Methods and systems for agent prioritization

Patent metadata

Field	Value
Publication number	US-12050468-B2
Application number	US-202318160456-A
Country	US
Kind code	B2
Filing date	Jan 27, 2023
Priority date	Dec 17, 2021
Publication date	Jul 30, 2024
Grant date	Jul 30, 2024

Abstract

Provided are methods for agent prioritization, which can include determining a primary agent set and generating, based on the primary agent set, a trajectory for the autonomous vehicle. Some methods described also include determining an interaction parameter of agents in the environment. Systems and computer program products are also provided.

First claim

What is claimed is:

1. A method for operating an autonomous vehicle, the method comprising:
obtaining, by at least one processor, sensor data associated with an environment in which an autonomous vehicle is operating,
determining, by the at least one processor, based on the sensor data, an agent set comprising a plurality of agents located in the environment;
determining, by the at least one processor, for each agent of the agent set, an interaction parameter indicative of a prediction of interaction of the respective agent with the autonomous vehicle;
determining, by the at least one processor, based on the plurality of interaction parameters, a primary agent set, wherein the primary agent set comprises a plurality of primary agents, the plurality of primary agents corresponding to a subset of the plurality of agents of the agent set;
ordering, by the at least one processor, the plurality of primary agents of the primary agent set based on a size of one or more homotopies;
generating, by the at least one processor, based on the ordered plurality of primary agents of the primary agent set, a trajectory for the autonomous vehicle; and
providing, based on the trajectory, control data to cause operation of the autonomous vehicle.

2. The method of claim 1 , wherein determining, based on the plurality of interaction parameters, the primary agent set comprises filtering out one or more agents of the agent set based on the plurality of interaction parameters.

3. The method of claim 2 , wherein filtering out one or more agents comprises filtering out one or more agents based on a criterion applied to the plurality of interaction parameters.

4. The method of claim 1 , wherein determining, based on the plurality of interaction parameters, the primary agent set comprises:
applying, to the agent set, a prioritization scheme based on the plurality of interaction parameters.

5. The method of claim 1 , wherein generating, based on the primary agent set, the trajectory for the autonomous vehicle comprises applying a first model to an agent of the primary agent set.

6. The method of claim 5 , further comprising determining, by the at least one processor, based on the plurality of interaction parameters, a secondary agent set, wherein the primary agent set and the secondary agent set are mutually exclusive subsets of the agent set.

7. The method of claim 6 , wherein generating, based on the primary agent set, the trajectory for the autonomous vehicle comprises:
generating, based on the primary agent set and the secondary agent set, the trajectory for the autonomous vehicle.

8. The method of claim 5 , wherein generating, based on the primary agent set and the secondary agent set, a trajectory for the autonomous vehicle comprises:
applying a second model to one or more agents of the secondary agent set.

9. The method of claim 8 , wherein the first model has a higher fidelity than the second model.

10. The method of claim 8 , wherein the first model or second model comprises one or more of: an agent recognition scheme, an agent prediction scheme, an agent projection scheme, a trajectory extraction scheme and a trajectory evaluation scheme.

11. The method of claim 1 , wherein determining, by the at least one processor, for each agent of the agent set, an interaction parameter indicative of a prediction of interaction of the respective agent with the autonomous vehicle comprises:
predicting an interaction of the respective agent with the autonomous vehicle; and
determining the interaction parameter based on the prediction.

12. The method of claim 11 , wherein predicting the interaction of the respective agent with the autonomous vehicle comprises:
determining one or more homotopy parameters indicative of a constraint applied by the respective agent on a trajectory of the autonomous vehicle; and
predicting the interaction based on the one or more homotopy parameters.

13. The method of claim 12 , wherein predicting the interaction based on the one or more homotopy parameters comprises predicting the interaction by inputting the one or more homotopy parameters or sensor data into a neural network model.

14. The method of claim 12 , wherein the one or more homotopy parameters are indicative of at least one of: a speed of the respective agent, an acceleration of the respective agent, or a location of the respective agent.

15. The method of claim 12 , wherein the method comprises clustering, based on the one or more homotopy parameters, agents of the plurality of agents.

16. The method of claim 1 , wherein an agent of the plurality of agents comprises an object capable of a dynamic movement over time.

17. The method of claim 1 , wherein the method comprises: operating, based on the trajectory, the autonomous vehicle.

18. A non-transitory computer readable medium comprising computer-executable instructions stored thereon that, when executed by at least one processor, cause the at least one processor to carry out operations comprising:
obtaining, by at least one processor, sensor data associated with an environment in which an autonomous vehicle is operating;
determining, by the at least one processor, based on the sensor data, an agent set comprising a plurality of agents located in the environment;
determining, by the at least one processor, for each agent of the agent set, an interaction parameter indicative of a prediction of interaction of the respective agent with the autonomous vehicle;
determining, by the at least one processor, based on the plurality of interaction parameters, a primary agent set, wherein the primary agent set comprises a plurality of primary agents, the plurality of primary agents corresponding to i-s-a subset of the plurality of agents of the agent set;
ordering, by the at least one processor, the plurality of primary agents of the primary agent set based on a size of one or more homotopies,
generating, by the at least one processor, based on the ordered plurality of primary agents of the primary agent set, a trajectory for the autonomous vehicle; and
providing, based on the trajectory, control data to cause operation of the autonomous vehicle.

19. A system, comprising at least one processor; and at least one memory storing computer-executable instructions thereon that, when executed by the at least one processor, cause the at least one processor to:
obtain sensor data associated with an environment in which an autonomous vehicle is operating;
determine, based on the sensor data, an agent set comprising a plurality of agents located in the environment;
determine, for each agent of the agent set, an interaction parameter indicative of a prediction of interaction of the respective agent with the autonomous vehicle;
determine, based on the plurality of interaction parameters, a primary agent set, wherein the primary agent set comprises a plurality of primary agents, the plurality of primary agents corresponding to a subset of the plurality of agents of the agent set;
order the plurality of primary agents of the primary agent set based on a size of one or more homotopies;
generate, based on the ordered plurality of primary agents of the primary agent set, a trajectory for the autonomous vehicle, and
provide, based on the trajectory, control data to cause operation of the autonomous vehicle.

20. The system of claim 19 , wherein to determine, based on the plurality of interaction parameters, the primary agent set, the computer-executable instructions cause the at least one processor to filte

Assignees

• Motional Ad Llc

Inventors

• Cserna Bence
• Kabzan Juraj
• Gall Kevin C
• Jespersen Thomas Kølbæk
• Bernasconi Gianmarco Alessandro
• Cui Henggang

Classifications

• G06N3/08

  Learning methods · CPC title

• G05D1/222

  operated by humans · CPC title

• G05D1/247

  using signals provided by artificial sources external to the vehicle, e.g. navigation beacons · CPC title

• G05D1/227

  Handing over between remote control and on-board control; Handing over between remote control arrangements · CPC title

• G05D1/0011

  associated with a remote control arrangement · CPC title