Object location analysis

1. A method for controlling a robotic device based on observed object locations, comprising:

observing an object in an environment over a period of time;

generating, for the object, a location cluster comprising a plurality of nodes, each node in the plurality of nodes associated with one observation of a plurality of observations performed over the period of time;

localizing the object in the environment based on a centroid of the location cluster;

generating a probability distribution for observed locations of the object based on observing the object over the period of time; and

overlaying a cost map on a map of the environment, the cost map associated with the probability distribution to identify the probability distribution of the object in relation to the localized object in the environment; and

controlling the robotic device to perform an action in the environment based on the cost map and the map of the environment.

2. The method of claim 1 , further comprising observing the object over a period of time.

3. The method of claim 2 , further comprising estimating a continuous distribution using the observations of the object over the period of time.

4. The method of claim 3 , in which the probability distribution is based on the continuous distribution.

5. The method of claim 1 , further comprising generating the cost map from the probability distribution.

6. The method of claim 5 , further comprising, controlling the robot to:

avoid a first area in the environment with a first object probability that is greater than a first threshold; or

navigate to a second area in the environment with a second object probability that is greater than a second threshold.

7. The method of claim 1 , further comprising controlling the robotic device to place an object in the environment based on the probability distribution.

8. An apparatus for controlling a robotic device based on observed object locations, the apparatus comprising:

a memory; and

at least one processor coupled to the memory, the at least one processor configured:

to observe an object in an environment over a period of time;

to generate, for the object, a location cluster comprising a plurality of nodes, each node in the plurality of nodes associated with one observation of a plurality of observations performed over the period of time;

to localize the object in the environment based on a centroid of the location cluster;

to generate a probability distribution for observed locations of the object based on observing the object over the period of time;

to overlay a cost map on a map of the environment, the cost map associated with the probability distribution to identify the probability distribution of the object in relation to the localized object in the environment; and

to control the robotic device to perform an action in the environment based on the cost map and the map of the environment.

9. The apparatus of claim 8 , in which the at least one processor is further configured to observe the object over a period of time.

10. The apparatus of claim 9 , in which the at least one processor is further configured to estimate a continuous distribution using the observations of the object over the period of time.

11. The apparatus of claim 10 , in which the probability distribution is based on the continuous distribution.

12. The apparatus of claim 8 , in which the at least one processor is further configured to generate the cost map from the probability distribution.

13. The apparatus of claim 12 , in which the at least one processor is further configured to control the robot to:

avoid a first area in the environment with a first object probability that is greater than a first threshold; or

navigate to a second area in the environment with a second object probability that is greater than a second threshold.

14. The apparatus of claim 8 , in which the at least one processor is further configured to control the robotic device to place an object in the environment based on the probability distribution.

15. A non-transitory computer-readable medium having program code recorded thereon for controlling a robotic device based on observed object locations, the program code executed by a processor and comprising:

program code to observe an object in an environment over a period of time;

program code to generate, for the object, a location cluster comprising a plurality of nodes, each node in the plurality of nodes associated with one observation of a plurality of observations performed over the period of time;

program code to localize the object in the environment based on a centroid of the location cluster

program code to generate a probability distribution for observed locations of the object based on observing the object over the period of time;

program code to overlay a cost map on a map of the environment, the cost map associated with the probability distribution to identify the probability distribution of the object in relation to the localized object in the environment; and

program code to control the robotic device to perform an action in the environment based on the cost map and the map of the environment.

16. The non-transitory computer-readable medium of claim 15 , in which the program code further comprises program code to observe the object over a period of time.

17. The non-transitory computer-readable medium of claim 16 , in which the program code further comprises program code to estimate a continuous distribution using the observations of the object over the period of time.

18. The non-transitory computer-readable medium of claim 17 , in which the probability distribution is based on the continuous distribution.

19. The non-transitory computer-readable medium of claim 15 , in which the program code further comprises program code to generate a cost map from the probability distribution.

20. The non-transitory computer-readable medium of claim 19 , in which the program code further comprises program code to control the robot to:

avoid a first area in the environment with a first object probability that is greater than a first threshold; or

navigate to a second area in the environment with a second object probability that is greater than a second threshold.