Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Methodology for Designing Knowledge-Driven Missions for Robots

Guillermo GP-Lenza, Carmen DR. Pita-Romero, Miguel Fernandez-Cortizas, Pascual Campoy

TL;DR

This work tackles the challenge of enhancing autonomous robotic missions by embedding knowledge graphs into ROS 2 in a brownfield setting. It delivers a comprehensive methodology that progresses from defining mission objectives to structuring tasks, planning sequences, representing data in a KG, and designing the mission with a high-level language. The authors implement the approach within Aerostack2, detailing ROS 2 KG components (Knowledge Base, Extractors, Retrievers) and a multi-agent KG fusion strategy, demonstrated on a Gazebo-based search-and-rescue scenario with drones. Results indicate improved explainability, situational awareness, and coordination, supported by released tooling and datasets, while outlining future work on automation and reasoning diversification.

Abstract

This paper presents a comprehensive methodology for implementing knowledge graphs in ROS 2 systems, aiming to enhance the efficiency and intelligence of autonomous robotic missions. The methodology encompasses several key steps: defining initial and target conditions, structuring tasks and subtasks, planning their sequence, representing task-related data in a knowledge graph, and designing the mission using a high-level language. Each step builds on the previous one to ensure a cohesive process from initial setup to final execution. A practical implementation within the Aerostack2 framework is demonstrated through a simulated search and rescue mission in a Gazebo environment, where drones autonomously locate a target. This implementation highlights the effectiveness of the methodology in improving decision-making and mission performance by leveraging knowledge graphs.

A Methodology for Designing Knowledge-Driven Missions for Robots

TL;DR

This work tackles the challenge of enhancing autonomous robotic missions by embedding knowledge graphs into ROS 2 in a brownfield setting. It delivers a comprehensive methodology that progresses from defining mission objectives to structuring tasks, planning sequences, representing data in a KG, and designing the mission with a high-level language. The authors implement the approach within Aerostack2, detailing ROS 2 KG components (Knowledge Base, Extractors, Retrievers) and a multi-agent KG fusion strategy, demonstrated on a Gazebo-based search-and-rescue scenario with drones. Results indicate improved explainability, situational awareness, and coordination, supported by released tooling and datasets, while outlining future work on automation and reasoning diversification.

Abstract

This paper presents a comprehensive methodology for implementing knowledge graphs in ROS 2 systems, aiming to enhance the efficiency and intelligence of autonomous robotic missions. The methodology encompasses several key steps: defining initial and target conditions, structuring tasks and subtasks, planning their sequence, representing task-related data in a knowledge graph, and designing the mission using a high-level language. Each step builds on the previous one to ensure a cohesive process from initial setup to final execution. A practical implementation within the Aerostack2 framework is demonstrated through a simulated search and rescue mission in a Gazebo environment, where drones autonomously locate a target. This implementation highlights the effectiveness of the methodology in improving decision-making and mission performance by leveraging knowledge graphs.
Paper Structure (14 sections, 5 figures, 2 tables)

This paper contains 14 sections, 5 figures, 2 tables.

Table of Contents

  1. INTRODUCTION
  2. Related work
  3. METHODOLOGY
  4. Definition
  5. Structuring
  6. Planning
  7. Representation
  8. Knowledge Extraction
  9. Concept Design
  10. Knowledge Mapping
  11. Mission Design
  12. ROS 2 KG Implementation
  13. USE CASE: SEARCH AND RESCUE SCENARIO
  14. CONCLUSIONS

Figures (5)

  • Figure 1: A full overview of the described methodology. Circles represent each step described in the proposed methodology while rectangular boxes contain the outcomes of each step.
  • Figure 2: The Gazebo simulation environment.
  • Figure 3: A graphical view of the application of the methodology to enhance Aerostack2. Green components are the ones introduced to the system through the application of the methodology, while blue ones are related to Aerostack2. Knowledge extraction is allocated in each one of the agents, while the knowledge base and the knowledge retrievers are centralized.
  • Figure 4: Initial state of the KG. It represents both drones in their respective home stations, with their batteries highly charged and ready to begin the search and rescue mission.
  • Figure 5: After locating the person, the KG should represent the fact that the two drones are close to each other, that one of them has located the person, and that both of them are still flying.