Techniques for determining machine operation using audio

Patent metadata

Field	Value
Publication number	US-2021390137-A1
Application number	US-202117342140-A
Country	US
Kind code	A1
Filing date	Jun 8, 2021
Priority date	Jun 10, 2020
Publication date	Dec 16, 2021
Grant date	—

Abstract

Techniques for determining machine operation using audio are disclosed. For example, a machine may be registered in a database and first audio data of the machine representing sound made by the machine during an operation may be received. A first digital signature associated with operation of the machine may be determined using the audio data. Instructional data may be received to correlate the first digital signature with the registered machine. Later, second audio data representing sound present in an environment of the machine may be obtained, and a status of the machine may be determined by comparing a second digital signature generated using the second audio data and the first digital signature. The status may indicate one of a normal operation of the machine or that operation of the machine is indeterminable. A notification including the status may be transmitted to a user device.

First claim

1 . An apparatus for training to determine machine operation, comprising:
a processor; a memory storing instructions that, when executed by the processor, configure the apparatus to:
register a machine in a database using registration information to create a machine registration in the database;
obtain audio data of the machine, the audio data representing sound made by the machine during a normal operation of the machine;
generate a base set of discrete signals using the audio data;
determine a digital signature associated with the normal operation of the machine using the base set of discrete signals;
receive instructional data to correlate the digital signature with the machine; and
correlate the digital signature with the machine registration using the instructional data.

2 . The apparatus of claim 1 , wherein:
the apparatus further comprises one or more microphones; and the instructions, when executed by the processor, further configure the apparatus to:
receive the sound made by the machine during the normal operation of the machine using the one or more microphones; and
generate the audio data using the sound made by the machine during the normal operation of the machine.

3 . The apparatus of claim 1 , wherein the audio data is obtained via a network.

4 . The apparatus of claim 1 , wherein the audio data comprises a waveform.

5 . The apparatus of claim 1 , wherein the audio data comprises a plurality of digital signals representing the sound made by the machine during the normal operation of the machine.

6 . The apparatus of claim 1 , wherein the instructions, when executed by the processor, further configure the apparatus to send the digital signature to a server.

7 . The apparatus of claim 1 , wherein the registration information is received via a voice command from a user.

8 . The apparatus of claim 1 , wherein the instructions, when executed by the processor, further configure the apparatus to generate the base set of discrete signals using the audio data by filtering ambient audio data from the audio data.

9 . The apparatus of claim 1 , wherein the instructions, when executed by the processor, further configure the apparatus to generate the base set of discrete signals using the audio data by removing excursions from an initial set of discrete signals generated using the audio data.

10 . The apparatus of claim 1 , wherein the instructional data is received via a voice command from a user.

11 . The apparatus of claim 1 , wherein the apparatus comprises an electric vehicle supply equipment (EVSE).

12 . A method of a device for training to determine machine operation, comprising:
registering a machine in a database using registration information to create a machine registration in the database; obtaining audio data of the machine, the audio data representing sound made by the machine during a normal operation of the machine; generating a base set of discrete signals using the audio data; determining a digital signature associated with the normal operation of the machine using the base set of discrete signals; receiving instructional data to correlate the digital signature with the machine; and correlating the digital signature with the machine registration using the instructional data.

13 . The method of claim 12 , further comprising:
receiving the sound made by the machine during the normal operation of the machine; and generating the audio data using the sound made by the machine during the normal operation of the machine.

14 . The method of claim 12 , wherein the audio data is obtained via a network.

15 . The method of claim 12 , wherein the audio data comprises a waveform.

16 . The method of claim 12 , wherein the audio data comprises a plurality of digital signals representing the sound made by the machine during the normal operation of the machine.

17 . The method of claim 12 , further comprising sending the digital signature to a server.

18 . The method of claim 12 , wherein the registration information is received via a voice command from a user.

19 . The method of claim 12 , wherein generating the base set of discrete signals using the audio data comprises filtering ambient audio data from the audio data.

20 . The method of claim 12 , wherein generating the base set of discrete signals using the audio data comprises removing excursions from an initial set of discrete signals generated using the audio data.

21 . The method of claim 12 , wherein the instructional data is received via a voice command from a user.

22 . The method of claim 12 , wherein the device comprises an electric vehicle supply equipment (EVSE).

23 . A non-transitory computer-readable storage medium comprising instructions that when executed by a device, cause the device to:
register a machine in a database using registration information to create a machine registration in the database; obtain audio data of the machine, the audio data representing sound made by the machine during a normal operation of the machine; generate a base set of discrete signals using the audio data; determine a digital signature associated with the normal operation of the machine using the base set of discrete signals; receive instructional data to correlate the digital signature with the machine; and correlate the digital signature with the machine registration using the instructional data.

Assignees

• Enel X North America Inc

Inventors

• Feldman Daniel

Classifications

• G06N3/044

  Recurrent networks, e.g. Hopfield networks · CPC title

• G06N3/045

  Combinations of networks · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06N3/0442

  characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU] · CPC title