Sub-bin refinement for autonomous machines

Patent metadata

Field	Value
Publication number	US-9404239-B2
Application number	US-201414300221-A
Country	US
Kind code	B2
Filing date	Jun 9, 2014
Priority date	Jun 9, 2014
Publication date	Aug 2, 2016
Grant date	Aug 2, 2016

Abstract

A computer-implemented method for determining a cut location for a machine implement is provided. The method may include comparing a target cut volume to a projected cut volume associated with each boundary of a selected bin, designating the cut location as the boundary most closely approximating the target cut volume if both of the projected cut volumes at the boundaries are either greater than or less than the target cut volume, and designating the cut location as an average of the boundaries if the projected cut volumes at the boundaries are greater than and less than the target cut volume.

First claim

What is claimed is:

1. A computer-implemented method for determining a cut location for a machine implement, comprising:
comparing, by a controller, a target cut volume to a projected cut volume associated with each boundary of a selected bin;
designating, by the controller, the cut location as the boundary most closely approximating the target cut volume if both of the projected cut volumes at the boundaries are either greater than or less than the target cut volume; and
designating, by the controller, the cut location as an average of the boundaries if the projected cut volumes at the boundaries are greater than and less than the target cut volume; and
instructing, by the controller, the machine to execute a cut at the designated cut location.

2. The computer-implemented method of claim 1 , wherein the selected bin is selected from a plurality of bins corresponding to a digitalization of an action space, each of the plurality of bins being equal in size and including a range of cut locations, each cut location having a corresponding projected cut volume.

3. The computer-implemented method of claim 2 , wherein the selected bin is selected such that the associated range of cut locations has corresponding projected cut volumes most closely approximating the target cut volume.

4. The computer-implemented method of claim 1 , wherein the average of the boundaries is a weighted average.

5. The computer-implemented method of claim 4 , wherein the weighted average is determined by performing an interpolation of at least the boundaries of the selected bin, assigning a weight to each of the boundaries based on the interpolation, and calculating an average of the weighted boundaries.

6. The computer-implemented method of claim 1 , further tracking a location of the machine implement and previously engaged cut locations.

7. The computer-implemented method of claim 1 , further communicating instructions to the machine implement for engaging a cut at the cut location.

8. A control system for determining a cut location for a machine implement, comprising:
a memory configured to retrievably store one or more algorithms; and
a controller in communication with the memory and, based on the one or more algorithms, configured to at least:
test each boundary of a selected bin and corresponding projected cut volumes in relation to a target cut volume,
designate the cut location as the boundary most closely approximating the target cut volume if both of the projected cut volumes at the boundaries are either greater than or less than the target cut volume,
designate the cut location as an average of the boundaries if the projected cut volumes at the boundaries are greater than and less than the target cut volume; and
instruct the machine to execute a cut at the designated cut location.

9. The control system of claim 8 , wherein the controller is configured to select the selected bin from a plurality of bins corresponding to a digitalization of an action space, each of the plurality of bins being equal in size and including a range of cut locations, each cut location having a corresponding projected cut volume.

10. The control system of claim 9 , wherein the controller is configured to select the selected bin such that the associated range of cut locations has corresponding projected cut volumes most closely approximating the target cut volume.

11. The control system of claim 8 , wherein the controller is configured to determine the average of the boundaries based on a weighted average of the boundaries.

12. The control system of claim 11 , wherein the controller is configured to determine the weighted average by performing an interpolation of at least the boundaries of the selected bin, assigning a weight to each of the boundaries based on the interpolation, and calculating an average of the weighted boundaries.

13. The control system of claim 8 , wherein the controller is configured to track a location of the machine implement and previously engaged cut locations.

14. The control system of claim 8 , wherein the controller is configured to communicate instructions to the machine implement for engaging a cut at the cut location.

15. A controller for determining a cut location for a machine implement, comprising:
a boundary test module configured to compare a target cut volume to a projected cut volume associated with each boundary of a selected bin;
a boundary selection module configured to designate the cut location as the boundary most closely approximating the target cut volume if the boundary test module indicates both of the projected cut volumes at the boundaries to be either greater than or less than the target cut volume;
a boundary average calculation module configured to designate the cut location as an average of the boundaries if the boundary test module indicates the projected cut volumes at the boundaries to be greater than and less than the target cut volume; and
wherein the controller further instructs the machine to execute a cut at the designated cut location.

16. The controller of claim 15 , further comprising a digitalization module configured to digitalize an action space into a plurality of bins, each of the plurality of bins being equal in size and including a range of cut locations, each cut location having a corresponding projected cut volume.

17. The controller of claim 16 , further comprising a bin selection module configured to select the selected bin such that the associated range of cut locations has corresponding projected cut volumes most closely approximating the target cut volume.

18. The controller of claim 15 , wherein the boundary average calculation module is configured to calculate a weighted average of the boundaries.

19. The controller of claim 18 , wherein the boundary average calculation module is configured to perform an interpolation of at least the boundaries, assign a weight to each of the boundaries based on the interpolation, and designate the cut location based on an average of the weighted boundaries.

20. The controller of claim 15 , further comprising a communications module configured to at least track a location of the machine implement and previously engaged cut locations.

Assignees

• Caterpillar Inc

Inventors

• Wei Mo
• Taylor Michael Allen
• Edara Thandava Krishna

Classifications

• E02F9/261

  Surveying the work-site to be treated · CPC title

• E02F9/2054

  Fleet management · CPC title

• G05D1/021

  specially adapted to land vehicles · CPC title

• E02F9/264

  Sensors and their calibration for indicating the position of the work tool · CPC title

• E02F9/205 Primary

  Remotely operated machines, e.g. unmanned vehicles (E02F3/8866 takes precedence) · CPC title