Driver Risk Assessment System and Method Having Calibrating Automatic Event Scoring

• US 20100250021A1
• Filed: 06/11/2010
• Published: 09/30/2010
• Est. Priority Date: 01/26/2009
• Status: Active Grant

• Current Assignee: Lytx, Inc.
• Original Assignee: Lytx, Inc.
• Inventors: Gilles, Louis, Cook, Bryon
• Granted Patent: US 8,508,353 B2
• Time in Patent Office: Days
• US Class Current: 701/1
• CPC Class Codes: G07C 5/008 communicating information to a remotely located station transmission systems for measured values G08C
• CPC Class Codes: G07C 5/0808 Diagnosing performance data testing of vehicles G01M17/00; testing of electrical installation on vehicles G01R31/005
• CPC Class Codes: G07C 5/085 using electronic data carriers

Abstract

The system and method provides robust and reliable event scoring and reporting, while also optimizing data transmission bandwidth. The system includes onboard vehicular driving event detectors that record data related to detected driving events, selectively store or transfer data related to said detected driving events. If elected, the onboard vehicular system will score a detected driving event, compare the local score to historical values previously stored within the onboard system, and upload selective data or data types to a remote server or user if the system concludes that a serious driving event has occurred. Importantly, the onboard event scoring system, if enabled, will continuously evolve and improve in its reliability by being periodically re-calibrated the ongoing reliability results of manual human review of automated predictive event reports. The system may further respond to independent user requests by transferring select data to said user at a variety of locations and formats.

Claims

1. A system for reducing risk in driving, comprising:

at least one event capture device associated with a vehicle, said event capture device or devices detecting data related to the physical condition of said vehicle;

at least one event detector device coupled with the vehicle and configured to communicate with said event capture device or devices and to predict, responsive to analysis of the group of data captured by said event capture devices, whether or not data captured by said event capture devices represents a driving event, and assigning a unique risk identifier selected from a group of risk identities to said driving event;

an event scoring module attached to or otherwise associated to communicate with said event detector for generating an event score based upon the assigned said risk identifier data from said event capture device or devices and applying risk confidence data related to the specific risk; and

an event data management module attached to or otherwise associated to communicate with said event detector for uploading data from said event capture devices to a remote computing device, with said uploading being responsive to said event score.

2. The system of claim 1, wherein said system further comprises a risk confidence data repository configured to communicate with said event scoring module to periodically update said risk confidence data.

3. The system of claim 2, wherein said risk confidence data is updated regularly after human review of captured event data.

4. The system of claim 3, wherein said human review of captured event data results in the assignment of a manual score to a specific driving event and the risk identifier assigned thereto by said event detector or event detectors.

5. The system of claim 4, wherein said human review of captured event data results in the assignment of an severity score related to the riskiness of the driving observed by the human reviewer in the data of the captured event.

6. The system of claim 5, wherein said risk confidence data comprises a confidence factor for each said risk identifier in said group of risk identities, said confidence factor being a fractional value representing the statistical likelihood that each said risk identifier in said group of risk identities is representative of a risky driving event.

7. The system of claim 6, wherein said risk confidence data for each risk identity includes the percentage of times that human review concurred that the driving event to which the predicted risk identifier was assigned was correct.

8. The system of claim 7, wherein said risk confidence data for each risk identity includes a number representing the severity of the risk of the driving event to which the predicted risk identifier was assigned.

9. The system of claim 6, comprising:

at least one said event capture device generating said selectively uploaded data associated with said vehicle physical condition; and

at least another said event capture device generating said regularly uploaded data associated with said vehicle physical condition.

10. The system of claim 9, wherein said event data management module uploads said selectively uploaded data responsive to a comparison between said generated event score and a set of representative event scores stored in a local data repository associated with said comparison subsystem.

11. The system of claim 9, wherein said event score is based only upon said regularly uploaded data.

12. The system of claim 9, wherein said selectively uploaded data is selected from the group of audio data video data and telemetry data.

13. The system of claim 9, wherein said regularly uploaded data comprises vehicle velocity, vehicle acceleration, vehicle special orientation, vehicle metadata and vehicle location.

14. A method for evaluating risk in driving, comprising the steps of:

capturing driving event data at one or more event capture devices coupled with said vehicle;

analyzing the driving event data with at least one event detector device to assign a predicted driving event identifier selected from a pre-determined group of driving event identifiers;

assigning an event score to said predicted driving event identifier, said event score being either the result of actual human review of said driving event data associated with said predicted driving event identifier, or the historical confidence that prior human review of driving event data leading to the same predicted driving event identifier was confirmed; and

selectively uploading all or part of said captured data to a remote computing device, said selective uploading being responsive to said assigned event score.

15. The method of claim 14, wherein said selective uploading is further responsive to a data request received by said event detector.

16. The method of claim 15, wherein said data request received by said event detector is sent from a computing device that is remote to said vehicle.

17. The method of claim 16, wherein said event score calculation excludes said selectively uploaded data.

18. The method of claim 17, wherein said selectively uploaded captured data comprises audio or video data.

19. The method of claim 15, wherein said selective uploading is further responsive to user request.

20. A method for reducing risk in driving, comprising:

obtaining a plurality of driving event data for the individual driver, said event data comprising selectively uploadable video or audio data and regularly uploaded velocity, acceleration, vehicle spatial orientation and heading, vehicle metadata and vehicle location data, the driving event data obtained from a data storage area comprising a plurality of captured driving event data for a plurality of drivers;

analyzing the plurality of regularly uploadable driving event data;

analyzing demographic data for the individual driver, said demographic data obtained from a data storage area comprising a plurality of driver records, said records comprising demographic data; and

scoring the individual driver based on the analysis of the plurality of driving event data and the analysis of the demographic data; and

uploading relevant portions of said selectively uploadable video and audio data to a remote computing device responsive to a two-stage analysis of said driving event data, said first stage being the numerical analysis of the driving event data in order to generate a predicted event identifier selected from a group of potential event identities, and the second stage being the assignment of a confidence score to the generated predicted event identifier, with said confidence score being responsive to intermittent human review of predicted driving events.