Map selection for vehicle pose system

1. A method for locating a vehicle, comprising:

accessing, by a pose state estimator executing at a computing device, a previous position for the vehicle at a first time, wherein the previous position is on a first sub-map of a plurality of sub-maps, and wherein the computing device comprises a processor and a data storage;

receiving, by the pose state estimator and from a first localizer, a first position estimate for the vehicle at a second time after the first time, wherein the first position estimate is on a second sub-map of the plurality of sub-maps; and

sending, by the pose state estimator and to a second localizer, a sub-map change message indicating the second sub-map.

2. The method of claim 1 , further comprising:

receiving, by the first localizer, first remote sensor data describing an environment of the vehicle;

matching, by the first localizer, the first remote sensor data to the first sub-map to determine a first sub-map position estimate for the vehicle;

matching, by the first localizer, the first remote sensor data to the second sub-map to determine a second sub-map position estimate for the vehicle; and

determining, by the first localizer, that an error of the second sub-map position is less than an error of the first sub-map position.

3. The method of claim 2 , wherein the determining that the error of the second sub-map position is less than the error of the first sub-map position is based at least in part on a covariance indicator for the first sub-map position for the vehicle and a covariance indicator for the second sub-map position for the vehicle.

4. The method of claim 1 , further comprising:

receiving, by the first localizer, first remote sensor data describing an environment of the vehicle;

matching, by the first localizer, the first remote sensor data to the first sub-map to generate a first candidate position estimate;

in response to determining that an error of the first candidate position estimate is greater than a threshold error, matching, by the first localizer, the first remote sensor data to the second sub-map to generate a second candidate position estimate;

determining, by the first localizer, that an error of the second candidate position estimate is less than the error of the first candidate position estimate; and

sending the second candidate position estimate to the pose state estimator as the first position estimate.

5. The method of claim 1 , further comprising:

receiving, by the second localizer, second remote sensor data describing an environment of the vehicle;

receiving, by the second localizer, the sub-map change message; and

matching, by the second localizer, the second remote sensor data describing the environment of the vehicle to the second sub-map to generate a second position estimate.

6. The method of claim 1 , further comprising:

receiving, by the pose state estimator, a first set of position estimates from the first localizer, wherein the first set of position estimates comprises the first position estimate for the vehicle, and a second position on the first sub-map;

receiving, by the pose state estimator, a second set of position estimates from the second localizer, wherein the second set of position estimates comprises a third position estimate on the first sub-map and a fourth position estimate on the second sub-map; and

selecting, by the pose state estimator, the second sub-map based at least in part on an error of the first position estimate, an error of the second position estimate, an error of the third position estimate, and an error of the fourth position estimate.

7. The method of claim 6 , further comprising receiving, by the pose state estimator and from the first localizer, a first pose estimate for the vehicle at the second time, wherein the first pose estimate comprises the first position estimate and a first attitude estimate.

8. A system for locating a vehicle, comprising:

a computing system comprising a processor and a data storage, wherein the computing system is programmed to execute components comprising:

a pose state estimator executing at a computing device comprising a processor and a data storage, wherein the pose state estimator is programmed to perform operations comprising:

accessing a previous position for the vehicle at a first time, wherein the previous position is on a first sub-map of a plurality of sub-maps, and wherein the computing device comprises a processor and a data storage;

receiving from a first localizer, a first position estimate for the vehicle at a second time after the first time, wherein the first position estimate is on a second sub-map of the plurality of sub-maps; and

sending to a second localizer, a sub-map change message indicating the second sub-map.

9. The system of claim 8 , wherein the computing system is further programmed to execute the first localizer, and wherein the first localizer is programmed to perform operations comprising:

receiving first remote sensor data describing an environment of the vehicle;

matching the first remote sensor data to the first sub-map to determine a first sub-map position estimate for the vehicle;

matching the first remote sensor data to the second sub-map to determine a second sub-map position estimate for the vehicle; and

determining that an error of the second sub-map position is less than an error of the first sub-map position.

10. The system of claim 8 , wherein the determining that an error of the second sub-map position is less than an error of the first sub-map position is based at least in part on a covariance indicator for the first sub-map position for the vehicle and a covariance indicator for the second sub-map position for the vehicle.

11. The system of claim 8 , wherein the computing system is further programmed to execute the first localizer, and wherein the first localizer is programmed to perform operations comprising:

receiving first remote sensor data describing an environment of the vehicle;

matching the first remote sensor data to the first sub-map to generate a first candidate position estimate;

in response to determining that an error of the first candidate position estimate is greater than a threshold error, matching the first remote sensor data to the second sub-map to generate a second candidate position estimate;

determining that an error of the second candidate position estimate is less than the error of the first candidate position estimate; and

sending the second candidate position estimate to the pose state estimator as the first position estimate.

12. The system of claim 8 , wherein the computing system is further programmed to execute the second localizer, and wherein the second localizer is programmed to perform operations comprising:

receiving second remote sensor data describing an environment of the vehicle;

receiving the sub-map change message; and

matching the second remote sensor data describing the environment of the vehicle to the second sub-map to generate a second position estimate.

13. The system of claim 8 , wherein the pose state estimator is further programmed to perform operations comprising:

receiving a first set of position estimates from the first localizer, wherein the first set of position estimates comprises the first position estimate for the vehicle, and a second position on the first sub-map;

receiving a second set of position estimates from the second localizer, wherein the second set of position estimates comprises a third position estimate on the first sub-map and a fourth position estimate on the second sub-map; and

selecting the second sub-map based at least in part on an error of the first position estimate, an error of the second position estimate, an error of the third position estimate, and an error of the fourth position estimate.

14. A method for locating a vehicle, comprising:

accessing, by a pose state estimator executing at a computing device, a first position for the vehicle at a first time, wherein the first position is described on a first sub-map of a plurality of sub-maps, and wherein the computing device comprises a processor and a data storage;

receiving, by the pose state estimator, sensor data describing a motion of the vehicle;

determining, by the pose state estimator, a second position for the vehicle at a second time after the first time based at least in part on the first position of the vehicle and at least in part on the sensor data;

determining, by the pose state estimator, that the second position of the vehicle is on a second sub-map of the plurality of sub-maps; and

sending, by the pose state estimator, a sub-map change message indicating the second sub-map to a first localizer and to a second localizer.

15. The method of claim 14 , further comprising sending, by the pose state estimator, a second sub-map change message indicating that the second sub-map is active to a second localizer.

16. The method of claim 14 , further comprising:

receiving, by first localizer, remote sensor data describing an environment of the vehicle;

receiving, by the first localizer, the first sub-map change message; and

matching, by the first localizer, the remote sensor data describing the environment of the vehicle to the second sub-map to generate a position estimate; and

sending the position estimate to the pose state estimator.

17. The method of claim 14 , wherein the sensor data comprises Global Navigation Satellite System (GNSS) data describing the vehicle.

18. The method of claim 14 , wherein the sensor data comprises motion sensor data describing motion of the vehicle.

19. The method of claim 14 , further comprising:

receiving, by the pose state estimator and from the first localizer, a pose estimate for the second time, wherein the pose estimate is on the second sub-map; and

determining, by the pose state estimator, a revised second position for the vehicle based at least in part on the sensor data and at least in part on the pose estimate.

20. The method of claim 14 , further comprising:

receiving, by the first localizer, the sub-map change message;

generating, by the first localizer, a first position estimate on the second sub-map;

receiving, by the second localizer, the sub-map change message; and

generating, by the second localizer, a second position estimate on the second sub-map.