Pecan Street has developed a residential electric use disaggregation algorithm training kit, previously available only to members of its university consortium and clients of its algorithm evaluation service, that is now publicly available. This unique kit includes a 15-minute whole home dataset and 1-minute interval circuit-level dataset for packages of 10 to 100 homes in Austin covering a one-year period. Algorithms can be run against the whole home data and the output compared to the circuit-level data. The circuit-level data includes furnace, compressor, refrigerator, oven, PV generation (when present), EV charging (when present) and other major circuits in the home, such as small kitchen appliances, lighting and bedrooms. Organizations interested in purchasing the training kit can contact Fisayo Fadelu at ffadelu@pecanstreet.org.
