IP Library › Granted Patent US 11,351,671
Granted Patent B2
US 11,351,671 · App. 16/173,866 · Granted Jun 7, 2022

Intent based control of a robotic device

Inventors: Allison Thackston (San Jose, CA); Samuel Zapolsky (Mountain View, CA); Katarina Miller (Palo Alto, CA); Laura Stelzner (Mountain View, CA); Ron Goldman (Los Altos, CA)
Assignee: TOYOTA RESEARCH INSTITUTE, INC.
B25J9/1689B25J9/1697G05D1/0016G05D1/0038G05D2201/0217
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 11,351,671
App. No.
16/173,866
Granted
Jun 7, 2022
Kind
B2
Abstract

A method for controlling a robotic device includes monitor an action of an operator of the robotic device. The method also includes inferring an intended target based on the monitored action. The method further includes determining an intended action for the intended target. The method still further includes controlling the robotic device to perform the intended action.

Claims (52)

1. A method performed by a robotic device, comprising:

monitoring for a movement of an operator of the robotic device;

identifying a current movement of the operator based on the monitoring;

inferring an intended object, for an action to be performed by the robotic device, from a plurality of objects within a current field of view of an environment based on the identified current movement and a first prior action performed with another object by the robotic device;

inferring an object type of the intended object, one or more other objects surrounding the intended object, and limitations of the environment;

determining the action to be performed with the intended object based on inferring the intended object and also inferring the object type, the one or more other objects surrounding the intended object, and the limitations of the environment; and

performing, by the robotic device, the action with the intended object.

2. The method of claim 1 , further comprising transmitting, to the operator, a feed of the current field of view from a vision sensor of the robotic device.

3. The method of claim 1 , further comprising receiving, from the operator, confirmation of the intended object, in which the action is determined based on receiving the confirmation.

4. The method of claim 1 , in which the action is further determined based on a learned approach or a combination of the learned approach and a modeled approach.

5. The method of claim 4 , in which:

the learned approach is based on a second prior action of the robotic device; and

the modeled approach is based on a set of actions corresponding to the intended object.

6. The method of claim 1 , further comprising receiving confirmation of the action in which the action is performed in response to the confirmation.

7. The method of claim 1 , further comprising performing another action after inferring the intended object and prior to performing the action.

8. A robotic device, comprising:

a processor;

a memory coupled with the processor; and

instructions stored in the memory and operable, when executed by the processor, to cause the robotic device:

to monitor for a movement of an operator of the robotic device;

to identify a current movement of the operator based on the monitoring;

to infer an intended object, for an action to be performed by the robotic device, from a plurality of objects within a current field of view of an environment based on the identified current movement and a first prior action performed with another object by the robotic device;

to infer an object type of the intended object, one or more other objects surrounding the intended object, and limitations of the environment;

to determine the action to be performed with the intended object based on inferring the intended object and also inferring the object type, the one or more other objects surrounding the intended object, and the limitations of the environment; and

to perform the action with the intended object.

9. The robotic device of claim 8 , in which execution of the instructions further cause the robotic device to transmit, to the operator, a feed of the current field of view from a vision sensor of the robotic device.

10. The robotic device of claim 8 , in which:

execution of the instructions further cause the robotic device to receive, from the operator, confirmation of the intended target; and

the action is determined based on receiving the confirmation.

11. The robotic device of claim 8 , in which execution of the instructions further cause the robotic device to determine the action based on a learned approach or a combination of the learned approach and a modeled approach.

12. The robotic device of claim 11 , in which:

the learned approach is based on a second prior action of the robotic device; and

the modeled approach is based on a set of actions corresponding to the intended object.

13. The robotic device of claim 8 , in which:

execution of the instructions further cause the robotic device to receive confirmation of the action; and

the action is performed based on receiving the confirmation.

14. The robotic device of claim 8 , in which execution of the instructions further cause the robotic device to perform another action after inferring the intended object and prior to performing the action.

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

program code to monitor for a movement of an operator of the robotic device;

program code to identify a current movement of the operator based on the monitoring

program code to infer an intended object, for an action to be performed by the robotic device, from a plurality of objects within a current field of view of an environment based on the identified current movement and a first prior action performed with another object by the robotic device;

program code to infer an object type of the intended object, one or more other objects surrounding the intended object, and limitations of the environment;

program code to determine the action to be performed with the intended object based on inferring the intended object and also inferring the object type, the one or more other objects surrounding the intended object, and the limitations of the environment; and

program code to perform the action with the intended object.

16. The non-transitory computer-readable of claim 15 , in which the program code further comprises program code to transmit, to the operator, a feed of the current field of view from a vision sensor of the robotic device.

17. The non-transitory computer-readable of claim 15 , in which:

the program code further comprises program code to receive, from the operator, confirmation of the intended target; and

the action is determined based on receiving the confirmation.

18. The non-transitory computer-readable of claim 15 , in which the program code further comprises program code to determine the action based on a learned approach or a combination of the learned approach and a modeled approach.

19. The non-transitory computer-readable of claim 18 , in which:

the learned approach is based on a second prior action of the robotic device; and

the modeled approach is based on a set of actions corresponding to the intended target.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 15, 2022
From: TOYOTA RESEARCH INSTITUTE, INC.
To: TOYOTA JIDOSHA KABUSHIKI KAISHA
Reel/Frame 060526/0845 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 29, 2018
From: THACKSTON, ALLISON; ZAPOLSKY, SAMUEL; MILLER, KATARINA; STELZNER, LAURA; GOLDMAN, RON
To: TOYOTA RESEARCH INSTITUTE, INC.
Reel/Frame 047344/0739 →
Continuity (1)
Related Publication 20200130190A1 · Apr 30, 2020
Cited By (1)
US 12,629,839