TRAINING A SELF-DRIVING VEHICLE

• US 20180237005A1
• Filed: 02/22/2017
• Published: 08/23/2018
• Est. Priority Date: 02/22/2017
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: DUAN, NING, HU, GUOQIANG, HUANG, JING CHANG, YAN, JUNCHI, ZHU, JUN
• Granted Patent: US 10,752,239 B2
• Time in Patent Office: Days
• CPC Class Codes: B60W 2040/0881 Seat occupation; Driver or passenger presence
• CPC Class Codes: B60W 2530/10 Weight
• CPC Class Codes: B60W 2554/00 Input parameters relating to objects
• CPC Class Codes: B60W 2556/45 External transmission of data to or from the vehicle
• CPC Class Codes: B60W 30/085 Taking automatic action to adjust vehicle attitude in preparation for collision, e.g. braking for nose dropping
• CPC Class Codes: B60W 30/09 Taking automatic action to avoid collision, e.g. braking and steering
• CPC Class Codes: B62D 15/025 Active steering aids, e.g. helping the driver by actively influencing the steering system after environment evaluation B62D1/28 takes precedence
• CPC Class Codes: B62D 15/0265 Automatic obstacle avoidance by steering
• CPC Class Codes: G08G 1/166 for active traffic, e.g. moving vehicles, pedestrians, bikes

Abstract

One or more sensors associated with a vehicle detect a roadway condition of a first roadway and an evasive maneuver is performed by the vehicle to avoid the detected roadway condition. In response to one or more processors determining that the evasive maneuver was successful, a record of the successful maneuver and the roadway condition are stored in a database. Subsequent to storing the record in the database, one or more computers associated with one or more vehicles is trained to execute the evasive maneuver, in response to determining that the one or more vehicles is exposed to the roadway condition experienced by the vehicle.

Claims

1. A computer-implemented method comprising:

detecting, by one or more sensors associated with a vehicle, a roadway condition of a first roadway;

detecting, by the one or more sensors associated with the vehicle, an evasive maneuver performed in response to detecting the roadway condition;

determining, by a computer associated with the vehicle, whether the evasive maneuver was successful in avoiding the roadway condition;

storing, by the computer, a record of the evasive maneuver and the roadway condition, in response to said determining whether the evasive maneuver was a successful maneuver for avoiding the roadway condition;

training one or more computers associated with one or more vehicles, to execute the evasive maneuver, in response to determining that the one or more vehicles is exposed to the roadway condition experienced by the vehicle.

2. The computer-implemented method of claim 1, further comprising:

in response to determining that the evasive maneuver was an unsuccessful maneuver, blocking transmission of the unsuccessful maneuver to the one or more vehicles.

3. The computer-implemented method of claim 1, wherein the roadway condition is a traffic hazard from a group consisting of:

a second vehicle proximate to the vehicle at a time of the evasive maneuver; and

a fixed obstruction on the first roadway; and

wherein

the evasive maneuver is performed to avoid the roadway condition.

4. The computer-implemented method of claim 1, wherein the roadway condition is determined to involve a hazard to a pedestrian proximate to the vehicle at a time of the evasive maneuver, and wherein the evasive maneuver is performed to avoid striking the pedestrian.

5. The computer-implemented method of claim 1, wherein the vehicle is a self-driving vehicle (SDV), and wherein the method further comprises:

detecting, by one or more processors, that an imminent collision by the SDV is imminent with a confidence C1;

determining, by one or more processors, whether the SDV has one or more occupants with a confidence C2;

identifying, by one or more processors, an object with which the imminent collision by the SDV is imminent with a confidence C3;

generating, by one or more processors, an ameliorative action for the SDV based on C1, C2, C3; and

directing the on-board processor associated with the SDV to execute the ameliorative action.

6. The computer-implemented method of claim 5, further comprising:

determining, by one or more processors, confidence C1 based on an analysis of sensor data received in real-time from one or more sensors on the SDV.

7. The computer-implemented method of claim 5, wherein the one or more occupants are selected from a group consisting of animate passengers and inanimate passengers.

8. The computer-implemented method of claim 5, wherein the SDV has one or more crumple zones, further comprising:

directing the on-board processor associated with the SDV to strike the object in a manner that causes one or more crumple zones of the SDV to maximize absorption of the collision with the object.

9. The computer-implemented method of claim 1, wherein the evasive maneuver comprises causing at least one tire on the vehicle to deflate.

10. The computer-implemented method of claim 5, wherein the ameliorative action prevents one or more airbags within the SDV from deploying in response to the SDV colliding with the object.

11. The computer-implemented method of claim 5, further comprising:

determining, by the computer, a weight ratio between the SDV and the object; and

adjusting, by the computer, the ameliorative action according to the weight ratio between the SDV and the object.

12. The computer-implemented method of claim 1, further comprising:

receiving, by the one or more computers associated with the one or more vehicles, computer executable instructions from the vehicle for implementing an evasive action similar to the evasive maneuver performed by the vehicle; and

directing, by at least one of the one or more computers, the one or more vehicles to execute the computer executable instructions for implementing the evasive action, in response to detecting the roadway condition.

13. The computer-implemented method of claim 1, wherein the computer-implemented method is implemented as a cloud-based service.

14. A computer program product for training one or more computers associated with one or more vehicles to perform an evasive maneuver, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions readable and executable by a processor to cause the processor to:

detect, by one or more sensors associated with a vehicle, a roadway condition of a first roadway;

detect, by the one or more sensors associated with the vehicle, an evasive maneuver performed in response to detecting the roadway condition;

determine whether the evasive maneuver was successful in avoiding the roadway condition;

store a record of the evasive maneuver and the roadway condition, in response to said determining whether the evasive maneuver was a successful maneuver for avoiding the roadway condition;

train one or more computers associated with one or more vehicles, to execute the evasive maneuver, in response to determining that the one or more vehicles is exposed to the roadway condition experienced by the vehicle.

15. The computer program product of claim 14, wherein the program instructions further cause the processor to:

in response to determining that the evasive maneuver was an unsuccessful maneuver, block transmission of the unsuccessful maneuver to the one or more vehicles.

16. The computer program product of claim 14, wherein the program instructions are provided as a service in a cloud environment.

17. A system comprising:

one or more processors;

one or more computer readable memories operably coupled to the one or more processors;

one or more computer readable storage mediums operably coupled to the one or more computer readable memories; and

program instructions stored on at least one of the one or more computer readable storage mediums for execution by at least one of the one or more processors via at least one of the one or more computer readable memories, the program instructions comprising;

program instructions to detect, by one or more sensors associated with a vehicle, a roadway condition of a first roadway;

program instructions to detect, by the one or more sensors associated with the vehicle, an evasive maneuver performed in response to detecting the roadway condition;

program instructions to determine whether the evasive maneuver was successful in avoiding the roadway condition;

program instructions to store a record of the evasive maneuver and the roadway condition, in response to said determining whether the evasive maneuver was a successful maneuver for avoiding the roadway condition;

program instructions to train one or more computers associated with one or more vehicles, to execute the evasive maneuver, in response to determining that the one or more vehicles is exposed to the roadway condition experienced by the vehicle.

18. The system of claim 17, wherein the program instructions further comprise:

program instructions to, in response to determining that the evasive maneuver was an unsuccessful maneuver, block transmission of the unsuccessful maneuver to the one or more vehicles.

19. The system of claim 17, wherein the vehicle is a self-driving vehicle (SDV), and wherein the program instructions further comprise:

program instructions to detect that an imminent collision by the SDV is imminent with a confidence C1;

program instructions to determine whether the SDV has one or more occupants with a confidence C2;

program instructions to identify an object with which the imminent collision by the SDV is imminent with a confidence C3;

program instructions to generate an ameliorative action for the SDV based on C1, C2, C3; and

program instructions to direct the SDV to execute the ameliorative action.

20. The system of claim 17, wherein the program instructions are provided as a service in a cloud environment.