Semantic place recognition and localization

1. A computer-implemented method comprising:

receiving observation data that represents an observation of a portion of a property, wherein the observation data was obtained using one or more sensors of a robot while located at a particular position of the property;

analyzing the observation data to detect one or more objects within the portion of the property;

determining, based at least on the one or more objects detected within the portion of the property, a semantic zone type for the portion of the property;

accessing mapping data for the property, wherein the mapping data identifies (i) a plurality of semantic zones of the property, and (ii) for each semantic zone of the property, a type of the semantic zone and one or more respective objects assigned to the semantic zone; and

selecting, from among the plurality of semantic zones of the property identified by the mapping data and based at least on (i) the semantic zone type that was determined for the portion of the property and (ii) a comparison of at least one object detected within the portion of the property to the respective objects assigned to different ones of the plurality of semantic zones of the property, a particular semantic zone of the property identified by the mapping data; and

setting, as a current location of the robot at the property, a particular location at the property associated with the particular semantic zone of the property identified by the mapping data.

2. The computer-implemented method of claim 1 , wherein the observation data comprises one or more images of the portion of the property.

3. The computer-implemented method of claim 1 , wherein the observation data comprises a plurality of LIDAR measurements obtained from the portion of the property.

4. The computer-implemented method of claim 1 , wherein determining the semantic zone type for the portion of the property comprises:

determining that the one or more objects detected within the portion of the property is insufficient to determine a semantic zone type for the portion of the property;

receiving additional observation data that represents an additional observation of a different portion of the property, wherein the additional observation data was obtained using the one or more sensors of the robot while located at the particular position of the property;

analyzing the additional observation data to detect one or more additional objects within the different portion of the property; and

determining, based at least on the one or more objects detected within the portion of the property and the one or more additional objects detected within the different portion of the property, the semantic zone type for the portion of the property.

5. The computer-implemented method of claim 1 , wherein selecting the particular semantic zone of the property identified by the mapping data comprises:

comparing the semantic zone type for the portion of the property to the semantic zone types for at least one of the semantic zones of the property identified by the mapping data; and

determining that the semantic zone type for the portion of the property matches a semantic zone type for the particular semantic zone of the property identified by the mapping hierarchy.

6. The computer-implemented method of claim 1 , wherein selecting the particular semantic zone of the property identified by the mapping data comprises:

determining that the mapping data identifies multiple semantic zones of the property that are of the semantic zone type for the portion of the property;

based on determining that the mapping data identifies multiple semantic zones of the property that are of the semantic zone type corresponding to the portion of the property, determining one or more characteristics of the portion of the property based on the observation data;

comparing one or more characteristics of the portion of the property to the one or more characteristics of each of the multiple semantic zones of the property that are of the semantic zone type corresponding to the portion of the property; and

determining, based at least on comparing the one or more characteristics of the portion of the property to the one or more characteristics of each of the multiple semantic zones of the property that are of the semantic zone type corresponding to the portion of the property, that one or more characteristics of the portion of the property correspond to one or more characteristics of a particular semantic zone of the property that is of the semantic zone type corresponding to the portion of the property.

7. The computer-implemented method of claim 6 , wherein the one or more characteristics of each of the multiple semantic zones of the property comprise one or more characteristics of three-dimensional representations of each of the multiple semantic zones of the property.

8. The computer-implemented method of claim 7 , wherein selecting the particular semantic zone of the property identified by the mapping data comprises:

determining, based at least on the observation, a three-dimensional representation of the portion of the property;

comparing one or more characteristics of the three-dimensional representation of the portion of the property to the one or more characteristics of the three-dimensional representations of each of the multiple semantic zones of the property; and

determining that one or more characteristics of the three-dimensional representation of the portion of the property corresponds to one or more characteristics of the three-dimensional representation of the particular semantic zone of the property identified by the mapping data.

9. The computer-implemented method of claim 1 , wherein the mapping data comprises a hierarchical data structure that describes hierarchical relationships among semantic zones of the property and various objects located within the semantic zones of the property.

10. The computer-implemented method of claim 1 , wherein setting, as a current location of the robot at the property, the particular location at the property associated with the particular semantic zone comprises setting the particular semantic zone as the current location of the robot.

11. A system comprising:

one or more processors configured to execute computer program instructions; and

one or more non-transitory computer-storage media encoded with computer programs that, when executed by the processor, cause the system to perform operations comprising:

receiving observation data that represents an observation of a portion of a property, wherein the observation data was obtained using one or more sensors of a robot while located at a particular position of the property;

analyzing the observation data to detect one or more objects within the portion of the property;

determining, based at least on the one or more objects detected within the portion of the property, a semantic zone type for the portion of the property;

accessing mapping data for the property, wherein the mapping data identifies (i) a plurality of semantic zones of the property, and (ii) for each semantic zone of the property, a type of the semantic zone and one or more respective objects assigned to the semantic zone; and

selecting, from among the plurality of semantic zones of the property identified by the mapping data and based at least on (i) the semantic zone type that was determined for the portion of the property and (ii) a comparison of at least one object detected within the portion of the property to the respective objects assigned to different ones of the plurality of semantic zones of the property, a particular semantic zone of the property identified by the mapping data; and

setting, as a current location of the robot at the property, a particular location at the property associated with the particular semantic zone of the property identified by the mapping data.

12. The system of claim 11 , wherein the observation data comprises one or more images of the portion of the property.

13. The system of claim 11 , wherein the observation data comprises a plurality of LIDAR measurements obtained from the portion of the property.

14. The system of claim 11 , wherein determining the semantic zone type for the portion of the property comprises:

determining that the one or more objects detected within the portion of the property is insufficient to determine a semantic zone type for the portion of the property;

receiving additional observation data that represents an additional observation of a different portion of the property, wherein the additional observation data was obtained using the one or more sensors of the robot while located at the particular position of the property;

analyzing the additional observation data to detect one or more additional objects within the different portion of the property; and

determining, based at least on the one or more objects detected within the portion of the property and the one or more additional objects detected within the different portion of the property, the semantic zone type for the portion of the property.

15. The system of claim 11 , wherein selecting the particular semantic zone of the property identified by the mapping data comprises:

comparing the semantic zone type for the portion of the property to the semantic zone types for at least one of the semantic zones of the property identified by the mapping data; and

determining that the semantic zone type for the portion of the property matches a semantic zone type for the particular semantic zone of the property identified by the mapping hierarchy.

16. The system of claim 11 , wherein selecting the particular semantic zone of the property identified by the mapping data comprises:

determining that the mapping data identifies multiple semantic zones of the property that are of the semantic zone type for the portion of the property;

based on determining that the mapping data identifies multiple semantic zones of the property that are of the semantic zone type corresponding to the portion of the property, determining one or more characteristics of the portion of the property based on the observation data;

comparing one or more characteristics of the portion of the property to the one or more characteristics of each of the multiple semantic zones of the property that are of the semantic zone type corresponding to the portion of the property; and

determining, based at least on comparing the one or more characteristics of the portion of the property to the one or more characteristics of each of the multiple semantic zones of the property that are of the semantic zone type corresponding to the portion of the property, that one or more characteristics of the portion of the property correspond to one or more characteristics of a particular semantic zone of the property that is of the semantic zone type corresponding to the portion of the property.

17. The system of claim 11 , wherein the mapping data comprises a hierarchical data structure that describes hierarchical relationships among semantic zones of the property and various objects located within the semantic zones of the property.

18. The system of claim 11 , wherein setting, as a current location of the robot at the property, the particular location at the property associated with the particular semantic zone comprises setting the particular semantic zone as the current location of the robot.

19. One or more non-transitory computer-readable devices storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to perform operations comprising:

receiving observation data that represents an observation of a portion of a property, wherein the observation data was obtained using one or more sensors of a robot while located at a particular position of the property;

analyzing the observation data to detect one or more objects within the portion of the property;

determining, based at least on the one or more objects detected within the portion of the property, a semantic zone type for the portion of the property;

accessing mapping data for the property, wherein the mapping data identifies (i) a plurality of semantic zones of the property, and (ii) for each semantic zone of the property, a type of the semantic zone and one or more respective objects assigned to the semantic zone; and

selecting, from among the plurality of semantic zones of the property identified by the mapping data and based at least on (i) the semantic zone type that was determined for the portion of the property and (ii) a comparison of at least one object detected within the portion of the property to the respective objects assigned to different ones of the plurality of semantic zones of the property, a particular semantic zone of the property identified by the mapping data; and

setting, as a current location of the robot at the property, a particular location at the property associated with the particular semantic zone of the property identified by the mapping data.