Tunnel Inspection Demo (Fukushima, Winter 2024)
- Description: In the recent field test, conducted in Fukushima Robot Test Field, we deployed our self-designed UAV and dynamic mapping system within a real tunnel environment to perform comprehensive inspection tasks. The drone leveraged advanced LIO-based positioning and an integrated LiDAR-camera sensor suite, which enabled real-time dynamic object detection and tracking. This sophisticated approach ensured safe navigation throughout the tunnel, even in the presence of both static and dynamic obstacles deliberately placed to simulate real-world conditions. As a result, our system successfully completed the autonomous inspection of the entire tunnel, demonstrating its robust performance and enhanced safety in challenging environments.
- Role and Contribution:
- Develop the dynamic obstacle detection and tracking module.
- Use dynamic removal to assist building the global occupancy map.
- Implemented dynamic removal for occupancy map.
- Validated in tunnel with static and dynamic obstacles.
- Demo Video:Google Drive
Tech Stack
Hardware
Software
Key Features
Autonomous navigation
Navigates complex environments without GPS, using LiDAR and visual sensors for real-time localisation and mapping.
High odometry frequency
FAST-LIO2 fused with the PX4 IMU raises the odometry rate, giving accurate and stable flight control.
Light-weight design
Optimised for minimal weight, extending flight duration and improving manoeuvrability.
Modular architecture
Hardware and software components can be upgraded or maintained independently.