Assisted labeling of devices with disaggregation

• US 9,691,030 B2
• Filed: 07/22/2016
• Issued: 06/27/2017
• Est. Priority Date: 11/26/2014
• Status: Active Grant

• Current Assignee: Sense Labs, Inc.
• Original Assignee: Sense Labs, Inc.
• Inventors: Micali, Christopher M., MacAulay, Vanessa Lee, Houlette, Ryan T., Phillips, Michael S.
• Primary Examiner(s): Hill, Stanley K
• Assistant Examiner(s): Bharadwaj, Kalpana
• Application Number:
  US15/217,420
• Publication Number:
  US 20160328651A1
• Time in Patent Office: 340 Days
• CPC Class Codes: G01R 21/133 by using digital technique
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06Q 10/20 Administration of product repair or maintenance
• CPC Class Codes: G06Q 50/06 Energy or water supply

Abstract

Electrical usage of devices in a building may be monitored to assist a user in labeling devices. In some implementations, a user may provide information about a first device, and this information may be used to obtain a plurality of mathematical models that each correspond to a device or a class of devices. The mathematical models may be used to process an electrical signal, and a score may be generated for each mathematical model. A mathematical model may be selected from the plurality of mathematical models using the scores, such as selecting a mathematical model with a highest score. The first device may be associated with the selected mathematical model.

Claims

1. One or more non-transitory computer-readable media comprising computer executable instructions that, when executed, cause at least one processor to perform actions comprising:

causing a prompt to be displayed to a user requesting that the user provide first information corresponding to a first device;

receiving the first information from the user corresponding to the first device;

obtaining a plurality of mathematical models using the first information, wherein each model of the plurality of mathematical models corresponds to a device or a class of devices;

measuring an electrical signal from the building, wherein the electrical signal powers a plurality of devices;

computing a plurality of features from the measured electrical signal;

computing a plurality of scores using the plurality of mathematical models and the plurality of features, wherein (i) the plurality of mathematical models comprises a first mathematical model and (ii) the plurality of scores comprises a first score computed using the first model and the plurality of features;

selecting the first mathematical model using the plurality of scores; and

associating the first device with the first mathematical model.

2. The one or more non-transitory computer-readable media of claim 1, the actions further comprising:

after causing the prompt to be displayed and before computing the plurality of features, causing a second prompt to be displayed to the user requesting that that user change a state of the first device.

3. The one or more non-transitory computer-readable media of claim 1, the actions further comprising:

receiving an indication from the user that the user has changed a state of the first device.

4. The one or more non-transitory computer-readable media of claim 1, wherein the prompt is displayed to the user in response to the user performing an action.

5. The one or more non-transitory computer-readable media of claim 1, wherein the first information corresponds to a class of the first device, a make of the first device, or a version of the first device.

6. The one or more non-transitory computer-readable media of claim 1, wherein the first mathematical model comprises a transition model or a device model.

7. The one or more non-transitory computer-readable media of claim 1, wherein selecting the first mathematical model using the plurality of scores comprises selecting a model with a highest score.

8. A computer-implemented method for labeling devices in a building utilizing an electrical signal representative of the electrical usage of the devices, the method comprising:

receiving first information from a user corresponding to a first device in the building;

obtaining a plurality of mathematical models using the first information, wherein each model of the plurality of mathematical models corresponds to a device or a class of devices;

computing a plurality of features from the electrical signal;

computing a plurality of scores using the plurality of mathematical models and the plurality of features;

selecting a first mathematical model from the plurality of mathematical models using the plurality of scores; and

associating the first device with the first mathematical model.

9. The method of claim 8, further comprising requesting a user to change a state of the first device.

10. The method of claim 9 further comprising receiving an indication from the user that the user has changed a state of the first device.

11. The method of claim 8, wherein the first information corresponds to a class of the first device, a make of the first device, or a version of the first device.

12. The method of claim 8, wherein the first mathematical model comprises a transition model or a device model.

13. The method of claim 8, wherein selecting a first mathematical model from the plurality of mathematical models using the plurality of scores further comprises selecting a model from the plurality of models with a highest score.

14. The method of claim 8, wherein the first information indicates that a make or version of the first device is unknown.

15. A system for labeling devices in a building utilizing an electrical signal representative of the electrical usage of the devices, the system comprising:

one or more computers comprising at least one processor and at least one memory, the one or more computers configured to;

receive first information from a user corresponding to a first device in the building;

obtain a plurality of mathematical models using the first information, wherein each model of the plurality of mathematical models corresponds to a device or a class of devices;

compute a plurality of features from the electrical signal;

compute a plurality of scores using the plurality of mathematical models and the plurality of features;

select a first mathematical model from the plurality of mathematical models using the plurality of scores; and

associate the first device with the first mathematical model.

16. The system of claim 15, wherein the one or more computers are further configured to request a user to change a state of the first device.

17. The system of claim 16, wherein the one or more computers are further configured to receive an indication from the user that the user has changed a state of the first device.

18. The system of claim 15, wherein the first information corresponds to a class of the first device, a make of the first device, or a version of the first device.

19. The system of claim 15, wherein the first mathematical model comprises a transition model or a device model.

20. The system of claim 15, wherein selecting a first mathematical model from the plurality of mathematical models using the plurality of scores further comprises selecting a model from the plurality of models with a highest score.