Real-time robot topological localization and mapping with limited visual sampling in simulated buried pipe networks

Area of Science:

• Robotics
• Geospatial Engineering
• Computer Vision

Background:

• Mapping underground utilities like pipe networks is crucial for infrastructure management.
• Existing methods often struggle with real-time performance and computational efficiency.
• Robotic systems offer a potential solution for autonomous exploration and mapping.

Purpose of the Study:

• To develop a real-time robotic localization and mapping system for buried pipe networks.
• To improve the accuracy and efficiency of pipe network mapping.
• To reduce computational requirements for robotic mapping systems.

Main Methods:

• Integration of non-vision-based exploration and navigation with active-vision-based localization.
• Selective activation of the localization and mapping algorithm at key topological locations (e.g., junctions).
• Utilizing non-vision-based sensors for junction detection to minimize visual data acquisition.

Main Results:

• Demonstrated feasibility of the proposed system through simulations.
• Achieved real-time performance with reduced computational load.
• Provided accurate and efficient mapping of a virtual pipe network.

Conclusions:

• The developed system offers a practical and efficient solution for real-time localization and mapping of buried pipe networks.
• The hybrid approach (non-vision and active-vision) effectively addresses challenges in underground environments.
• Simulations confirm the system's potential for autonomous infrastructure mapping.