Method and system for hierarchical decomposition of tasks and action planning in a robotic network

Patent metadata

Field	Value
Publication number	US-2020005162-A1
Application number	US-201916460531-A
Country	US
Kind code	A1
Filing date	Jul 2, 2019
Priority date	Jul 2, 2018
Publication date	Jan 2, 2020
Grant date	—

Abstract

This disclosure relates generally to robotic network, and more particularly to a method and system for hierarchical decomposition of tasks and task planning in a robotic network. While a centralized system is used for action planning in a robotic network, any communication network issues can adversely affect working of the robotic network. Further, hardcoding one or more specific tasks to a robot restricts use of the robots irrespective of capabilities of the robots. The robotic agent decomposes a goal assigned to the robot to multiple sub-goals, and for each sub-goal, identifies one or more tasks to be executed/performed by the robot. An action plan is generated based on all such tasks identified, and the robot executes the action plan, in response to the goal assigned to the robot.

First claim

What is claimed is:

1 . A processor-implemented method for action planning in a robotic network, comprising:
assigning at least one goal to at least one robot in the robotic network, via one or more hardware processors; collecting the assigned at least one goal, via the one or more hardware processors, by a robotic agent of the at least one robot; decomposing the at least one goal to a plurality of sub-goals, via the one or more hardware processors, by the robotic agent; for each sub-goal:
identifying at least one target and at least one action on the target, as requirements corresponding to the sub-goal;
identifying a plurality of attributes corresponding to the at least one target, comprising:
fetching said plurality of attributes from a local database of the robot, if said plurality of attributes are available in the local database;
fetching said plurality of attributes from a global database of the at least one robot, if said plurality of attributes are not available in the local database of the at least one robot, and are available in the global database; and
obtaining said plurality of attributes dynamically based on at least one real-time input obtained based on perception of the at least one robot;

identifying a set of attributes corresponding to the at least one action, comprising:
fetching said set of attributes from the local database of the at least one robot, if said set of attributes are available in the local database;
fetching said set of attributes from the global database of the at least one robot, if said set of attributes are not available in the local database of the at least one robot, and are available in the global database; and
obtaining said set of attributes dynamically based on at least one real-time input obtained based on perception of the at least one robot, if the set of attributes are not available in the local database and the global database;

determining capability of the at least one robot, in terms of at least one specification of the at least one robot, by querying a robot capability knowledge base; and
generating an action plan corresponding to the collected at least one goal, based on (i) the set of attributes corresponding to the at least one action, (ii) the plurality of attributes corresponding to the at least one target, (iii) the determined capability of the at least one robot, (iv) data pertaining to an environment in which the at least one robot is deployed, and (v) data pertaining to at least one algorithm to be executed by the at least one robot for executing the action plan.

2 . The method as claimed in claim 1 , wherein the plurality of attributes corresponding to the at least one target comprises of at least one object attribute representing at least one characteristic of the at least one target.

3 . The method as claimed in claim 1 , wherein the set of attributes corresponding to the at least one action comprises of a navigation plan and a manipulation plan.

4 . The method as claimed in claim 3 , wherein the navigation plan is used for the at least one robot to reach the at least one target.

5 . The method as claimed in claim 3 , wherein the manipulation plan comprises data pertaining to at least one action to be performed on the at least one target, by the at least one robot.

6 . The method as claimed in claim 1 , wherein the global database comprises of:
i. a world model database comprising: 1) an environment map, 2) data pertaining to facility description, and one or more point cloud templates; ii. an object template database comprising data pertaining to 1) object description of one or more objects, 2) at least one feature and at least one template describing one or more characteristics of the one or more objects; iii. task description database, comprising data pertaining to 1) data pertaining to description of one or more tasks, and 2) mapping of the task descriptions with features of the environment and at least one object; iv. robot description database, comprising data pertaining to 1) robot model, 2) one or more components of the robot, 3) and specification representing one or more capability of the robot to perform one or more tasks; and v. data pertaining to one or more algorithms to be executed as part of handling one or more goals by the at least one robot.

7 . A robotic agent ( 100 ) for action planning in a robotic network, comprising:
a memory module ( 101 ) storing a plurality of instructions; one or more communication interfaces ( 103 ); and one or more hardware processors ( 102 ) coupled to the memory module ( 101 ) via the one or more communication interfaces ( 103 ), wherein the one or more hardware processors ( 102 ) are configured by the instructions to:
collect data pertaining to at least one goal that has been assigned to a robot with which the robotic agent is associated;
decompose the at least one goal to a plurality of sub-goals;
for each sub-goal:
identify at least one target and at least one action on the target, as requirements corresponding to the sub-goal;
identify a plurality of attributes corresponding to the at least one target, comprising:
fetching said plurality of attributes from a local database of the at least one robot, if said plurality of attributes are available in the local database;
fetching said plurality of attributes from a global database of the at least one robot, if said plurality of attributes are not available in the local database of the robot, and are available in the global database; and
obtaining said plurality of attributes dynamically based on at least one real-time input obtained based on perception of the at least one robot;

identify a set of attributes corresponding to the at least one action, comprising:
fetching said set of attributes from the local database of the robot, if said plurality of attributes are available in the local database;
fetching said plurality of attributes from the global database of the robot, if said plurality of attributes are not available in the local database of the robot, and are available in the global database; and
obtaining said plurality of attributes dynamically based on at least one real-time input obtained based on perception of the at least one robot, if the plurality of parameters are not available from the local database and the global database;

determine capability of the at least one robot, in terms of at least one specification of the at least one robot, by querying a robot capability knowledge base; and
generate an action plan corresponding to the collected at least one goal, based on (i) the set of attributes corresponding to the at least one action, (ii) the plurality of attributes corresponding to the at least one target, (iii) the determined capability of the robot, (iv) data pertaining to environment in which the at least one robot is deployed, and (v) data pertaining to at least one algorithm to be executed by the at least one robot for executing the action plan.

8 . The robotic agent as claimed in claim 7 , wherein the robotic agent is configured to identify at least one object attribute representing at least one characteristic of the at least one target as the plurality of attributes corresponding to the at least one target.

9 . The robotic agent as claimed in claim 7 , wherein the robotic agent is configured to identify a navigation plan and a manipulation plan as the set of attributes corresponding to the at least one action.

10 . The robotic agent as claimed in claim 9 , wherein the robotic agent is configured to use the navigation plan to navigate the at least one robot to reach the at least one target.

11 . The robotic agent as claimed in claim 9 , wherei

Assignees

• Tata Consultancy Services Ltd

Inventors

• Kattepur Ajay
• Dey Sounak
• Purushothaman Balamuralidhar

Classifications

• G05B2219/40113

  Task planning · CPC title

• G06F16/2455

  Query execution · CPC title

• G05B19/4155

  characterised by program execution, i.e. part program or machine function execution, e.g. selection of a program · CPC title

• G06N5/02 Primary

  Knowledge representation; Symbolic representation · CPC title

• G05B2219/39536

  Planning of hand motion, grasping · CPC title