Facilitating device control

• US 10,500,726 B2
• Filed: 04/20/2017
• Issued: 12/10/2019
• Est. Priority Date: 04/25/2016
• Status: Active Grant

• Current Assignee: Ocado Innovation Limited (Ocado Group PLC)
• Original Assignee: Kindred Systems Inc (LifePoint Health, Inc.)
• Inventors: Gildert, Suzanne
• Primary Examiner(s): Sample, Jonathan L
• Application Number:
  US15/492,619
• Publication Number:
  US 20170305014A1
• Time in Patent Office: 964 Days
• Field of Search: 700245-264
• CPC Class Codes: B25J 9/161 Hardware, e.g. neural networks, fuzzy logic, interfaces, processor
• CPC Class Codes: B25J 9/1664 characterised by motion, path, trajectory planning
• CPC Class Codes: B25J 9/1692 Calibration of manipulator
• CPC Class Codes: G05B 2219/39024 Calibration of manipulator
• CPC Class Codes: Y10S 901/01 Mobile robot

Abstract

Systems, devices, articles, and methods are illustrated and described herein. A method of operation in a robotic system including a processor, a first device, and a second device involves receiving, by the processor, a training set including a first plurality of positions in a first configuration space that represents physical configurations of the first device, a second plurality of positions in a second configuration space that represents physical configurations of the second device, and information that represents pairs of positions. A representative pair includes a first representative position in the first configuration space and a second representative position in the second configuration space. The method involves creating, by the processor, from the training set, information that represents a map between a first run-time position in the first configuration space, and a second run-time position in the second configuration space, and returning, by the processor, the information that represents the map.

Claims

1. A method of operation in a robotic system comprising, a processor, a first device in communication with the processor, and a second device in communication with the first device and the processor, the method comprising:

receiving, by the processor, a training set including;

a first plurality of positions in a first configuration space that represents physical configurations of the first device,

a second plurality of positions in a second configuration space that represents physical configurations of the second device, and

information that represents a plurality of pairs of positions, wherein a representative pair in the plurality of pairs of positions includes a first representative position in the first configuration space and a second representative position in the second configuration space;

creating, by the processor, from the training set, information that represents a map between a first run-time position in the first configuration space and a second run-time position in the second configuration space;

creating, by the processor, information that represents a plurality of nearest neighbour cells, wherein each cell in the plurality of nearest neighbour cells includes a position from the first plurality of positions in the first configuration space;

including, by the processor, in the information that represents the map, the information that represents the plurality of nearest neighbour cells; and

returning, by the processor, the information that represents the map.

2. The method of claim 1, wherein receiving the training set further comprises:

receiving, by the processor, the first plurality of positions in the first configuration space;

receiving, by the processor, the second plurality of positions in the second configuration space;

receiving, by the processor, the information that represents the plurality of pairs of positions; and

storing, by the processor, the information that represents the plurality of pairs of positions on at least one tangible computer readable storage device.

3. The method of claim 2, wherein:

the first device comprises an operator interface in communication with the processor;

the first configuration space includes a configuration space for the operator interface; and

receiving the first plurality of positions in the first configuration space further comprises;

receiving, by the processor from the operator interface, the first plurality of positions in the configuration space for the operator interface.

4. The method of claim 2, wherein:

the second device comprises an operator controllable device in communication with the processor;

the second configuration space includes a configuration space for the operator controllable device; and

receiving the second plurality of positions in the second configuration space further comprises;

receiving, by the processor from the operator controllable device, the second plurality of positions in the configuration space for the operator controllable device.

5. The method of claim 1 further comprising:

processing, by the processor, the training set into a processed training set.

6. The method of claim 1 further comprising:

creating, by the processor, information that represents a cut off distance in the first configuration space; and

including, by the processor, in the information that represents the map, the information that represents the cut off distance in the first configuration space.

9. The method of claim 1 further comprising:

selecting, by the processor, information that represents a plurality of parameters for a Gaussian process; and

including, by the processor, the information that represents the map, the information that represents the plurality of parameters for the Gaussian process.

10. The method of claim 1 further comprising:

storing, by the processor, the information that represents the map on at least one tangible computer readable storage media.

7. A method of operation in a robotic system comprising, a processor, a first device in communication with the processor, and a second device in communication with the first device and the processor, the method comprising:

receiving, by the processor, a training set including;

a first plurality of positions in a first configuration space that represents physical configurations of the first device,

a second plurality of positions in a second configuration space that represents physical configurations of the second device, and

information that represents a plurality of pairs of positions, wherein a representative pair in the plurality of pairs of positions includes a first representative position in the first configuration space and a second representative position in the second configuration space;

creating, by the processor, from the training set, information that represents a map between a first run-time position in the first configuration space and a second run-time position in the second configuration space;

creating, by the processor, a dictionary of atoms and a plurality of weights that in linear combination define at least one representative pair in the plurality of pairs of positions;

including, by the processor, in the information that represents the map, the information that represents the dictionary of atoms; and

returning, by the processor, the information that represents the map.

8. The method of claim 7 further comprising:

selecting, by the processor, a fit function that includes the dictionary of atoms and the plurality of weights; and

improving, by the processor, the fit function wherein in response to the improving by the processor, the processor creates the dictionary of atoms and the plurality of weights.

11. A method of operation in a robotic system comprising, a processor, an operator interface in communication with the processor, and an operator controllable device in communication with the processor and the operator interface, the method comprising:

receiving, by the processor, information that represents a map between a first position in a first configuration space associated with physical configurations for the operator interface and a second position in a second configuration space associated with physical configurations for the operator controllable device;

receiving, by the processor, a first run-time position in the first configuration space;

receiving, by the processor, information that represents a plurality of nearest neighbour cells, wherein each cell in the plurality of nearest neighbour cells includes a calibration position from a first plurality of calibration positions in the first configuration space;

receiving, by the processor, information that represents a second plurality of calibration positions in the second configuration space, wherein a representative calibration position from the second plurality of calibration positions in the second configuration space is associated with a representative calibration position from the first plurality of calibration positions in the first configuration space;

creating, by the processor, a second run-time position in the second configuration space from the information that represents the map, the first run-time position in the first configuration space, and the second plurality of calibration positions in the second configuration space;

causing, by the processor, an action to be taken with the second run-time position in the second configuration space; and

producing, by the processor, a signal including information that represents an action for the operator controllable device, wherein the information when interpreted causes the operator controllable device to assume the second run-time position.

12. The method of claim 11 further comprising:

receiving, by the processor from the operator interface, the first run-time position in the first configuration space for the operator interface.

13. The method of claim 11 further comprising:

creating, by the processor, the second run-time position in the second configuration space for the operator controllable device.

14. The method of claim 11 further comprising:

receiving, by the processor, information that represents a cut off distance in the first configuration space;

selecting by the processor a cell from the plurality of nearest neighbour cells, wherein the first run-time position is within the cell and the cell includes the calibration position from the first plurality of calibration positions in the first configuration space;

finding, by the processor, a distance between the first run-time position and the calibration position; and

if the distance is less than the cut off distance, including, by the processor, in the second run-time position, information that represents an error.

17. The method of claim 11 further comprising:

receiving, by the processor, information that represents a plurality of parameters for a Gaussian process;

receiving, by the processor, a training set including information that represents a plurality of pairs of positions, wherein a representative pair in the plurality of pairs of positions includes a first representative position in the first configuration space and a second representative position in the second configuration space; and

creating, by the processor, the second run-time position in the second configuration space from;

the information that represents the map,

the information that represents the plurality of parameters for the Gaussian process,

the training set, and

the first run-time position in the first configuration space.

18. The method of claim 11, wherein causing the action to be taken with the second run-time position in the second configuration space, further comprises:

storing, by the processor, the second run-time position in the second configuration space on at least one tangible computer readable storage media, or

returning, by the processor, the second run-time position in the second configuration space.

19. The method of claim 11, wherein causing the action to be taken with the second run-time position in the second configuration space, further comprises:

causing, by the processor, the operator controllable device to move to the second run-time position in the second configuration space.

15. A method of operation in a robotic system comprising, a processor, an operator interface in communication with the processor, and an operator controllable device in communication with the processor and the operator interface, the method comprising:

receiving, by the processor, information that represents a map between a first position in a first configuration space associated with physical configurations for the operator interface and a second position in a second configuration space associated with physical configurations for the operator controllable device;

receiving, by the processor, a first run-time position in the first configuration space;

creating, by the processor, a second run-time position in the second configuration space from the information that represents the map and the first run-time position in the first configuration space;

causing, by the processor, an action to be taken with the second run-time position in the second configuration space;

producing, by the processor, a signal including information that represents an action for the operator controllable device, wherein the information when interpreted causes the operator controllable device to assume the second run-time position;

receiving, by the processor, information that represents a dictionary of atoms that in linear combination with a plurality of weights define a position in the first configuration space, and a position in the second configuration space, wherein the dictionary of atoms includes a first plurality of atoms associated with the first configuration space; and

deriving, by the processor, the plurality of weights from a system of equations that includes;

the first run-time position,

the first plurality of atoms, and

the plurality of weights.

16. The method of claim 15 further comprising:

creating, by the processor, the second run-time position in the second configuration space from a linear combination of the dictionary of atoms and the plurality of weights.