US 2026/0310688 A1Published application

System and method for real-time monitoring utilizing machine-learning based upon sensor data

AssigneeState Farm Mutual Automobile Insurance Company

InventorsEdward W. Breitweiser, Julia Marsaglia, Eric R. Moore, Andrew T. Swartz, Harrison Lin, Eric A. Finley, Jennifer L. Crawford

Life of this patent

Estimating the term…

patentgrep

Loading patent…

Abstract

System and methods for analyzing data to generate personalized response(s) are disclosed. The method may include, such as by one or more processors, transceivers, and/or sensors: (1) receiving collected sensor data from external data stores; (2) utilizing an NLP model and a computer vision model to categorize the collected data into improvement category, positive category, or neutral category; (3) retrieving policy data corresponding to the user from data store(s); (4) inputting the categorized data and the policy data into a trained machine-learning model to generate personalized responses; (5) generating a presentation of the personalized responses in a device; (6) receiving updated sensor data from the sensor(s) associated with the user; (7) analyzing the updated sensor data to determine a compliance level for the categories to the personalized response(s); (8) comparing the compliance level to a category threshold for the categories; and/or (9) outputting an updated presentation based upon the comparing.

Claims21

Dependent claims

A computer-implemented method for analyzing data to generate one or more personalized responses, the computer-implemented method performed by one or more processors of a computing system in communication with one or more data sources, the computer-implemented method comprising:

receiving, by the one or more processors, collected sensor data from one or more external data stores, wherein the collected sensor data corresponds to one or more sensors associated with at least one user, and wherein the collected sensor data includes voice data and visual data associated with the at least one user;
utilizing, by the one or more processors, a natural language processing (NLP) model and a computer vision model to categorize the collected sensor data into one or more categories, wherein the one or more categories include at least one of (i) an improvement category, (ii) a positive category, or (iii) a neutral category, wherein the computer vision model analyzes the visual data to detect one or more facial expressions of the at least one user to identify one or more emotional indicators and body posture data that indicate physical comfort or discomfort, and wherein the NLP model analyzes the voice data to detect one or more changes in speech tone and a speech pace of the at least one user that indicate relaxation or stress, and wherein the categorizing is based upon the facial expressions and the one or more changes in speech tone and the speech pace;
retrieving, by the one or more processors, policy data corresponding to the at least one user from one or more data stores;
inputting, by the one or more processors, the categorized data and the policy data into a trained machine-learning model to generate the one or more personalized responses for the at least one user;
generating, by the one or more processors, a presentation of the one or more personalized responses in a user interface of at least one device associated with the at least one user;
receiving, by the one or more processors, updated sensor data from the one or more sensors associated with the at least one user, wherein the updated sensor data includes updated voice data and updated visual data;
analyzing, by the one or more processors, the updated sensor data to determine a compliance level for at least one of the one or more categories to the one or more personalized responses, wherein the analyzing includes processing the updated voice data and the updated visual data to determine adherence by the at least one user to the one or more personalized responses;
comparing, by the one or more processors, the compliance level to a category threshold corresponding to the one or more categories, wherein the compliance level below the category threshold triggers one or more adaptive strategies including one or more alternative recommendations;
and
outputting, by the one or more processors, an updated presentation based upon the comparing in the user interface of the at least one device.

7 more elements

A computer system for analyzing data to generate one or more personalized responses, comprising:

one or more processors of a computing system in communication with one or more data sources;
and
at least one non-transitory computer readable medium storing instructions which, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
receiving collected sensor data from one or more external data stores, wherein the collected sensor data corresponds to one or more sensors associated with at least one user, and wherein the collected sensor data includes voice data and visual data associated with the at least one user;
utilizing a natural language processing (NLP) model and a computer vision model to categorize the collected sensor data into one or more categories, wherein the one or more categories include at least one of (i) an improvement category, (ii) a positive category, or (iii) a neutral category, wherein the computer vision model analyzes the visual data to detect one or more facial expressions of the at least one user to identify one or more emotional indicators and body posture data that indicate physical comfort or discomfort, and wherein the NLP model analyzes the voice data to detect one or more changes in speech tone and a speech pace of the at least one user that indicate relaxation or stress, and wherein the categorizing is based upon the facial expressions and the one or more changes in speech tone and the speech pace;
retrieving policy data corresponding to the at least one user from one or more data stores;
inputting the categorized data and the policy data into a trained machine-learning model to generate the one or more personalized responses for the at least one user;
generating a presentation of the one or more personalized responses in a user interface of at least one device associated with the at least one user;
receiving updated sensor data from the one or more sensors associated with the at least one user, wherein the updated sensor data includes updated voice data and updated visual data;
analyzing the updated sensor data to determine a compliance level for at least one of the one or more categories to the one or more personalized responses, wherein the analyzing includes processing the updated voice data and the updated visual data to determine adherence by the at least one user to the one or more personalized responses;
comparing the compliance level to a category threshold corresponding to the one or more categories, wherein the compliance level below the category threshold triggers one or more adaptive strategies including one or more alternative recommendations;
and
outputting an updated presentation based upon the comparing in the user interface of the at least one device.

10 more elements

A non-transitory computer readable medium for analyzing data to generate one or more personalized responses, the non-transitory computer readable medium storing instructions which, when executed by one or more processors of a computing system in communication with one or more data sources, cause the one or more processors to perform operations comprising:

receiving collected sensor data from one or more external data stores, wherein the collected sensor data corresponds to one or more sensors associated with at least one user, and wherein the collected sensor data includes voice data and visual data associated with the at least one user;
utilizing a natural language processing (NLP) model and a computer vision model to categorize the collected sensor data into one or more categories, wherein the one or more categories include at least one of (i) an improvement category, (ii) a positive category, or (iii) a neutral category, wherein the computer vision model analyzes the visual data to detect one or more facial expressions of the at least one user to identify one or more emotional indicators and body posture data that indicate physical comfort or discomfort, and wherein the NLP model analyzes the voice data to detect one or more changes in speech tone and a speech pace of the at least one user that indicate relaxation or stress, and wherein the categorizing is based upon the facial expressions and the one or more changes in speech tone and the speech pace;
retrieving policy data corresponding to the at least one user from one or more data stores;
inputting the categorized data and the policy data into a trained machine-learning model to generate the one or more personalized responses for the at least one user;
generating a presentation of the one or more personalized responses in a user interface of at least one device associated with the at least one user;
receiving updated sensor data from the one or more sensors associated with the at least one user, wherein the updated sensor data includes updated voice data and updated visual data;
analyzing the updated sensor data to determine a compliance level for at least one of the one or more categories to the one or more personalized responses, wherein the analyzing includes processing the updated voice data and the updated visual data to determine adherence by the at least one user to the one or more personalized responses;
comparing the compliance level to a category threshold corresponding to the one or more categories, wherein the compliance level below the category threshold triggers one or more adaptive strategies including one or more alternative recommendations;
and
outputting an updated presentation based upon the comparing in the user interface of the at least one device.

7 more elements

Dependencies are inferred from the claim text.