Telematics service detection and action messaging based on machine learning for assisting car sharing platform
AssigneeState Farm Mutual Automobile Insurance Company
InventorsJoseph Robert Brannan, Ryan Gross
Life of this patent
Estimating the term…
Abstract
A method of causing a user of a car sharing platform to perform an action with respect to a vehicle includes receiving a first data set, analyzing the first data set to determine vehicle imperatives, generating actions corresponding to the vehicle imperatives, analyzing the actions using trained machine learning models to generate action values, displaying the actions to the user, receiving acceptance of the user, receiving a second data set; and based on the acceptance and analyzing the second data set to identify changes to the vehicle imperatives indicating completed actions, releasing the action value to the user for each of the completed actions.
Claims20
A computer implemented method of causing a user of a car sharing platform to perform an action with respect to a vehicle, the method comprising:
- training, by one or more processors, a plurality of artificial neural network models by analyzing historical data, wherein each model is trained to generate respective actions values for actions corresponding to different respective vehicle imperative types, and wherein at least one of the artificial neural network models is trained to generate action values for actions corresponding to a vehicle imperative type other than relocation service imperatives,
- receiving, by the one or more processors, a first data set corresponding to the vehicle,
- analyzing, by the one or more processors, the first data set to determine one or more vehicle imperatives,
- generating, by the one or more processors and based on the one or more determined vehicle imperatives, one or more actions, each corresponding to one of the one or more determined vehicle imperatives,
- generating, by the one or more processors, a respective action value for each of the one or more generated actions by analyzing each of the one or more generated actions using a respective one of the plurality of trained artificial neural network models according to a vehicle imperative type associated with each of the one or more generated actions,
- displaying, by the one or more processors, the one or more generated actions to the user,
- receiving, by the one or more processors, an indication of acceptance of the user of at least one of the one or more generated actions,
- receiving, by the one or more processors, a second data set corresponding to the vehicle;
- analyzing, by the one or more processors, the second data set corresponding to the vehicle to identify changes to the one or more determined vehicle imperatives indicating one or more completed actions;
- and
- releasing, by the one or more processors and based on the analyzing of the second data set and the acceptance of the user, the respective generated action value to the user for each of the one or more completed actions.
A non-transitory computer readable medium containing program instructions that when executed, cause a computer system to:
- train, by one or more processors, a plurality of artificial neural network models by analyzing historical data, wherein each model is trained to generate respective actions values for actions corresponding to different respective vehicle imperative types, and wherein at least one of the artificial neural network models is trained to generate action values for actions corresponding to a vehicle imperative type other than relocation service imperatives,
- receive, by the one or more processors, a first data set corresponding to the vehicle,
- analyze, by the one or more processors, the first data set to determine one or more vehicle imperatives,
- generate, by the one or more processors and based on the one or more determined vehicle imperatives, one or more actions, each corresponding to one of the one or more determined vehicle imperatives,
- generate, by the one or more processors, a respective action value for each of the one or more generated actions by analyzing each of the one or more generated actions using a respective one of the plurality of trained artificial neural network models according to a vehicle imperative type associated with each of the one or more generated actions,
- display, by the one or more processors, the one or more generated actions to the user,
- receive, by the one or more processors, an indication of acceptance of the user of at least one of the one or more generated actions,
- receive, by the one or more processors, a second data set corresponding to the vehicle,
- analyze, by the one or more processors, the second data set corresponding to the vehicle to identify changes to the one or more determined vehicle imperatives indicating one or more completed actions;
- and
- release, by the one or more processors and based on the analysis of the second data set and the acceptance of the user, the respective generated action value to the user for each of the one or more completed actions.
Dependencies are inferred from the claim text.