Robot Intent Inference Using Knowledge Graph and ECA Rules
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies struggle to accurately infer user intent from human activity and situational information for controlling robots, requiring significant domain knowledge and time, and often fail to discover new rules for advanced service robots.
Innovation Solution
A robot control apparatus that utilizes a knowledge graph and a rule creation model to infer user intent by combining knowledge-driven and data-driven approaches, updating event-condition-action (ECA) rules based on neural compositional rule learning (NCRL) and rule mining models, to enhance intent perception and task execution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional domain knowledge approaches are used to infer user intent, then intent perception can be achieved, but the time and cost required to obtain domain knowledge is significant
Solution Approach 1:
The system pre-builds a knowledge graph containing event instances, context instances, and their relationships before runtime inference. This preliminary structuring of domain knowledge allows the system to avoid time-consuming knowledge acquisition during actual intent perception, as the knowledge is already organized and ready for querying.
Solution Approach 2:
The patent introduces an ECA (Event-Condition-Action) rule-based intermediary layer that mediates between raw user activity data and intent inference. This rule-based system translates complex domain knowledge into structured inference rules, reducing the need for time-intensive manual knowledge engineering while maintaining accurate intent perception.
2Device complexity
If only user activity data is used for intent inference, then the inference process is simple, but the accuracy of intent perception is insufficient
Solution Approach 1:
The system merges multiple data sources including user activity data, event information from the knowledge graph, and context information into a unified inference framework. This combination of diverse information sources enhances intent perception accuracy while the rule-based ECA structure keeps the overall process manageable.
Solution Approach 2:
The patent adds contextual dimensions to the inference process by incorporating event instances and context instances from the knowledge graph alongside traditional user activity data. This multi-dimensional approach enriches the inference input without overwhelming complexity through structured ECA rules.
3Productivity
If existing ECA rules are used without updates, then the system operates efficiently, but new rules for advanced service robots cannot be discovered
Solution Approach 1:
The system implements a feedback mechanism where intent inference results are continuously fed back to update and refine ECA rules. This allows the system to learn from actual usage patterns and discover new rules while maintaining efficient operation through the structured rule-based framework.
Solution Approach 2:
The ECA rule set is designed to be dynamic rather than static, allowing rules to be added, modified, and discovered over time. This dynamic nature enables the system to adapt to new service scenarios while maintaining the efficiency benefits of rule-based processing through incremental updates.
Data Source
AI summary
A robot control apparatus can include a memory that stores computer-executable instructions and at least one processor that executes the instructions by accessing the memory. The at least one processor can apply event information about activity of a user, which can be identified from user activity data perceived from a robot, and context information about time and space in which the activity occurs, to a knowledge graph formed by a relation between an event instance regarding the event information and a context instance regarding the context information, obtain user intent data regarding intent of the activity by applying the event instance among instances included in the knowledge graph to a rule creation model for creating information about the intent of the activity, and control the robot such that the robot performs a target task related to expected activity, which can follow the activity, based on the user intent data.