PLAID: the Plug Load Appliance Identification Dataset

A Public Dataset of High Resolution for Load Identification Research

PLAID currently includes current and voltage measurements sampled at 30 kHz from 11 different appliance types present in more than 60 households in Pittsburgh, Pennsylvania, USA. Data collection took place during the summer of 2013, and winter of 2014. Each appliance type is represented by dozens of different instances of varying make/models. For each appliance, three to six measurements were collected for each state transition. These measurements were then post-processed to extract a few-second-long window containing both the steady-state operation and the startup transient (when available). Measurements with significant noise in the voltage due to measurement errors were removed, after which we were left with 1074 instances in PLAID 1 and 719 instances in PLAID 2.

The paper is avialable here. Please cite PLAID in your publications if it helps your research.


The IPython code to parse the data is available here. Check the Github Repo for more details.


The IPython code for GlobalSip 2015 paper titled "A feasibility study of automated plug-load identification from high-frequency measurements" is available here.


PLAID 2 has been released! You can download the dataset by clicking the Download Tab.

Check SustainIT 2017 paper titled "Handling imbalance in extended PLAID dataset" for more details.     The IPython code for this paper is available here.


Our PLAID Dataset has been re-organized and publised in Nature Scientific Data!

A sample instance from PLAID


A Fridge instance for 1 second

P-Q plane feature


Five random instances from each appliance type are plotted in P-Q plane.

VI trajectory feature


Current versus voltage plot of one steady state from four different appliances.


Template for Bootstrap, by @mdo.