Behavior planning for autonomous vehicles in yield scenarios

1. A method comprising:

identifying a scenario for a first vehicle based at least on analyzing sensor data generated using at least one sensor of the first vehicle in an environment;

determining based at least on the scenario, a first path for the first vehicle to traverse an area in the environment, a second path for a second vehicle to traverse the area, and a wait state corresponding to a state machine including at least a first state and a second state that correspond to respective categories of yield behavior for controlling the first vehicle with respect to one or more contention points between the first path and the second path;

determining, from the first path, a first trajectory that satisfies the respective categories of yield behavior represented by the wait state in the area based at least on a first location of the first vehicle at a time;

determining, from the second path, a second trajectory in the area for the second vehicle based at least on a second location of the second vehicle at the time;

evaluating whether there is a conflict between the first trajectory and the second trajectory;

selecting the wait state for the first vehicle based at least on the evaluating; and

based at least on the selecting of the wait state, controlling the first vehicle to traverse the area using the wait state.

2. The method of claim 1 , furthering comprising:

determining, from the first path, a third trajectory that satisfies at least one category of yield behavior represented by a second wait state in the area based at least on the first location of the first vehicle at the time,

wherein the selecting of the wait state is further based at least on determining a conflict exists between the third trajectory and one or more trajectories corresponding to the second path.

3. The method of claim 1 , wherein the selecting of the wait state is from a plurality of wait states.

4. The method of claim 1 , wherein the controlling includes transitioning between the first state and the second state of the state machine based at least on evaluating one or more transition conditions associated with the wait state.

5. The method of claim 1 , the wait state further corresponds to at least one category of yield behavior for controlling the second vehicle with respect to the one or more contention points between the first path and the second path, and the second trajectory is determined to satisfy the at least one category of yield behavior represented by the wait state in the area.

6. The method of claim 1 , wherein the scenario includes one or more of a crossing scenario with the area including an intersection or a merging scenario with the area including merging lanes.

7. The method of claim 1 , wherein the identifying the scenario includes:

localizing the first vehicle in a map of the environment that identifies a location of the scenario, wherein the map encodes a set of traffic rules for the location;

detecting dynamic traffic signal states of the environment; and

selecting, for the evaluating, at least the wait state as a subset of a plurality of wait states based at least on the wait state corresponding to the set of traffic rules for the location and the dynamic traffic signal states.

8. A processor comprising:

one or more circuits to:

identify a scenario for a first vehicle based at least on analyzing sensor data generated using at least one sensor of the first vehicle in an environment;

determine based at least on the scenario, a first path for the first vehicle to traverse an area in the environment, a second path for a second vehicle to traverse the area, and a wait state corresponding to a state machine including at least a first state and a second state that correspond to respective categories of yield behavior for controlling the first vehicle with respect to one or more contention points between the first path and the second path;

employ the first path to determine a first trajectory that satisfies the respective categories of yield behavior represented by the wait state in the area based at least on a first location of the first vehicle at a time;

employ the second path to determine a second trajectory in the area for the second vehicle based at least on a second location of the second vehicle at the time;

evaluate whether there is a conflict between the first trajectory and the second trajectory;

select the wait state for the first vehicle based at least on the evaluating; and

based at least on the selecting of the wait state, control the first vehicle to traverse the area using the wait state.

9. The processor of claim 8 , wherein the one or more circuits are further to:

determine, from the first path, a third trajectory that satisfies at least one category of yield behavior represented by a second wait state in the area based at least on the first location of the first vehicle at the time,

wherein the selecting of the wait state is further based at least on determining a conflict exists between the third trajectory and one or more trajectories corresponding to the second path.

10. The processor of claim 8 , wherein selecting of the wait state is from a plurality of wait states.

11. The processor claim 8 , wherein the one or more circuits are further to:

determine a first time interval that indicates a first set of temporal coordinates characterizing the first trajectory traversing the area;

determine a second time interval that indicates a second set of temporal coordinates characterizing the second trajectory traversing the area;

determine an intersection of the first set of temporal coordinates and the second set of temporal coordinates; and

determine the conflict between the first trajectory and the second trajectory based on detecting that the intersection of the first set of temporal coordinates and the second set of temporal coordinates is a non-null set.

12. The processor of claim 8 , wherein the first trajectory and the second trajectory are each trajectories within a four-dimensional manifold with at least two spatial dimensions and at least two temporal dimensions.

13. The processor of claim 8 , wherein the scenario includes one or more of a crossing scenario with the area including an intersection or a merging scenario with the area including merging lanes.

14. The processor of claim 8 , wherein the second trajectory simulates the second vehicle accelerating after the time.

15. A system comprising:

one or more processing units to execute actions comprising:

identifying a scenario for a first vehicle based at least on analyzing sensor data generated using at least one sensor of the first vehicle in an environment;

determining based at least on the scenario, a first path for the first vehicle to traverse an area in the environment, a second path for a second vehicle to traverse the area, and a wait state corresponding to a state machine including at least a first state and a second state that correspond to respective categories of yield behavior for controlling the first vehicle with respect to one or more contention points between the first path and the second path;

determining, from the first path, a first trajectory that satisfies the respective categories of yield behavior represented by the wait state in the area based at least on a first location of the first vehicle at a time;

determining, from the second path, a second trajectory in the area for the second vehicle based at least on a second location of the second vehicle at the time;

evaluating whether there is a conflict between the first trajectory and the second trajectory;

selecting the wait state for the first vehicle based at least on the evaluating; and

based at least on the selecting of the wait state, controlling the first vehicle to traverse the area using the wait state.

16. The system of claim 15 , wherein the actions further include:

determining, from the first path, a third trajectory that satisfies at least one category of yield behavior represented by a second wait state in the area based at least on the first location of the first vehicle at the time,

wherein the selecting of the wait state is further based at least on determining a conflict exists between the third trajectory and one or more trajectories corresponding to the second path.

17. The system of claim 15 , wherein the selecting of the wait state is from a plurality of wait states.

18. The system of claim 15 , the actions further comprising:

determining a first time interval that indicates a first set of temporal coordinates characterizing the first trajectory traversing the area;

determining a second time interval that indicates a second set of temporal coordinates characterizing the second trajectory traversing the area;

determining an intersection of the first set of temporal coordinates and the second set of temporal coordinates; and

identifying the conflict between the first trajectory and the second trajectory based on detecting that the intersection of the first set of temporal coordinates and the second set of temporal coordinates is a non-null set.

19. The system of claim 15 , wherein the first trajectory and the second trajectory are each trajectories within a four-dimensional manifold with at least two spatial dimensions and at least two temporal dimensions.

20. The system of claim 15 , comprised in at least one of:

a control computing system for an autonomous or semi-autonomous machine;

a perception computing system for an autonomous or semi-autonomous machine;

a system for performing simulation operations;

a system for performing deep learning operations;

a system implemented using an edge computing device;

a system implemented using a robot;

a system incorporating one or more virtual machines (VMs);

a system implemented at least partially in a data center; or

a system implemented at least partially using cloud computing resources.