Automated system and method for modeling the behavior of vehicles and other agents

• US 9,495,874 B1
• Filed: 04/13/2012
• Issued: 11/15/2016
• Est. Priority Date: 04/13/2012
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Zhu, Jiajun, Ferguson, David I., Dolgov, Dmitri A.
• Primary Examiner(s): Beaulieu, Yonel
• Assistant Examiner(s): Mahne, Kevin P
• Application Number:
  US13/446,494
• Time in Patent Office: 1,677 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: B60W 2420/403 Image sensing, e.g. optical camera
• CPC Class Codes: B60W 2420/408 Radar; Laser, e.g. lidar
• CPC Class Codes: B60W 2554/00 Input parameters relating to objects
• CPC Class Codes: B60W 2554/20 Static objects
• CPC Class Codes: B60W 2554/4023 large-size vehicles, e.g. trucks
• CPC Class Codes: B60W 2554/4029 Pedestrians
• CPC Class Codes: B60W 2554/80 Spatial relation or speed relative to objects
• CPC Class Codes: B60W 2555/20 Ambient conditions, e.g. wind or rain
• CPC Class Codes: B60W 2556/10 Historical data
• CPC Class Codes: B60W 30/08 Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
• CPC Class Codes: B60W 30/09 Taking automatic action to avoid collision, e.g. braking and steering
• CPC Class Codes: B60W 30/095 Predicting travel path or likelihood of collision
• CPC Class Codes: B60W 30/0956 the prediction being responsive to traffic or environmental parameters
• CPC Class Codes: B60W 50/0097 Predicting future conditions
• CPC Class Codes: B60W 60/00133 for resting
• CPC Class Codes: B60W 60/00274 considering possible movement changes
• CPC Class Codes: B60W 60/00276 for two or more other traffic participants
• CPC Class Codes: G01R 19/25 using digital measurement techniques
• CPC Class Codes: G05D 1/0088 characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours using knowledge based models G06N5/00
• CPC Class Codes: G05D 1/0212 with means for defining a desired trajectory involving a plurality of land vehicles G05D1/0287
• CPC Class Codes: G05D 1/0214 : in accordance with safety or protection criteria, e.g. avoiding hazardous areas monitoring the location of vehicles within a certain area, e.g. forbidden or allowed areas, in traffic control systems for road vehicles G08G1/13
• CPC Class Codes: G05D 1/0289 : with means for avoiding collisions between vehicles vehicle fittings for automatically controlling speed including means for detecting potential obstacles B60K31/0008; avoiding obstacles by action on the steering system B62D; radar, sonar, lidar systems designed for anti-collision purposes G01S13/93, G01S15/93, G01S17/93
• CPC Class Codes: G06N 20/00 : Machine learning
• CPC Class Codes: G06N 3/00 : Computing arrangements based on biological models
• CPC Class Codes: G06V 10/255 : Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
• CPC Class Codes: G06V 20/56 : exterior to a vehicle by using sensors mounted on the vehicle
• CPC Class Codes: G06V 20/58 : Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
• CPC Class Codes: G06V 20/582 : of traffic signs
• CPC Class Codes: G06V 20/584 : of vehicle lights or traffic lights
• CPC Class Codes: G06V 20/588 : Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
• CPC Class Codes: G08G 1/09623 : Systems involving the acquisition of information from passive traffic signs by means mounted on the vehicle G08G1/0967 takes precedence
• CPC Class Codes: G08G 1/09626 : where the origin of the information is within the own vehicle, e.g. a local storage device, digital map
• CPC Class Codes: G08G 1/096725 : where the received information generates an automatic action on the vehicle control
• CPC Class Codes: G08G 1/096741 : where the source of the transmitted information selects which information to transmit to each vehicle
• CPC Class Codes: G08G 1/096775 : where the origin of the information is a central station
• CPC Class Codes: G08G 1/164 : Centralised systems, e.g. external to vehicles
• CPC Class Codes: G08G 1/165 : for passive traffic, e.g. including static obstacles, trees
• CPC Class Codes: G08G 1/166 : for active traffic, e.g. moving vehicles, pedestrians, bikes

Abstract

A method and apparatus are provided for determining one or more behavior models used by an autonomous vehicle to predict the behavior of detected objects. The autonomous vehicle may collect and record object behavior using one or more sensors. The autonomous vehicle may then communicate the recorded object behavior to a server operative to determine the behavior models. The server may determine the behavior models according to a given object classification, actions of interest performed by the object, and the object'"'"'s perceived surroundings.

Claims

1. A system comprising one or more computing devices, having one or more processors, the one or more computing devices being configured to:

store behavior model data for a plurality of object classifications;

receive object data from a monitoring source, wherein the object data comprises world-view data and actions data for a detected object;

determine the detected object'"'"'s classification based on the object data;

access a first behavior model from the stored behavior model data, wherein the first behavior model is accessed based on the detected object'"'"'s classification;

filter the actions data for the detected object to identify actions of interest;

update the first behavior model based on the world-view data and the identified actions of interest for the detected object, so as to generate an updated behavior model, wherein the updated behavior model is used for predicting how other objects, having a same classification as the detected object, will behave; and

use the first behavior model to autonomously control a vehicle.

2. The system of claim 1, wherein the actions data relates to the geographic position and movement of the detected object.

3. The system of claim 1, wherein the monitoring source is one or more autonomous vehicles.

4. The system of claim 1, wherein the detected object is one of an automobile, a pedestrian, or a bicycle.

5. The system of claim 1, wherein the first behavior model identifies a plurality of probabilities for potential actions of the detected object, and wherein the plurality of probabilities are based on observed actions taken by previously monitored objects having a classification that is the same as the detected object'"'"'s classification.

18. The system of claim 1, wherein the one or more computing devices are incorporated into the vehicle.

19. The system of claim 1, further comprising:

the vehicle; and

a server computing device,

wherein at least a first one of the one or more computing devices configured to use the first behavior model to autonomously control the vehicle is incorporated into the vehicle, and wherein at least a second one of the one or more computing devices configured to update the first behavior model is the server computing device.

6. A computer implemented method comprising:

receiving, by one or more processors, object data from a monitoring source, wherein the object data comprises classification data for a detected object and world-view data for the object;

accessing, by the one or more processors, a first behavior model corresponding to the classification data;

identifying an action of interest in the object data of the object, wherein data relating to the action of interest is only a subset of the object data;

updating, by the one or more processors, the first behavior model based on the identified action of interest and the world-view data so as to create an updated behavior model; and

autonomously controlling a vehicle, by the one or more processors, based on the updated behavior model predicting how another object, having a same classification as the detected object, will behave.

7. The method of claim 6, wherein the object data further comprises actions data indicating that the detected object has been positioned or has moved in a predefined manner relative to one or more road graph elements.

8. The method of claim 7, wherein the road graph elements comprise a first lane of traffic and a second lane of traffic, and wherein the actions data indicates that the detected object has changed from the first lane of traffic to the second lane of traffic.

9. The method of claim 6, wherein the monitoring source is an autonomous vehicle.

10. The method of claim 6, wherein the detected object is one of an automobile, a pedestrian, or a bicycle.

11. The method of claim 6, wherein:

the first behavior model is based on paths of travel for one or more previously monitored objects, and wherein the first behavior model identifies a plurality of probabilities for potential actions of the detected object.

12. The method of claim 11, wherein at least one of the plurality of probabilities are based on first world-view data for the previously monitored objects.

13. The method of claim 11, wherein updating the first behavior model further comprises adjusting the plurality of probabilities based on actions taken by the detected object.

14. The method of claim 6 wherein accessing the updated behavior model occurs in connection with receiving a request for the updated behavior model from one or more computing devices associated with one or more autonomous vehicles.

15. A method comprising:

controlling, by one or more computing devices having one or more processors, an autonomous vehicle in accordance with a control strategy;

detecting, by the one or more computing devices, a first object that is external to the autonomous vehicle using one or more sensors;

storing, by the one or more computing devices, object data for the first object;

determining, by the one or more computing devices, that the first object is of a first classification based on the object data;

selecting, by the one or more computing devices, a behavior model of the first object, wherein the selection is based on the first object being of the first classification;

predicting, by the one or more computing devices, that the first object will perform one or more actions, wherein the prediction is based on the selected behavior model and the world-view data;

altering, by the one or more computing devices, the control strategy based on the prediction that the first object will perform one or more actions;

providing, by the one or more computing devices, object data to be used in updating the behavior model, wherein the updated behavior model is used to predict actions of a second object having a same classification as the first object; and

updating, by the one or more computing devices, the selected behavior model based on an action of interest identified in the object data of the first object, wherein data relating to the action of interest is only a subset of the object data.

16. The method of claim 15, further comprising:

accessing, by the one or more computing devices, map data having a one or more road graph elements;

comparing, by the one or more computing devices, position and movement data for the first object with the road graph element;

determining, by the one or more computing devices, based on the comparison, that the first object has performed an action of interest; and

storing, by the one or more computing devices, actions-of-interest data relating to the determined action of interest.

17. The method of claim 15 wherein the object data indicates that the first object is at a particular geographic location, and wherein the behavior model is based on behavior data for a plurality of objects that have been located at or near the particular geographic location.