Activating a security mode for a vehicle based on driver identification
AssigneeMicron Technology, Inc.
InventorRobert Richard Noel Bielby
Life of this patent
Estimating the term…
Abstract
A controller of a vehicle detects persons that are approaching the vehicle and uses sensors to collect data regarding the persons to determine whether a security risk exists. In one approach, the collected data is used to determine whether each person is associated with the vehicle. In response to determining that at least one of the persons is unknown, a security risk is evaluated based on data from monitoring of the environment of the vehicle. In response to determining that a security risk score exceeds a threshold, one or more actions for the vehicle are performed to make the vehicle more secure.
Claims20
A method comprising:
- training a machine-learning model using prior data collected for an authorized first user of a vehicle;
- collecting sensor data from at least one sensor of the vehicle;
- collecting data corresponding to a person that is approaching the vehicle;
- determining, based on the collected data, whether the person is associated with the vehicle;
- determining a context of the vehicle, including a time of day, a current location of the vehicle, a route used by the vehicle to reach the currents location, and a presence of other vehicles at the current location;
- in response to determining that the person is the first user, determining a security risk based on monitoring an environment of the vehicle, wherein the security risk is determined based on an output from the machine-learning model, and wherein the machine-learning model has inputs including the determined context and the sensor data;
- and
- in response to determining the security risk, performing at least one action for the vehicle.
A system comprising:
- at least one sensor;
- at least one memory configured to store a database of persons associated with a vehicle;
- and
- at least one processing device;
- the at least one memory containing instructions configured to instruct the at least one processing device to:
- train a machine-learning model using prior data collected for a first person that is associated with the vehicle;
- collect sensor data using the at least one sensor, wherein the collected sensor data corresponds to the first person and a second person that are each approaching the vehicle;
- determine, based on the collected sensor data and by accessing the database, that the first person is associated with the vehicle, and that the second person is not associated with the vehicle;
- determine a context of the vehicle;
- determine a security risk based on an output from the machine-learning model, wherein the machine-learning model has inputs including the determined context and the sensor data;
- and
- in response to determining that the second person is not associated with the vehicle, perform, based on the security risk, at least one action, wherein performing the at least one action comprises selecting software based on the determined context, and launching the selected software.
(canceled)
A non-transitory computer-readable medium storing instructions which, when executed on a computing device, cause the computing device to at least:
- train a machine-learning model using prior data collected for an authorized driver of a vehicle;
- collect sensor data corresponding to a person that is approaching a vehicle;
- determine, based on the collected sensor data, whether the person is the authorized driver;
- determine a context of the vehicle;
- in response to determining that the person is the authorized driver, determine a security risk based on monitoring an environment of the vehicle, wherein the security risk is determined based on an output from the machine-learning model, and wherein the machine-learning model has inputs including the determined context and the sensor data;
- determine a classification of the security risk;
- and
- in response to determining the security risk, select software based on the classification, and launch the selected software.
Dependencies are inferred from the claim text.