Commonsense reasoning about task instructions
AssigneeHonda Motor Co., Ltd.
InventorsRakesh Gupta, Ken Hennacy
Life of this patent
Estimating the term…
Abstract
A system and method enable an autonomous machine such as an indoor humanoid robot to systematically process user commands and respond to situations. The method captures distributed knowledge from human volunteers, referred to as “commonsense knowledge.” The commonsense knowledge comprises classes such as steps for tasks, responses to situations, and locations and uses of objects. Filtering refines the commonsense knowledge into useful class rules. A second level of rules referred to as meta-rules performs reasoning by responding to user commands or observed situations, orchestrating the class rules and generating a sequence of task steps. A task sequencer processes the generated task steps and drives the mechanical systems of the autonomous machine.
Claims29
A method for enabling an autonomous machine to perform a task in an environment, the method comprising:
- receiving data collected by a distributed capture technique, said data comprising at least one class of knowledge associated with the task;
- automatically processing said at least one class of knowledge associated with the task to produce one or more categories of class rules, wherein said processing comprises filtering and extracting the at least one class of knowledge, and wherein the class rules comprise specific relationships within the at least one class of knowledge;
- and
- formulating a set of meta-rules that extract information from two or more categories of class rules to orchestrate an application of said class rules within the environment, each meta-rule identifying information from two or more categories of class rules;
- responsive to a first category of class rules missing information associated with the task, applying a meta-rule to retrieve the missing information associated with the task from a second category of class rules.
An apparatus for an autonomous machine to perform a task in an environment, the apparatus comprising:
- an input module receiving data collected by a distributed capture technique, said data comprising at least one class of knowledge associated with the task;
- and
- a processor module coupled to the input module and to a memory device;
- the memory device having computer program instructions embodied thereon to cause the processor module to implement a method for enabling the autonomous machine to perform the task in the environment, the method comprising the steps of:
- automatically processing said at least one class of knowledge associated with the task to produce one or more categories of class rules , wherein said processing comprises filtering and extracting the at least one class of knowledge, and wherein the class rules comprise specific relationships within the at least one class of knowledge, and
- formulating a set of meta-rules that extract information from two or more categories of class rules to orchestrate an application of said class rules within the environment, each meta-rule identifying information from two or more categories of class rules;
- responsive to a first category of class rules missing information associated with the task, applying a meta-rule to retrieve the missing details information associated with the task from a second category of class rules.
An apparatus for an autonomous machine to perform a task in an environment, the apparatus comprising:
- means to receive data collected by a distributed capture technique, said data comprising at least one class of knowledge associated with the task;
- means to automatically process said at least one class of knowledge associated with the task to produce one or more categories of class rules, wherein said processing comprises filtering and extracting the at least one class of knowledge, and wherein the class rules comprise specific relationships within the at least one class of knowledge;
- and
- means to formulate a set of meta-rules that extract information from two or more categories of class rules to orchestrate an application of said class rules within the environment, each meta-rule identifying information from two or more categories of class rules;
- means to, responsive to a first category of class rules missing information associated with the task, apply a meta-rule to retrieve the missing information associated with the task from a second category of class rules.
A computer program product, comprising a memory device having computer program instructions embodied thereon to cause a computer processor to implement a method for enabling an autonomous machine to perform a task in an environment, the method comprising:
- receiving data collected by a distributed capture technique, said data comprising at least one class of knowledge associated with the task;
- automatically processing said at least one class of knowledge associated with the task to produce one or more categories of class rules , wherein said processing comprises filtering and extracting the at least one class of knowledge, and wherein the class rules comprise specific relationships within the at least one class of knowledge;
- and
- formulating a set of meta-rules that extract information from two or more categories of class rules to orchestrate an application of said class rules within the environment, each meta-rule identifying information from two or more categories of class rules;
- responsive to a first category of class rules missing information associated with the task, applying a meta-rule to retrieve the missing information associated with the task from a second category of class rules.
Dependencies are inferred from the claim text.