Autonomous vehicle operational management
Autonomous vehicle operational management may include an autonomous vehicle traversing a vehicle transportation network, which may include operating a scenario-specific operational control evaluation module instance, which is an instance of a scenario-specific operational control evaluation module instantiated for an occurrence of a distinct vehicle operational scenario by allocating computing resources to, and populating, the scenario-specific operational control evaluation module instance with operational environment information corresponding to the occurrence of the distinct vehicle operational scenario. Traversing the vehicle transportation network may include determining a current vehicle control action by receiving a candidate vehicle control action from the scenario-specific operational control evaluation module instance and identifying the candidate vehicle control action as the current vehicle control action, and traversing a portion of the vehicle transportation network in accordance with the current vehicle control action, wherein the portion of the vehicle transportation network includes the occurrence of the distinct vehicle operational scenario.
1 . A method for use in traversing a vehicle transportation network, the method comprising:
traversing, by an autonomous vehicle, a vehicle transportation network, wherein traversing the vehicle transportation network includes:
operating a scenario-specific operational control evaluation module instance, wherein the scenario-specific operational control evaluation module instance is an instance of a scenario-specific operational control evaluation module instantiated for an occurrence of a distinct vehicle operational scenario by allocating computing resources to the scenario-specific operational control evaluation module instance and populating the scenario-specific operational control evaluation module instance with operational environment information representing aspects of an operational environment of the autonomous vehicle corresponding to the occurrence of the distinct vehicle operational scenario, wherein the aspects include:
one or more aspects representing an external object within a defined distance from the autonomous vehicle; and
one or more aspects representing vehicle transportation network geometry for a portion of the vehicle transportation network that includes the occurrence of the distinct vehicle operational scenario;
determining a current vehicle control action, wherein determining the current vehicle control action includes:
receiving, as a candidate vehicle control action, a vehicle control action from the scenario-specific operational control evaluation module instance; and
identifying the candidate vehicle control action as the current vehicle control action, wherein the current vehicle control action is a stop vehicle control action, an ‘advance’ vehicle control action, an ‘accelerate’ vehicle control action, a ‘decelerate’ vehicle control action, a ‘maintain’ vehicle control action, or a ‘proceed’ vehicle control action; and
traversing the portion of the vehicle transportation network in accordance with the current vehicle control action.
2 . The method of claim 1 , wherein traversing the portion of the vehicle transportation network includes traversing the portion of the vehicle transportation network in accordance with an identified route, wherein the current vehicle control action is with respect to an expected path of the autonomous vehicle identified in accordance with the identified route.
3 . The method of claim 1 , wherein the aspects representing the external object indicate an expected path for the external object identified in accordance with a current trajectory for the external object.
4 . The method of claim 1 , wherein traversing the vehicle transportation network includes storing a plurality of scenario-specific operational control evaluation modules that includes the scenario-specific operational control evaluation module.
5 . The method of claim 4 , wherein a respective scenario-specific operational control evaluation module from the plurality of scenario-specific operational control evaluation modules encapsulates a respective model of a respective distinct vehicle operational scenario independent of an occurrence of the respective distinct vehicle operational scenario.
6 . The method of claim 1 , wherein the scenario-specific operational control evaluation module instance includes an instance of a model of the distinct vehicle operational scenario.
7 . The method of claim 6 , wherein the model of the distinct vehicle operational scenario is configured to model sensor uncertainty.
8 . The method of claim 6 , wherein populating the scenario-specific operational control evaluation module instance with operational environment information includes:
receiving sensor information from a sensor of the autonomous vehicle by an operational environment monitor operated by the autonomous vehicle;
generating the operational environment information by abstracting the sensor information by the operational environment monitor; and
representing the operational environment information in the instance of the model of the distinct vehicle operational scenario.
9 . The method of claim 8 , wherein populating the scenario-specific operational control evaluation module instance with operational environment information includes:
receiving vehicle transportation network information representing the vehicle transportation network; and
representing the operational environment information and the vehicle transportation network information in the instance of the model of the distinct vehicle operational scenario.
10 . The method of claim 6 , wherein populating the scenario-specific operational control evaluation module instance with operational environment information includes:
receiving a probability of availability for the portion of the vehicle transportation network from a blocking monitor instance; and
representing the probability of availability in the instance of the model of the distinct vehicle operational scenario.
11 . The method of claim 6 , wherein the model is a Partially Observable Markov Decision Process model.
12 . The method of claim 11 , wherein the Partially Observable Markov Decision Process model models the distinct vehicle operational scenario using a set of states, a set of actions, a set of state transition probabilities, a reward function, a set of observations, and a set of conditional observation probabilities.
13 . The method of claim 1 , wherein:
the scenario-specific operational control evaluation module instance is a first scenario-specific operational control evaluation module instance;
traversing the vehicle transportation network includes operating a second scenario-specific operational control evaluation module instance instantiated concurrently with operating the first scenario-specific operational control evaluation module instance; and
determining the current vehicle control action includes:
receiving, as a second candidate vehicle control action, a second vehicle control action from the second scenario-specific operational control evaluation module instance; and
identifying the candidate vehicle control action or the second candidate vehicle control action as the current vehicle control action.
14 . The method of claim 13 , wherein the second scenario-specific operational control evaluation module instance is a second instance of the scenario-specific operational control evaluation module.
15 . The method of claim 13 , wherein:
the scenario-specific operational control evaluation module is a first scenario-specific operational control evaluation module from a plurality of scenario-specific operational control evaluation modules; and
the second scenario-specific operational control evaluation module instance is an instance of a second scenario-specific operational control evaluation module from the plurality of scenario-specific operational control evaluation modules.
16 . The method of claim 13 , wherein:
the first scenario-specific operational control evaluation module instance omits data representing a second external object within the defined distance from the autonomous vehicle; and
the second scenario-specific operational control evaluation module instance includes data representing the second external object and omits data representing the first external object.
17 . The method of claim 13 , wherein:
the first scenario-specific operational control evaluation module instance omits data representing other external objects within the defined distance from the autonomous vehicle; and
the second scenario-specific operational control evaluation module instance includes data representing the first external object and omits data representing other external objects.
18 . An autonomous vehicle comprising:
a processor that executes instructions stored on a non-transitory computer readable medium to:
operate a scenario-specific operational control evaluation module instance, wherein the scenario-specific operational control evaluation module instance is an instance of a scenario-specific operational control evaluation module instantiated for an occurrence of a distinct vehicle operational scenario, wherein to instantiate the scenario-specific operational control evaluation module instance the processor executes the instructions to allocate computing resources to the scenario-specific operational control evaluation module instance and populate the scenario-specific operational control evaluation module instance with operational environment information representing aspects of an operational environment of the autonomous vehicle corresponding to the occurrence of the distinct vehicle operational scenario, wherein the aspects include:
one or more aspects representing an external object within a defined distance from the autonomous vehicle; and
one or more aspects representing vehicle transportation network geometry;
determine a current vehicle control action, wherein to determine the current vehicle control action the processor executes the instructions to:
receive, as a candidate vehicle control action, a vehicle control action from the scenario-specific operational control evaluation module instance; and
identify the candidate vehicle control action as the current vehicle control action, wherein the current vehicle control action is a stop vehicle control action, an ‘advance’ vehicle control action, an ‘accelerate’ vehicle control action, a ‘decelerate’ vehicle control action, a ‘maintain’ vehicle control action, or a ‘proceed’ vehicle control action; and
traverse a portion of a vehicle transportation network in accordance with the current vehicle control action, wherein the portion of the vehicle transportation network includes the occurrence of the distinct vehicle operational scenario.
19 . The autonomous vehicle of claim 18 , wherein the scenario-specific operational control evaluation module instance includes an instance of a model of the distinct vehicle operational scenario.
20 . The autonomous vehicle of claim 19 , wherein the model of the distinct vehicle operational scenario is configured to model sensor uncertainty.
21 . The autonomous vehicle of claim 19 , wherein, to populate the scenario-specific operational control evaluation module instance with operational environment information, the processor executes the instructions to:
receive sensor information from a sensor of the autonomous vehicle by an operational environment monitor operated by the autonomous vehicle;
receive vehicle transportation network information representing the vehicle transportation network;
generating the operational environment information by abstracting the sensor information by the operational environment monitor; and
represent the operational environment information and the vehicle transportation network information in the instance of the model of the distinct vehicle operational scenario.
22 . A non-transitory computer-readable storage medium, comprising executable instructions that are executed by a processor to perform:
controlling an autonomous vehicle to traverse a vehicle transportation network by:
operating a scenario-specific operational control evaluation module instance, wherein the scenario-specific operational control evaluation module instance is an instance of a scenario-specific operational control evaluation module instantiated for an occurrence of a distinct vehicle operational scenario by allocating computing resources to the scenario-specific operational control evaluation module instance and populating the scenario-specific operational control evaluation module instance with operational environment information representing aspects of an operational environment of the autonomous vehicle corresponding to the occurrence of the distinct vehicle operational scenario, wherein the aspects include:
one or more aspects representing an external object within a defined distance from the autonomous vehicle; and
one or more aspects representing vehicle transportation network geometry for a portion of the vehicle transportation network that includes the occurrence of the distinct vehicle operational scenario;
determining a current vehicle control action, wherein determining the current vehicle control action includes:
receiving, as a candidate vehicle control action, a vehicle control action from the scenario-specific operational control evaluation module instance; and
identifying the candidate vehicle control action as the current vehicle control action, wherein the current vehicle control action is a stop vehicle control action, an ‘advance’ vehicle control action, an ‘accelerate’ vehicle control action, a ‘decelerate’ vehicle control action, a ‘maintain’ vehicle control action, or a ‘proceed’ vehicle control action; and
controlling the autonomous vehicle to traverse the portion of the vehicle transportation network in accordance with the current vehicle control action.