Methods and systems for pattern-based identification of a driver of a vehicle

• US 10,845,381 B2
• Filed: 08/08/2018
• Issued: 11/24/2020
• Est. Priority Date: 06/24/2014
• Status: Active Grant

• Current Assignee: Cambridge Mobile Telematics, Inc.
• Original Assignee: Truemotion, Inc.
• Inventors: Cordova, Brad, Sahoo, Sanujit
• Primary Examiner(s): Ward, Eric A.
• Application Number:
  US16/058,745
• Publication Number:
  US 20190041423A1
• Time in Patent Office: 839 Days
• CPC Class Codes: B60W 40/09 Driving style or behaviour
• CPC Class Codes: G01P 15/18 in two or more dimensions

Abstract

A system for pattern-based identification of a driver of a vehicle includes a mobile device; and a server configured to communicate with the mobile device. The mobile device collects movement information during a driving trip via one or more sensors in the mobile device, and communicates the collected movement information to a server. The server analyzes the movement information via a classifier, identifies driving features for the driver based at least in part on the analysis of the movement information, creates an identification model for the driver based on identified driving features, and stores the identification model.

Claims

1. A method for identifying a driver that is driving a vehicle using collected information, the method comprising:

operating a mobile device to collect movement information and non-movement information during a driving trip via one or more sensors in the mobile device, wherein the non-movement information includes an interaction between the driver and the mobile device during the driving trip;

communicating the collected movement information and the non-movement information to a server;

analyzing the movement information and the non-movement information at the server via a classifier;

identifying driving features for the driver based at least in part on an analysis of the movement information and the non-movement information, wherein identifying driving features for the driver comprises;

determining a relative quaternion that represents an axis in a reference frame around which the mobile device turned when the mobile device entered the vehicle, wherein the relative quaternion further represents an angle of rotation of the mobile device about the axis;

determining a yaw angle of the mobile device using a gravity vector and the relative quaternion;

determining, based on the yaw angle of the mobile device, a probability that the mobile device entered the vehicle using a side of the vehicle that corresponds to a driving position within the vehicle;

creating an identification model for the driver based on identified driving features;

storing the identification model; and

identifying the driver during a subsequent driving trip based on the identification model.

2. The method of claim 1, further comprising collecting the non-movement information via one or more of a light sensor, a microphone, and a pressure gauge of the mobile device and communicating the collected movement information to the server, wherein the non-movement information is related to driving behavior of the driver.

3. The method of claim 1, wherein the non-movement information comprises driver interactions with the mobile device.

4. The method of claim 1, wherein the non-movement information comprises information collected from one or more smart devices linked to the mobile device.

5. The method of claim 1, further comprising receiving external information associated with the driving trip, and

analyzing, via the classifier, the external information as part of the identification model.

6. The method of claim 5, wherein the external information comprises at least one of traffic condition data, weather data, and road map data.

7. The method of claim 1, further comprising storing the movement information and the non-movement information collected during the driving trip on the mobile device, and

communicating the collected movement information and the non-movement information to the server subsequent to the driving trip.

8. The method of claim 1, further comprising collecting second movement information and second non-movement information during a subsequent driving trip and communicating the second movement information and the second non-movement information collected during the subsequent driving trip to the server for analysis,

analyzing, via the classifier, the second movement information and the second non-movement information;

and

based on the analyzing, identifying the driver based a comparison of features identified during the subsequent driving trip and the features of the stored identification model for the driver.

9. The method of claim 8, further comprising updating the identification model for the driver based the features identified during the subsequent driving trip.

10. The method of claim 1, wherein the identification model comprises demographic information input to the mobile device by the driver and communicated to the server.

11. A method for updating a driver identification model for a driver of a vehicle using collected information, the method comprising:

operating a mobile device placed in the vehicle to collect first movement information and first non-movement information during a first driving trip via one or more sensors in the mobile device, wherein the first non-movement information includes a first interaction between the driver of the vehicle and the mobile device during the first driving trip;

analyzing the first movement information and the first non-movement information with a classifier;

identifying driving features for the driver of the vehicle for the first driving trip based at least in part on an analysis, wherein identifying driving features for the driver comprises;

determining a relative quaternion that represents an axis in a reference frame around which the mobile device turned when the mobile device entered the vehicle, wherein the relative quaternion further represents an angle of rotation of the mobile device about the axis;

determining a yaw angle of the mobile device using a gravity vector and the relative quaternion;

determining, based on the yaw angle of the mobile device, a probability that the mobile device entered the vehicle using a side of the vehicle that corresponds to a driving position within the vehicle;

creating an identification model for the driver of the vehicle for the first driving trip based on identified driving features and storing the identification model;

collecting second movement information and second non-movement information during a second driving trip, wherein the second non-movement information includes a second interaction between the driver of the vehicle and the mobile device during the second driving trip;

analyzing the second movement information and the second non-movement information with the classifier to generate a feature representation for the driver of the vehicle for the second driving trip;

comparing, with the classifier, the generated feature representation with the stored identification model;

matching, by the classifier, the generated feature representation with the stored identification model; and

updating the identification model with additional driving features from the generated feature representation.

12. The method of claim 11, wherein the classifier matches features of the generated feature representation with a subset of driving features in the identification model, and

wherein non-matching features of the generated feature representation are added to the identification model.

13. The method of claim 11, wherein the classifier matches features of the generated feature representation with a subset of driving features in the identification model, and

wherein one or more driving features in the identification model are removed from the identification model after failing to match features of one or more generated feature representations.

14. The method of claim 11, further comprising collecting the first and second non-movement information via one or more of a light sensor, a microphone, and a pressure gauge of the mobile device, wherein the first and second non-movement information is related to driving behavior of the driver.

15. The method of claim 11, wherein the first and second non-movement information comprises information collected from one or more smart devices linked to the mobile device.

16. The method of claim 11, further comprising receiving first external information associated with the first driving trip,

analyzing, via the classifier, the first external information as part of the identification model;

receiving second external information associated with the second driving trip; and

analyzing, via the classifier, the second external information as part of the feature representation.

17. The method of claim 16, wherein the first and second external information comprises at least one of traffic condition data, weather data, and road map data.

18. The method of claim 11, wherein the classifier comprises a server, and

the collected first and second movement information and the first and second non-movement information is transferred to the server for analysis by the classifier.