Semantic mapping of environments for autonomous devices
Methods, systems, and apparatus for receiving a reference to an object located in an environment of a robot, accessing mapping data that indicates, for each of a plurality of object instances, respective probabilities of the object instance being located at one or more locations in the environment, wherein the respective probabilities are based at least on an amount of time that has passed since a prior observation of the object instance was made, identifying one or more particular object instances that correspond to the referenced object, determining, based at least on the mapping data, the respective probabilities of the one or more particular object instances being located at the one or more locations in the environment, selecting, based at least on the respective probabilities, a particular location in the environment where the referenced object is most likely located, and directing the robot to navigate to the particular location.
1. A computer-implemented method comprising:
receiving, by a system configured to facilitate operation of a robot in an environment, a reference to an object located in the environment of the robot;
accessing, by the system, mapping data that indicates, for each of a plurality of object instances, respective probabilities of the object instance being located at one or more locations in the environment, wherein the respective probabilities of the object instance being located at each of the one or more locations in the environment are based at least on an amount of time that has passed since a prior observation of the object instance was made;
identifying, by the system and from among the plurality of object instances, one or more particular object instances that correspond to the referenced object;
determining, by the system and based at least on the mapping data, the respective probabilities of the one or more particular object instances being located at the one or more locations in the environment;
selecting, by the system and based at least on the respective probabilities of the one or more particular object instances being located at the one or more locations in the environment, a particular location in the environment where the referenced object is most likely located; and
directing, by the system, the robot to navigate to the particular location.
2. The computer-implemented method of claim 1 , wherein each of the plurality of object instances is an instance of an object class.
3. The computer-implemented method of claim 1 , comprising:
controlling, by the system, navigation of the robot to the particular location.
4. The computer-implemented method of claim 1 , wherein the respective probabilities of an object instance being located at one or more locations in the environment are based at least on a temporal characteristic associated with an object class of the object instance.
5. The computer-implemented method of claim 1 , wherein the respective probabilities of an object instance being located at one or more locations in the environment are based at least on a spatial relationship between the object instance and one or more other object instances.
6. The computer-implemented method of claim 1 , wherein the respective probabilities of an object instance being located at one or more locations in the environment are based at least on a relationship between an object class of the object instance and an area type of a location associated with the probability.
7. The computer-implemented method of claim 1 , wherein the respective probabilities of an object instance being located at one or more locations in the environment are based at least on a spatiotemporal relationship between an object class of the object instance and a location associated with the probability.
8. The computer-implemented method of claim 1 , wherein the respective probabilities of an object instance being located at one or more locations in the environment are based at least on an identification confidence associated with one or more prior observations of the object instance.
9. The computer-implemented method of claim 1 , wherein the mapping data is generated based on observations of a plurality of robots.
10. The computer-implemented method of claim 1 , wherein the system accesses the mapping data at a cloud-based computing system.
11. A system comprising:
one or more processors configured to execute computer program instructions; and
one or more computer-storage media encoded with computer programs that, when executed by the one or more processors, cause the system to perform operations comprising:
receiving, by a system configured to facilitate operation of a robot in an environment, a reference to an object located in the environment of the robot;
accessing, by the system, mapping data that indicates, for each of a plurality of object instances, respective probabilities of the object instance being located at one or more locations in the environment, wherein the respective probabilities of the object instance being located at each of the one or more locations in the environment are based at least on an amount of time that has passed since a prior observation of the object instance was made;
identifying, by the system and from among the plurality of object instances, one or more particular object instances that correspond to the referenced object;
determining, by the system and based at least on the mapping data, the respective probabilities of the one or more particular object instances being located at the one or more locations in the environment;
selecting, by the system and based at least on the respective probabilities of the one or more particular object instances being located at the one or more locations in the environment, a particular location in the environment where the referenced object is most likely located; and
directing, by the system, the robot to navigate to the particular location.
12. The system of claim 11 , wherein each of the plurality of object instances is an instance of an object class.
13. The system of claim 11 , wherein the operations comprise:
controlling, by the system, navigation of the robot to the particular location.
14. The system of claim 11 , wherein the respective probabilities of an object instance being located at one or more locations in the environment are based at least on a temporal characteristic associated with an object class of the object instance.
15. The system of claim 11 , wherein the respective probabilities of an object instance being located at one or more locations in the environment are based at least on a spatial relationship between the object instance and one or more other object instances.
16. The system of claim 11 , wherein the respective probabilities of an object instance being located at one or more locations in the environment are based at least on a relationship between an object class of the object instance and an area type of a location associated with the probability.
17. The system of claim 11 , wherein the respective probabilities of an object instance being located at one or more locations in the environment are based at least on a spatiotemporal relationship between an object class of the object instance and a location associated with the probability.
18. The system of claim 11 , wherein the respective probabilities of an object instance being located at one or more locations in the environment are based at least on an identification confidence associated with one or more prior observations of the object instance.
19. One or more 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, by a system configured to facilitate operation of a robot in an environment, a reference to an object located in the environment of the robot;
accessing, by the system, mapping data that indicates, for each of a plurality of object instances, respective probabilities of the object instance being located at one or more locations in the environment, wherein the respective probabilities of the object instance being located at each of the one or more locations in the environment are based at least on an amount of time that has passed since a prior observation of the object instance was made;
identifying, by the system and from among the plurality of object instances, one or more particular object instances that correspond to the referenced object;
determining, by the system and based at least on the mapping data, the respective probabilities of the one or more particular object instances being located at the one or more locations in the environment;
selecting, by the system and based at least on the respective probabilities of the one or more particular object instances being located at the one or more locations in the environment, a particular location in the environment where the referenced object is most likely located; and
directing, by the system, the robot to navigate to the particular location.
20. The computer-readable device of claim 19 , wherein each of the plurality of object instances is an instance of an object class.