IP Library › Granted Patent US 12,743,305
Granted Patent B2
US 12,743,305 · App. 18/311,792 · Granted Sep 22, 2026

Method and apparatus for collaborative task planning for artificial intelligence agents

Inventor: Yong-Ju Lee (Daejeon, KR)
Assignee: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE
G06F9/4881G06F8/30
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 12,743,305
App. No.
18/311,792
Granted
Sep 22, 2026
Kind
B2
Abstract

Disclosed herein is a method for task planning for collaboration of artificial intelligence (AI) agents. The method includes generating a scene graph using an image acquired by an AI agent and a human instruction and generating a machine instruction set for objects in the scene graph, and the scene graph includes relevance information between each of the objects in the scene graph and the human instruction.

Claims (30)

1 . A method for task planning for collaboration of artificial intelligence (AI) agents, comprising:

generating a scene graph using an image acquired by an AI agent;

generating a relevance map for the scene graph based on a human instruction; and

generating a machine instruction set using relevance of each of a plurality of objects in the relevance map,

wherein generating the relevance map comprises calculating the relevance between each of the plurality of objects in the scene graph and the human instruction using a pretrained AI neural network, and

wherein generating the machine instruction set comprises requesting collaboration from an additional AI agent located within a preset distance from the AI agent when an object, the relevance of which is greater than a preset value, is not present among the plurality of objects in the relevance map.

2 . The method of claim 1 , wherein the machine instruction set includes machine instructions corresponding to a sub-task level of the human instruction to be executed by the AI agent.

3 . The method of claim 1 , wherein generating the machine instruction set comprises generating the machine instruction set for objects, the relevance information of which is greater than the preset value.

4 . The method of claim 1 , wherein generating the machine instruction set comprises generating the machine instruction set based on information about the additional AI agent.

5 . The method of claim 4 , wherein the information about the additional AI agent includes information about a location thereof and information about whether collaboration is possible.

6 . The method of claim 5 , wherein generating the machine instruction set comprises generating the machine instruction set based on a number of AI agents that are located within the preset distance from the AI agent and capable of collaborating with the AI agent.

7 . The method of claim 1 , further comprising:

requesting collaboration from the additional AI agent when complexity of performing the machine instruction set is greater than a preset value.

8 . The method of claim 1 , wherein generating the machine instruction set is performed using an AI neural network trained using training data configured with images, human instructions, and machine instruction sets.

9 . The method of claim 5 , wherein generating the machine instruction set comprises estimating a number of AI agents required for performing the human instruction based on calculation of complexity of the human instruction and generating the machine instruction set when a number of additional AI agents located within the preset distance and capable of collaboration is greater than the number of AI agents required for performing the human instruction.

10 . An apparatus for task planning for collaboration of artificial intelligence (AI) agents, comprising:

memory in which at least one program is recorded; and

a processor for executing the program,

wherein:

the program includes instructions for performing generating a scene graph using an image acquired by an AI agent, generating a relevance map for the scene graph based on a human instruction, and generating a machine instruction set using relevance of each of a plurality of objects in the relevance map,

generating the relevance map comprises calculating the relevance between each of the plurality of objects in the scene graph and the human instruction using a pretrained AI neural network, and

generating the machine instruction set comprises requesting collaboration from an additional AI agent located within a preset distance from the AI agent when an object, the relevance of which is greater than a preset value, is not present among the plurality of objects in the relevance map.

11 . The apparatus of claim 10 , wherein the machine instruction set includes machine instructions corresponding to a sub-task level of the human instruction to be executed by the AI agent.

12 . The apparatus of claim 10 , wherein generating the machine instruction set comprises generating the machine instruction set for objects, the relevance information of which is greater than the preset value.

13 . The apparatus of claim 10 , wherein generating the machine instruction set comprises generating the machine instruction set based on information about the additional AI agent.

14 . The apparatus of claim 13 , wherein the information about the additional AI agent includes information about a location thereof and information about whether collaboration is possible.

15 . The apparatus of claim 14 , wherein generating the machine instruction set comprises generating the machine instruction set based on a number of AI agents that are located within the preset distance from the AI agent and capable of collaborating with the AI agent.

16 . The apparatus of claim 10 , wherein the program further includes an instruction for performing requesting collaboration from the additional AI agent when complexity of performing the machine instruction set is greater than a preset value.

17 . The apparatus of claim 10 , wherein generating the machine instruction set is performed using an AI neural network trained using training data configured with images, human instructions, and machine instruction sets.

18 . The apparatus of claim 14 , wherein generating the machine instruction set comprises estimating a number of AI agents required for performing the human instruction based on calculation of complexity of the human instruction and generating the machine instruction set when a number of additional AI agents located within the preset distance and capable of collaboration is greater than the number of AI agents required for performing the human instruction.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded May 3, 2023
From: LEE, YONG-JU
To: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE
Reel/Frame 063524/0272 →
Priority Claims (1)
KR 10-2022-0159460 · Nov 24, 2022 · national
Continuity (1)
Related Publication 20240176653A1 · May 30, 2024
References Cited (31)
US 10192195B1 · Brazeau · 2019 [cited by examiner]
US 11899839B1 · Skorheim · 2024 [cited by examiner]
US 12070847B1 · Ebrahimi Afrouzi · 2024 [cited by examiner]
US 12198430B1 · Kottur · 2025 [cited by examiner]
US 20120143910A1 · Park et al. · 2012 [cited by applicant]
US 20170083005A1 · Hickman · 2017 [cited by examiner]
US 20190080252A1 · Shinn · 2019 [cited by examiner]
US 20200103911A1 · Ma et al. · 2020 [cited by applicant]
US 20200171650A1 · Hallock · 2020 [cited by examiner]
US 20200293995A1 · El Shawwa · 2020 [cited by examiner]
US 20200356829A1 · Costabello · 2020 [cited by examiner]
US 20210086353A1 · Shah · 2021 [cited by examiner]
US 20210224579A1 · Jones et al. · 2021 [cited by applicant]
US 20210229281A1 · Natarajan · 2021 [cited by examiner]
US 20220055221A1 · Ko et al. · 2022 [cited by applicant]
US 20220126445A1 · Zhu · 2022 [cited by examiner]
US 20220198813A1 · Chiu et al. · 2022 [cited by applicant]
US 20230134924A1 · Sun · 2023 [cited by examiner]
US 20230162494A1 · Banerjee · 2023 [cited by examiner]
US 20230311335A1 · Hausman · 2023 [cited by examiner]
US 20240078429A1 · Finn et al. · 2024 [cited by applicant]
US 20240408762A1 · Landernäs · 2024 [cited by examiner]
US 20260031916A1 · Li · 2026 [cited by examiner]
JP 2022170790 · 2022 [cited by applicant]
JP 2022173244 · 2022 [cited by applicant]
KR 1020190044683 · 2019 [cited by applicant]
KR 102063181 · 2020 [cited by applicant]
KR 1020200087298 · 2020 [cited by applicant]
KR 1020220146726 · 2022 [cited by applicant]
Chang, Yun, Luca Ballotta, and Luca Carlone. “D-Lite: Navigation-Oriented Compression of 3D Scene Graphs for Multi-Robot Collaboration.” arXiv preprint arXiv:2209.06111 (2022). (Year: 2022). [cited by examiner]
Guo, Junfu, Changhao Li, Xi Xia, Ruizhen Hu, and Ligang Liu. “Asynchronous Collaborative Autoscanning with Mode Switching for Multi-Robot Scene Reconstruction.” arXiv preprint arXiv:2210.04413 (2022). (Year: 2022). [cited by examiner]