Robot Behaviour Tree Generation Using a Knowledge Graph

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Manual generation of behaviour trees for robot devices is time-consuming, expertise-dependent, prone to human error, and lacks flexibility, making it difficult to adapt to dynamic situations.

Innovation Solution

A method involving a behaviour tree knowledge graph and knowledge graph embedding to automatically generate behaviour trees from a text prompt, enriched with domain-specific knowledge, allowing for efficient and accurate adaptation to new scenarios.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If behaviour trees are generated manually by experts, then the behaviour trees can be customized for specific requirements, but the process becomes time-consuming and resource-intensive

Engineering Contradiction:

Improvecustomization capabilityVSAvoidgeneration time

Core Design Contradiction:

Ease of manufactureVSLoss of time

Solution Approach 1:

The system pre-processes and structures domain knowledge into a knowledge graph before behaviour tree generation. This preliminary organization of information enables rapid automatic generation of behaviour trees without requiring time-consuming manual expert input for each new scenario.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

A knowledge graph serves as an intermediary between domain expertise and behaviour tree generation. The knowledge graph captures and structures domain knowledge, allowing automatic systems to generate customized behaviour trees without direct human intervention, thus reducing time while maintaining customization capability.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If behaviour trees are generated manually, then expertise can be applied to ensure quality, but human error increases and flexibility decreases

Engineering Contradiction:

Improvequality assuranceVSAvoidflexibility

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The system incorporates feedback mechanisms where the knowledge graph is continuously updated and refined based on generated behaviour trees and domain validation. This ensures quality assurance through structured knowledge while maintaining flexibility to adapt to new scenarios by updating the knowledge base.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The knowledge graph and behaviour tree generation system are designed to be dynamic and adaptable. The system can automatically adjust to new scenarios by querying and applying relevant knowledge from the knowledge graph, eliminating the rigidity of manually fixed behaviour trees while maintaining quality through structured domain knowledge.

Inventive Principle:

Principle #15Dynamics

3Loss of information

If manual behaviour tree generation is used, then detailed domain knowledge can be incorporated, but the process lacks efficiency and automation

Engineering Contradiction:

Improvedomain knowledge retentionVSAvoidgeneration efficiency

Core Design Contradiction:

Loss of informationVSProductivity

Solution Approach 1:

The system replaces manual mechanical processes of behaviour tree creation with automated computational processes. The knowledge graph embedding model automatically queries and applies domain knowledge from the knowledge graph, eliminating manual work while preserving detailed domain knowledge through structured representation.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system transforms domain knowledge into a structured knowledge graph representation with specific parameters and relationships. This parameterized knowledge can be efficiently queried and applied automatically, maintaining detailed domain information while enabling high-speed automated behaviour tree generation through computational processes.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20250242491A1Method for generating a behaviour tree for controlling a robot device

Publication Date: 2025.07.31 ROBERT BOSCH GMBH

AI summary

A method for generating a behaviour tree for controlling a robot device. The method includes: combining a plurality of predetermined behaviour trees and background knowledge into a behaviour tree knowledge graph, representing the behaviour tree knowledge graph in a latent space; extracting, from a prompt describing a desired behaviour of the robot device, a prompt representation graph; supplementing the prompt representation graph according to relations specified by the behaviour tree knowledge graph; selecting a sub graphs of the behaviour tree knowledge graph depending on a similarity to the supplemented prompt representation graph; and generating the behaviour tree for controlling the robot device by adjusting the selected sub graph according to knowledge from the prompt and the behaviour tree knowledge graph.