Risk-aware executor with action set recommendations
A first distinct vehicle operational scenario is identified for an autonomous vehicle (AV). A first set of candidate vehicle control actions are received from a model that provides a first solution to the first distinct vehicle operational scenario. An action is selected from the first set of candidate vehicle control actions. The AV is controlled based on the action. The first solution is obtained offline in a first idealized situation that is decoupled from a current context of the AV.
1. A method for use in traversing a vehicle transportation network by an autonomous vehicle (AV), the method comprising:
identifying a first distinct vehicle operational scenario for the AV;
receiving, from a model that provides a first solution to the first distinct vehicle operational scenario, a first set of candidate vehicle control actions,
wherein the first solution is obtained offline in a first idealized situation that is decoupled from a current context of the AV;
selecting an action from the first set of candidate vehicle control actions; and controlling the AV based on the action.
2. The method of claim 1 , further comprising:
receiving, from a second model that provides a second solution to a second distinct vehicle operational scenario, a second set of candidate vehicle control actions,
wherein the second solution is obtained offline in a second idealized situation that is decoupled from the current context of the AV; and
determining a third set of actions as an intersection between the first set of candidate vehicle control actions and the second set of candidate vehicle control actions.
3. The method of claim 2 , wherein the second model is selected in response to receiving sensor information corresponding to an external object.
4. The method of claim 2 , wherein determining the third set of actions as the intersection between the first set of candidate vehicle control actions and the second set of candidate vehicle control actions comprises:
on a condition that the intersection is empty, adding a default action to the third set of actions.
5. The method of claim 4 , wherein controlling the AV based on the action comprises:
on a condition that the action is to stop, controlling the AV to be stationary;
on a condition that the action is to advance, controlling the AV to traverse a defined cautionary distance in the vehicle transportation network at a defined cautionary rate; and
on a condition that the action is to proceed, controlling the AV to traverse the vehicle transportation network in accordance with a previously identified vehicle control action.
6. The method of claim 1 , wherein the model uses state information of the AV.
7. The method of claim 6 , wherein the state information comprises at least one of a first state of a sensor of the AV, a second state of an actuator of the AV, a third state of availability of a tele-operator, or a fourth state relating to an autonomy competence level of the AV.
8. An autonomous vehicle, comprising:
a processor configured to execute instructions stored on a non-transitory computer readable medium to:
identify a first distinct vehicle operational scenario for the autonomous vehicle;
receive, from a model that provides a first solution to the first distinct vehicle operational scenario, a first set of candidate vehicle control actions,
wherein the first solution is obtained offline in a first idealized situation that is decoupled from a current context of the autonomous vehicle;
select an action from the first set of candidate vehicle control actions; and
control the autonomous vehicle based on the action.
9. The autonomous vehicle of claim 8 , wherein the processor is further configured to execute instructions stored in the non-transitory computer readable medium to:
receive, from a second model that provides a second solution to a second distinct vehicle operational scenario, a second set of candidate vehicle control actions,
wherein the second solution is obtained offline in a second idealized situation that is decoupled from the current context of the autonomous vehicle; and
determine a third set of actions as an intersection between the first set of candidate vehicle control actions and the second set of candidate vehicle control actions.
10. The autonomous vehicle of claim 9 , wherein the second model is selected in response to receiving sensor information corresponding to an external object.
11. The autonomous vehicle of claim 9 , wherein to determine the third set of actions as the intersection between the first set of candidate vehicle control actions and the second set of candidate vehicle control actions comprises to:
on a condition that the intersection is empty, add a default action to the third set of actions.
12. The autonomous vehicle of claim 11 , wherein controlling the autonomous vehicle based on the action comprises:
on a condition that the action is to stop, controlling the autonomous vehicle to be stationary;
on a condition that the action is to advance, controlling the autonomous vehicle to traverse a defined cautionary distance at a defined cautionary rate; and
on a condition that the action is to proceed, controlling the autonomous vehicle to proceed in accordance with a previously identified vehicle control action.
13. The autonomous vehicle of claim 8 , wherein the model uses state information of the autonomous vehicle.
14. The autonomous vehicle of claim 13 , wherein the state information comprises at least one of a first state of a sensor of the autonomous vehicle, a second state of an actuator of the autonomous vehicle, a third state of availability of a tele-operator, or a fourth state relating to an autonomy competence level of the autonomous vehicle.
15. A non-transitory computer readable medium storing instructions operable to cause one or more processors to perform operations, comprising:
identifying a first distinct vehicle operational scenario for an autonomous vehicle (AV);
receiving, from a model that provides a first solution to the first distinct vehicle operational scenario, a first set of candidate vehicle control actions,
wherein the first solution is obtained offline in a first idealized situation that is decoupled from a current context of the AV;
selecting an action from the first set of candidate vehicle control actions; and
controlling the AV based on the action.
16. The non-transitory computer readable medium of claim 15 , further comprising:
receiving, from a second model that provides a second solution to a second distinct vehicle operational scenario, a second set of candidate vehicle control actions,
wherein the second solution is obtained offline in a second idealized situation that is decoupled from the current context of the AV; and
determining a third set of actions as an intersection between the first set of candidate vehicle control actions and the second set of candidate vehicle control actions.
17. The non-transitory computer readable medium of claim 16 , wherein the second model is selected in response to receiving sensor information corresponding to an external object.
18. The non-transitory computer readable medium of claim 16 , wherein determining the third set of actions as the intersection between the first set of candidate vehicle control actions and the second set of candidate vehicle control actions comprises:
on a condition that the intersection is empty, adding a default action to the third set of actions.
19. The non-transitory computer readable medium of claim 18 , wherein controlling the AV based on the action comprises:
on a condition that the action is to stop, controlling the AV to be stationary;
on a condition that the action is to advance, controlling the AV to traverse a defined cautionary distance at a defined cautionary rate; and
on a condition that the action is to proceed, controlling the AV to proceed in accordance with a previously identified vehicle control action.
20. The non-transitory computer readable medium of claim 15 , wherein the model uses state information of the AV, wherein the state information comprises at least one of a first state of a sensor of the AV, a second state of an actuator of the AV, a third state of availability of a tele-operator, or a fourth state relating to an autonomy competence level of the AV.