Systems and methods for configuring autonomous vehicle operation

Patent metadata

Field	Value
Publication number	US-12371065-B2
Application number	US-202017073657-A
Country	US
Kind code	B2
Filing date	Oct 19, 2020
Priority date	Oct 19, 2020
Publication date	Jul 29, 2025
Grant date	Jul 29, 2025

Abstract

Systems, methods, and non-transitory computer-readable media can detect an occurrence of a condition in an environment based on sensor data captured by a vehicle. A determination is made whether the occurrence of the condition satisfies a threshold associated with a likelihood that a behavior associated with an object in the environment will occur based on an interaction between the condition and the object, wherein the likelihood is based on prior observations of one or more objects. Subsequent to determining that the threshold is satisfied, a vehicle operation that is associated with the likelihood that the behavior associated with the object will occur is performed.

First claim

What is claimed is:

1. A computer-implemented method comprising:
detecting, by a computing system, an occurrence of a condition associated with a first object in an environment based on sensor data captured by a vehicle;
determining, by the computing system, whether the occurrence of the condition satisfies a threshold associated with a likelihood that a behavior associated with a second object in the environment will occur in response to the occurrence of the condition associated with the first object, wherein the likelihood is based on prior observations of one or more objects and an interaction between the condition and the second object;
subsequent to determining that the threshold is satisfied, reconfiguring, by the computing system, a perception component of an autonomy stack associated with the vehicle, wherein the reconfiguring comprises adjusting at least one of parameters, detection range, or focus areas of the perception component for observing the occurrence of the behavior associated with the second object; and
performing, by the computing system, a vehicle operation to guide the vehicle to avoid collision with the second object, wherein the vehicle operation is associated with the likelihood that the behavior associated with the second object will occur and data collected by the reconfigured perception component.

2. The computer-implemented method of claim 1 , wherein determining whether the occurrence of the condition satisfies the threshold comprises:
referencing, by the computing system, a semantic map associated with the environment, wherein the semantic map includes at least a priors layer that encodes information describing the condition and the likelihood that the behavior associated with the second object will occur, and
the semantic map is continuously updated based on new data collected by a fleet of vehicles.

3. The computer-implemented method of claim 1 , wherein the condition corresponds to an occurrence of an event associated with the first object in the environment, and wherein the occurrence of the at least one object or event within the environment is associated with the likelihood that the behavior associated with the second object will occur.

4. The computer-implemented method of claim 1 , wherein reconfiguring the perception component causes the perception component to at least one of increase or decrease a range of perception for sensors associated with the vehicle, expand perception processing to focus on an area of interest, change a perception model implemented by the perception component, change model parameters for a perception model implemented by the perception component, change an object classification model implemented by the perception component, activate or deactivate one or more sensors, or reallocate on-board resources associated with the perception component.

5. The computer-implemented method of claim 1 , wherein performing the vehicle operation comprises:
reconfiguring, by the computing system, a prediction component of the autonomy stack associated with the vehicle, wherein reconfiguring the prediction component changes at least one operation performed by the prediction component with respect to the second object in the environment based on the likelihood that the behavior associated with the second object will occur.

6. The computer-implemented method of claim 5 , wherein reconfiguring the prediction component causes the prediction component to apply a specialized prediction model to predict a location and movement of the second object instead of a generalized prediction model.

7. The computer-implemented method of claim 1 , wherein performing the vehicle operation comprises:
reconfiguring, by the computing system, a planning component of the autonomy stack associated with the vehicle, wherein reconfiguring the planning component changes at least one trajectory to be performed by the planning component based on the likelihood that the behavior associated with the second object will occur.

8. The computer-implemented method of claim 1 , further comprising:
determining, by the computing system, a plurality of probabilities that the behavior associated with the second object in the environment will occur based on an interaction between the second object and a plurality of conditions, including the condition associated with the first object; and
determining, by the computing system, the likelihood that the behavior associated with the second object will occur based on an interaction between the condition and the second object based on a highest probability from the plurality of probabilities.

9. The computer-implemented method of claim 1 , wherein the prior observations of the one or more objects are captured by sensors of one or more vehicles that navigated the environment, and wherein the second object is at least similar to the one or more objects, and a prior observation is associated with an interaction between the one or more objects and at least the condition.

10. A system comprising:
at least one processor; and
a memory storing instructions that, when executed by the at least one processor, cause the system to perform operations comprising:
detecting an occurrence of a condition associated with a first object in an environment based on sensor data captured by a vehicle;
determining whether the occurrence of the condition satisfies a threshold associated with a likelihood that a behavior associated with a second object in the environment will occur in response to the occurrence of the condition associated with the first object, wherein the likelihood is based on prior observations of one or more objects and an interaction between the condition and the second object;
subsequent to determining that the threshold is satisfied, reconfiguring a perception component of an autonomy stack associated with the vehicle, wherein the reconfiguring comprises adjusting at least one of parameters, detection range, or focus areas of the perception component for observing the occurrence of the behavior associated with the second object; and
performing a vehicle operation to guide the vehicle to avoid collision with the second object, wherein the vehicle operation is associated with the likelihood that the behavior associated with the second object will occur and data collected by the reconfigured perception component.

11. The system of claim 10 , wherein determining whether the occurrence of the condition satisfies the threshold comprises:
referencing a semantic map associated with the environment, wherein the semantic map includes at least a priors layer that encodes information describing the condition and the likelihood that the behavior associated with the second object will occur, and
the semantic map is continuously updated based on new data collected by a fleet of vehicles.

12. The system of claim 10 , wherein the condition corresponds to an occurrence of an event associated with the first object in the environment, and wherein the occurrence of the event is associated with the likelihood that the behavior associated with the second object will occur.

13. The system of claim 11 , wherein reconfiguring the perception component causes the perception component to at least one of increase or decrease a range of perception for sensors associated with the vehicle, expand perception processing to focus on an area of interest, change a perception model implemented by the perception component, change model parameters for a perception model implemented by the perception component, change an object classification model implemented by the perception component, activate or deactivate one or more sensors, or reallo

Assignees

• Lyft Inc

Inventors

• Agon Alan
• Ghadar Nastaran
• Jiang Yunjian
• Lee Mason
• De Magalhaes Massera Filho Carlos Alberto
• Stover Nachenberg Carey
• Omari Sammy
• Rufino Ferreira Ana Sofia
• Tao Meng

Classifications

• B60W2420/408

  Radar; Laser, e.g. lidar · CPC title

• B60W2420/403

  Image sensing, e.g. optical camera · CPC title

• G06V20/56

  exterior to a vehicle by using sensors mounted on the vehicle · CPC title

• B60W2554/4026

  Cycles · CPC title

• B60W2554/4029

  Pedestrians · CPC title