Passive Landmark Geometry Optimization and Evaluation for Reliable Autonomous Navigation in Mining Tunnels Using 2D...

Area of Science:

• Robotics and Autonomous Systems
• Geospatial Navigation
• Artificial Intelligence

Background:

• Mining tunnels lack GPS and natural landmarks, complicating autonomous navigation.
• Existing solutions often require power-intensive beacons, posing safety and maintenance risks.

Purpose of the Study:

• To develop an autonomous navigation approach for mining tunnels using optimized artificial passive landmarks.
• To ensure drift-free localization for mobile units equipped with lidar scanners.

Main Methods:

• Optimizing artificial landmark geometry and positioning using a genetic algorithm.
• Employing scan matching techniques for pose estimation.
• Validating the approach through simulation and real-world experiments with a robotic excavator.

Main Results:

• Optimized triangular and symmetrical landmarks improved positioning accuracy by 87.5% per 100m traveled.
• Demonstrated feasibility with inexpensive, low-maintenance passive landmarks.
• Showcased the inadequacy of natural features alone for unambiguous localization.

Conclusions:

• Optimized artificial landmarks are crucial for reliable autonomous navigation in smoothly bored mining tunnels.
• The proposed method enables the implementation of grid-based SLAM (Simultaneous Localization and Mapping) techniques.
• This approach offers a practical and effective solution for challenging underground environments.