MB)
• Microgrid dataset
  • Electrical data over a single 24-hour period from 443 unique homes: microgrid.tar.gz (3.5 MB)

Solar-TK Dataset

This dataset contains solar generation data for 81 homes across the United States. The files are in CSV format and have three columns: Timestamps, Local_Time, and Solar. The files are named Home_N_X_Y.csv, where N is the home number, X and Y are the coarse latitude and longitude. The current version just has the letters X and Y. The coarse coordinates for the homes will be uploaded in the coming days.

The dataset is available here.

Publications:

• [Home dataset] Smart*: An Open Data Set and Tools for Enabling Research in Sustainable Homes. Sean Barker, Aditya Mishra, David Irwin, Emmanuel Cecchet, Prashant Shenoy, and Jeannie Albrecht. Proceedings of the 2012 Workshop on Data Mining Applications in Sustainability (SustKDD 2012), Beijing, China, August 2012. pdf
• [Home dataset] SmartCharge: Cutting the Electricity Bill in Smart Homes with Energy Storage. Aditya Mishra, David Irwin, Prashant Shenoy, Jim Kurose, and Ting Zhu. Proceedings of the 3rd International Conference on Future Energy Systems (e-Energy 2012), Madrid, Spain, May 2012. pdf
• [Home dataset] SmartCap: Flattening Peak Electricity Demand in Smart Homes. Sean Barker, Aditya Mishra, David Irwin, Prashant Shenoy, and Jeannie Albrecht. Proceedings of the 10th IEEE International Conference on Pervasive Computing and Communications (PerCom 2012), Lugano, Switzerland, March 2012. pdf
• [Home dataset] Exploiting Home Automation Protocols for Load Monitoring in Smart Buildings. David Irwin, Anthony Wu, Sean Barker, Aditya Mishra, Prashant Shenoy, and Jeannie Albrecht. Proceedings of the 3rd ACM Workshop on Embedded Sensing Systems for Energy-Efficiency in Buildings (BuildSys 2011), Seattle, WA, November 2011. pdf
• [Home dataset] The Case for Efficient Renewable Energy Management for Smart Homes. Ting Zhu, Aditya Mishra, David Irwin, Navin Sharma, Prashant Shenoy, and Don Towsley. Proceedings of the 3rd ACM Workshop on Embedded Sensing Systems for Energy-Efficiency in Buildings (BuildSys 2011), Seattle, WA, November 2011. pdf
• [Solar dataset] SunSpot: Exposing the Location of Anonymous Solar-powered Homes. Dong Chen, Srinivasan Iyengar, David Irwin, Prashant Shenoy. Proceedings of the ACM International Conference on Systems for Energy Efficient Build Environments (Buildsys 2016), Palo Alto, CA, November 16-17, 2016 pdf
• [NIOM dataset] Non-Intrusive Occupancy Monitoring using Smart Meters. Dong Chen, Sean Barker, Adarsh Subbaswamy, David Irwin, and Prashant Shenoy. Proceedings of ACM BuildSys 2013, Rome, Italy, November 11-15, 2013 pdf
• [SunDance dataset] SunDance: Black-box Behind-the-Meter Solar Disaggregation. Dong Chen, David Irwin. In Proceedings of the eighth ACM International Conference on Future Energy Systems (e-Energy'17), Hong Kong, 2017.
• [Weatherman Dataset] Weatherman: Exposing Weather-based Privacy Threats in Big Energy Data. Dong Chen, David Irwin. In Proceedings of 2017 IEEE International Conference on Big Data (BigData'17), Boston, MA, USA, Dec 11-14, 2017.
• [Physical-blackbox Model Dataset] Staring at the Sun: A Physical Black-box Solar Performance Model. Dong Chen, Joseph Breda, David Irwin. In Proc. of the 2018 ACM International Conference on Systems for Energy-Efficient Built Environments (BuildSys'18), 2018.
• [DeepRoof Dataset] DeepRoof: A Data-driven Approach For Solar Potential Estimation for Rooftop Imagery. Stephen Lee, Srinivasan Iyengar, Menghong Feng, Prashant Shenoy, Subhransu Maji. Proceedings of 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD 2019), Anchorage, Alaska, August 2019.