Granted Patent B1

US 12,702,118 · App. 18/981,397 · Granted Aug 11, 2026

Inventors: Ali Ganjavian (London, GB); Fernando Key Portilla Kawamura (Basel, CH); Pablo Carrascal (Madrid, ES); Iñigo Javier Puente Henales (Madrid, ES); Anatoli Arkhipenko (New York, NY)

Assignee: MOGGIE LTD

A01K29/005A01K27/001H04W4/12

Abstract

Claims (42)

1 . A pet activity monitoring system, comprising:

a sensor platform comprising:

an inertial measurement unit configured to record time-series motion data; and

a fastening device for attaching the sensor platform to a collar;

a user device comprising a display; and

a server, configured to:

receive the time-series motion data generated by the sensor platform;

classify the time-series motion data as pertaining to an activity type of an animal;

generate a message based on the activity type using a natural language processing model; and

push the message to the user device;

where the user device is configured to display the message using the display.

2 . The activity monitoring system of claim 1 , wherein the server is further configured to:

continuously aggregate time-series motion data from the sensor platform;

classify the aggregated time-series motion data as pertaining to a sequence activity types; and

generate the message based on the sequence of activity types.

3 . The activity monitoring system of claim 1 , wherein the server is further configured to select the natural language processing model from a plurality of natural language processing models, where each natural language processing model in the plurality of natural language processing models is trained to output text in a different personality.

4 . The activity monitoring system of claim 1 , wherein at least one personality is a veterinarian personality.

5 . The activity monitoring system of claim 1 , wherein at least one personality is a personality type of the animal.

6 . The activity monitoring system of claim 1 , wherein the natural language processing model is a transformer model, and wherein the activity type is provided to the transformer model as an activity token.

7 . The activity monitoring system of claim 1 , wherein the server is further configured to generate a meme image of the animal based on the activity type and an image of the animal using a generative model.

8 . The activity monitoring system of claim 1 , wherein the sensor platform further comprises a temperature sensor configured to produce time-series temperature data; and

wherein the server is further configured to classify the time-series motion data using the time-series temperature data.

9 . The activity monitoring system of claim 1 , further comprising a base station configured to obtain data from the sensor platform and provide the data to the server.

10 . The activity monitoring system of claim 1 , wherein the animal is a cat.

11 . A pet activity monitoring method, comprising:

recording time-series motion data using a sensor platform mounted to an animal's collar, where the sensor platform comprises an inertial measurement unit configured to generate time-series motion data;

receiving the time-series motion data generated by the sensor platform at a server;

classifying the time-series motion data as pertaining to an activity type of an animal using a machine learning model of the server;

generating a message based on the activity type using a natural language processing model of the server;

pushing the message from the server to a user device; and

displaying the message using a display of the user device.

12 . The pet activity monitoring method of claim 11 , further comprising continuously aggregating time-series motion data from the sensor platform;

classifying the aggregated time-series motion data as pertaining to a sequence activity types; and

generating the message based on the sequence of activity types.

13 . The pet activity monitoring method of claim 11 , further comprising selecting the natural language processing model from a plurality of natural language processing models, where each natural language processing model in the plurality of natural language processing models is trained to output text in a different personality.

14 . The pet activity monitoring method of claim 11 , wherein at least one personality is a veterinarian personality.

15 . The pet activity monitoring method of claim 11 , wherein at least one personality is a personality type of the animal.

16 . The pet activity monitoring method of claim 11 , wherein the natural language processing model is a transformer model, and wherein the activity type is provided to the transformer model as an activity token.

17 . The pet activity monitoring method of claim 11 , further comprising generating a meme image of the animal based on the activity type and an image of the animal using a generative model of the server.

18 . The pet activity monitoring method of claim 11 , further comprising recording time-series temperature data using a temperature sensor of the sensor platform; and classifying the time-series motion data using the time-series temperature data.

19 . The pet activity monitoring method of claim 11 , further comprising using a base station to relay data from the sensor platform to the server.

20 . The pet activity monitoring method of claim 11 , wherein the animal is a cat.

Assignments (2)

Continuity (1)

References Cited (30)