Systems and methods for autonomous vehicle validation
An example method includes obtaining log data descriptive of an exemplar action of an exemplar vehicle in an environment, the exemplar action occurring in an initial state of the environment; determining, using the operational system, a planned action for a simulated vehicle in the initial state of the environment; simulating an SUT state of the environment resulting from the simulated vehicle executing the planned action in the initial state of the environment and an actor performing an actor action subsequent to the planned action and an exemplar state of the environment resulting from the simulated vehicle executing the exemplar action in the initial state of the environment and the actor performing the actor action subsequent to the exemplar action; determining a test score based on the SUT state and a reference score based on the exemplar state; evaluating the operational system based on the test score and the reference score.
1. A method for validating an operational system under test (SUT) for operating an autonomous vehicle, the method comprising:
(a) obtaining log data descriptive of an exemplar action of an exemplar vehicle in an environment, the exemplar action occurring in an initial state of the environment;
(b) determining, using the operational system, a planned action for a simulated vehicle in the initial state of the environment;
(c) simulating:
(i) a response of the simulated vehicle to an actor performing an actor action subsequent to the simulated vehicle executing the planned action in the initial state of the environment to obtain an SUT state of the environment; and
(ii) a response of the simulated vehicle to the actor performing the actor action subsequent to the simulated vehicle executing the exemplar action in the initial state of the environment to obtain an exemplar state of the environment;
(d) determining a test score based on the SUT state of the environment and a reference score based on the exemplar state of the environment; and
(e) evaluating the operational system based on the test score and the reference score.
2. The method of claim 1 , wherein the actor action comprises a hazard affecting the simulated vehicle.
3. The method of claim 2 , wherein the exemplar action is a guarding driving behavior to guard against occurrence of the hazard.
4. The method of claim 1 , wherein (c) comprises simulating (i) and (ii) using a model that approximates motion of the simulated vehicle.
5. The method of claim 1 , wherein (c) comprises injecting the actor action into the initial state of the environment.
6. The method of claim 5 , wherein:
the actor action is a synthesized actor action based on a set of possible actions that could be taken by the actor in the initial state of the environment.
7. The method of claim 5 , wherein:
the actor action is based on a recorded actor action obtained from other log data.
8. The method of claim 1 , wherein the operational system comprises at least one model of an autonomous vehicle control system configured to receive sensor data and control movement of the autonomous vehicle.
9. The method of claim 1 , wherein the test score is based on a state of the simulated vehicle in the SUT state of the environment and the reference score is based on a state of the simulated vehicle in the exemplar state of the environment.
10. The method of claim 9 , wherein (d) comprises:
evaluating the state of the simulated vehicle in the SUT state of the environment and the state of the simulated vehicle in the exemplar state of the environment with a cost function.
11. The method of claim 10 , wherein the cost function is determined based on a distance to an actor boundary.
12. The method of claim 10 , wherein the cost function is determined based on a severity measure of an intersection with an actor boundary.
13. The method of claim 1 , wherein (e) comprises categorizing, based on one or more features of the initial state of the environment, the test score and the reference score.
14. The method of claim 13 , comprising:
(f) determining one or more categories for improvement associated with suboptimal test scores; and
(g) adjusting one or more parameters of the operational system corresponding to operation of the autonomous vehicle in the one or more categories for improvement.
15. One or more non-transitory computer-readable media storing instructions that are executable to cause one or more processors to perform operations for validating an operational system under test (SUT) for operating an autonomous vehicle, the operations comprising:
(a) obtaining log data descriptive of an exemplar action of an exemplar vehicle in an environment, the exemplar action occurring in an initial state of the environment;
(b) determining, using the operational system, a planned action for a simulated vehicle in the initial state of the environment;
(c) simulating:
(i) a response of the simulated vehicle to an actor performing an actor action subsequent to the simulated vehicle executing the planned action in the initial state of the environment to obtain an SUT state of the environment; and
(ii) a response of the simulated vehicle to the actor performing the actor action subsequent to the simulated vehicle executing the exemplar action in the initial state of the environment to obtain an exemplar state of the environment;
(d) determining a test score based on the SUT state of the environment and a reference score based on the exemplar state of the environment; and
(e) evaluating the operational system based on the test score and the reference score.
16. The one or more non-transitory computer-readable media of claim 15 , wherein the actor action comprises a hazard affecting the simulated vehicle.
17. The one or more non-transitory computer-readable media of claim 16 , wherein (c) comprises determining a reaction to the hazard affecting the simulated vehicle.
18. The one or more non-transitory computer-readable media of claim 16 , wherein the exemplar action is a guarding driving behavior to guard against occurrence of the hazard.
19. The one or more non-transitory computer-readable media of claim 15 , wherein the actor action is selected from a plurality of actions having a probability below a threshold probability.
20. An autonomous vehicle control system for controlling an autonomous vehicle, the autonomous vehicle control system comprising:
one or more processors; and
one or more non-transitory computer-readable media storing instructions that are executable by the one or more processors to cause the autonomous vehicle control system to control a motion of the autonomous vehicle using an operational system;
wherein the operational system was validated by:
(a) obtaining log data descriptive of an exemplar action of an exemplar vehicle in an environment, the exemplar action occurring in an initial state of the environment;
(b) determining, using the operational system, a planned action for a simulated vehicle in the initial state of the environment;
(c) simulating:
(i) a response of the simulated vehicle to an actor performing an actor action subsequent to the simulated vehicle executing the planned action in the initial state of the environment to obtain an SUT state of the environment; and
(ii) a response of the simulated vehicle to the actor performing the actor action subsequent to the simulated vehicle executing the exemplar action in the initial state of the environment to obtain an exemplar state of the environment;
(d) determining a test score based on the SUT state of the environment and a reference score based on the exemplar state of the environment; and
(e) evaluating the operational system based on the test score and the reference score.