The EMBED (Energy Monitoring through Building Electricity Disaggregation) dataset is released to facilitate research and performance benchmarking in the disaggregation domain. EMBED contains data from three apartment units in Los Angeles, California, collected for at least two weeks. Aggregate power files, fully labeled appliance event timestamps, and plug load consumption for a variety of monitored appliances is provided in this page.

The dataset attributes is provided below.

Feature Value
Current sampling rate 12kHz
Voltage sampling rate 12kHz
Power resolution 60 Hz
Plug load sampling rate 1-2 Hz
Power metrics Real(P), Reactive(Q)
Calculated power harmonic content Up to 9th harmonics
Apt 1 Duration 14 days
Apt 2 Duration 21 days
Apt 3 Duration 27 days

For more information about the dataset, please refer to the paper:

F. Jazizadeh, M. Afzalan, B. Becerik-Gerber, L. Soibelman, “EMBED: A Dataset for Energy Monitoring through Building Electricity Disaggregation”, in Proceedings of the Ninth ACM International Conference on Future Energy Systems (ACM e-Energy’18), Karlsruhe, Germany, 2018.


Apartment 1 power files

Apartment 2 power files

Apartment 3 power files

Ground truth plug load consumption

Labeled event timestamps

Appliance label description