Detecting anomalous events using a microcontroller

Patent metadata

Field	Value
Publication number	US-12140940-B2
Application number	US-202117142132-A
Country	US
Kind code	B2
Filing date	Jan 5, 2021
Priority date	Jan 5, 2021
Publication date	Nov 12, 2024
Grant date	Nov 12, 2024

Abstract

In one embodiment, a method performed by a microcontroller of an electronic device includes accessing one or more real-time sensor data associated with one or more sensors of the electronic device, determining, by a machine-learning model running on the microcontroller, that an anomalous event has occurred on the electronic device by processing the one or more real-time sensor data with the machine-learning model, and sending, upon the determination that the anomalous event has occurred, a notification regarding the anomalous event to an application running on the electronic device.

First claim

What is claimed is:

1. A battery-powered electronic device comprising:
one or more non-transitory computer-readable storage media;
one or more processors coupled to the storage media;
one or more sensors; and
a microcontroller configured, while the battery-powered electronic device is in an idle mode, to:
access one or more real-time sensor data associated with one or more hardware components of the battery-powered electronic device collected by the one or more sensors, each monitoring a corresponding hardware component of the battery-powered electronic device;
determine, by a machine-learning model running on the microcontroller, that an anomalous event on the battery-powered electronic device has occurred by processing the one or more real-time sensor data associated with one or more hardware components of the battery-powered electronic device with the machine-learning model; and
send, upon the determination that the anomalous event has occurred on the battery-powered electronic device, a notification regarding the anomalous event to an application running on the battery-powered electronic device, wherein the notification causes the battery-powered electronic device to switch from the idle mode to an active mode.

2. The battery-powered electronic device of claim 1 , wherein the one or more real-time sensor data associated with one or more hardware components of the battery-powered electronic device comprise a current clock speed for one of the one or more processors, a current storage media utilization, a current battery level, a current network utilization, a current status for one of one or more displays, or a current device temperature.

3. The battery-powered electronic device of claim 1 , wherein the microcontroller triggers a system interrupt to send the notification.

4. The battery-powered electronic device of claim 1 , wherein the machine-learning model has been trained using sensor data collected from electronic devices having hardware configurations substantially similar to the battery-powered electronic device, and wherein the collected sensor data is labelled.

5. The battery-powered electronic device of claim 1 , wherein the machine-learning model has been trained using sensor data collected from electronic devices having software configurations substantially similar to the battery-powered electronic device.

6. The battery-powered electronic device of claim 1 , wherein the application takes further action to determine whether the anomalous event is associated with an abnormal operation mode of the battery-powered electronic device, and wherein the application provides a result of the determination to the microcontroller.

7. The battery-powered electronic device of claim 6 , wherein the further action comprises:
providing a visual notification to a user on one or more of displays associated with the battery-powered electronic device, wherein the visual notification comprises one or more choices for the user to provide a confirmation or a rejection to the visual notification; and
receiving a selection among the one or more choices from the user.

8. The battery-powered electronic device of claim 6 , wherein the microcontroller is further configured to:
receive the result from the application; and
update the machine-learning model based on the received result.

9. The battery-powered electronic device of claim 1 , wherein the battery-powered electronic device is a mobile electronic device.

10. The battery-powered electronic device of claim 1 , wherein the microcontroller is a sensor hub associated with the battery-powered electronic device.

11. The battery-powered electronic device of claim 1 , wherein the battery-powered electronic device enters into the idle mode when the battery-powered electronic device has not detected any activity on installed applications for a pre-determined duration of time, wherein the battery-powered electronic device enters into the active mode when the battery-powered electronic device detects an activity on any of the installed applications, and wherein the microcontroller keeps the determining regardless of a mode the battery-powered electronic device is on.

12. The battery-powered electronic device of claim 1 , wherein the one or more processors are configured to access second sensor data from external sensors including one or more cameras, touch sensors, microphones, or motion detection sensors, and wherein the second sensor data is not used by the microcontroller for determining that the anomalous event on the battery-powered electronic device has occurred.

13. A method comprising, by a microcontroller of a battery-powered electronic device, while the battery-powered electronic device is in an idle mode:
accessing one or more real-time sensor data associated with one or more hardware components of the battery-powered electronic device collected by one or more sensors of the battery-powered electronic device, each monitoring a corresponding hardware component of the battery-powered electronic device;
determining, by a machine-learning model on the microcontroller, that an anomalous event on the battery-powered electronic device has occurred by processing the one or more real-time sensor data associated with one or more hardware components of the battery-powered electronic device with the machine-learning model; and
sending, upon the determination that the anomalous event has occurred on the battery-powered electronic device, a notification regarding the anomalous event to an application running on the battery-powered electronic device, wherein the notification causes the battery-powered electronic device to switch from the idle mode to an active mode.

14. The method of claim 13 , wherein the microcontroller triggers a system interrupt to send the notification.

15. The method of claim 13 , wherein the application takes further action to determine whether the anomalous event is associated with an abnormal operation mode of the battery-powered electronic device, and wherein the application provides a result of the determination to the microcontroller.

16. The method of claim 15 , further comprising:
receiving the result from the application; and
updating the machine-learning model based on the received result.

17. A computer-readable non-transitory storage media comprising instructions executable by a microcontroller of a battery-powered electronic device, while the battery-powered electronic device is in an idle mode, to:
access one or more real-time sensor data associated with one or more hardware components of the battery-powered electronic device collected by one or more sensors of the battery-powered electronic device, each monitoring a corresponding hardware component of the battery-powered electronic device;
determine, by a machine-learning model running on the microcontroller, that an anomalous event on the battery-powered electronic device has occurred by processing the one or more real-time sensor data associated with one or more hardware components of the battery-powered electronic device with the machine-learning model; and
send, upon the determination that the anomalous event has occurred on the battery-powered electronic device, a notification regarding the anomalous event to an application running on the battery-powered electronic device, wherein the notification causes the battery-powered electronic device to switch from the idle mode to an active mode.

18. The media of claim 17 , wherein the microcontroller triggers a system interrupt to send the notification.

19. The

Assignees

• Samsung Electronics Co Ltd

Inventors

• Wong Erick
• O'Brien Sean
• Wong Julia

Classifications

• G06N20/00

  Machine learning · CPC title

• G05B23/0243

  model based detection method, e.g. first-principles knowledge model · CPC title

• G05B23/0259 Primary

  characterized by the response to fault detection · CPC title