Wenyi Liu,Yunfan Ren,Rui Guo,Vickie W. W. Kong,Anthony S. P. Hung,Fangcheng Zhu,Yixi Cai,Yuying Zou,Fu Zhang
Abstract:This work presents a LiDAR-based quadrotor system for slope inspection in dense vegetation environments. Cities like Hong Kong are vulnerable to climate hazards, which often result in landslides. To mitigate the landslide risks, the Civil Engineering and Development Department (CEDD) has constructed steel flexible debris-resisting barriers on vulnerable natural catchments to protect residents. However, it is necessary to carry out regular inspections to identify any anomalies, which may affect the proper functioning of the barriers. Traditional manual inspection methods face challenges and high costs due to steep terrain and dense vegetation. Compared to manual inspection, unmanned aerial vehicles (UAVs) equipped with LiDAR sensors and cameras have advantages such as maneuverability in complex terrain, and access to narrow areas and high spots. However, conducting slope inspections using UAVs in dense vegetation poses significant challenges. First, in terms of hardware, the overall design of the UAV must carefully consider its maneuverability in narrow spaces, flight time, and the types of onboard sensors required for effective inspection. Second, regarding software, navigation algorithms need to be designed to enable obstacle avoidance flight in dense vegetation environments. To overcome these challenges, we develop a LiDAR-based quadrotor, accompanied by a comprehensive software system. The goal is to deploy our quadrotor in field environments to achieve efficient slope inspection. To assess the feasibility of our hardware and software system, we conduct functional tests in non-operational scenarios. Subsequently, invited by CEDD, we deploy our quadrotor in six field environments, including five flexible debris-resisting barriers located in dense vegetation and one slope that experienced a landslide. These experiments demonstrated the superiority of our quadrotor in slope inspection.
What problem does this paper attempt to address?
### Problems the Paper Aims to Solve
The paper aims to address the challenges of slope inspection in dense vegetation environments. Specifically, the main goal of slope inspection is to assess the condition of man-made structures on slopes (such as flexible rockfall barriers) to determine if slope maintenance is needed. This plays a crucial role in ensuring the safety and daily life of residents.
### Background and Challenges
1. **Climate Disasters**: Cities like Hong Kong are prone to landslides due to climate disasters such as extreme rainfall and typhoons. To reduce the risk of landslides, the Geotechnical Engineering Office (GEO) has constructed steel flexible rockfall barriers in disaster-prone natural catchments.
2. **Limitations of Traditional Inspection Methods**: Traditional manual inspection methods face numerous challenges, including steep terrain and high costs and work difficulty due to dense vegetation. Additionally, dense vegetation and harsh environments (such as steep stairs, insects, etc.) are not suitable for prolonged human presence.
3. **Advantages of UAVs**: Compared to manual inspections, unmanned aerial vehicles (UAVs) equipped with LiDAR sensors and cameras have the following advantages:
- Maneuverability in complex terrain, capable of accessing narrow areas and high places.
- Ability to collect detailed data on terrain and obstacles.
- Reduced manpower requirements, lower operational risks, and increased work efficiency.
- Potentially eliminates the need for constructing maintenance access paths, saving costs and reducing criminal activities.
### Main Issues
1. **Hardware Design**: The overall design of the UAV must carefully consider its maneuverability in narrow spaces, flight time, and the types of sensors required for effective inspection.
2. **Software Design**: Navigation algorithms need to be designed to achieve obstacle avoidance flight in dense vegetation environments. Specifically, this includes:
- **Localization**: In dense vegetation environments, traditional GPS methods are unreliable, and the UAV needs to rely on onboard sensors (such as LiDAR and cameras) for accurate localization.
- **Mapping**: The onboard sensor data must be used to quickly and accurately update map information, especially regarding small objects (such as thin branches, ropes, and fine nets).
- **Planning and Control**: The UAV must effectively cope with wind disturbances and swiftly and smoothly avoid dynamic obstacles, ensuring safe navigation among swaying branches and other natural interferences.
### Solution
To address the above issues, the authors developed a LiDAR-based quadrotor UAV and its accompanying software system, including localization, mapping, planning, and control modules. Specifically:
1. **Hardware Design**: Developed a quadrotor UAV equipped with a LiDAR sensor, with a wheelbase of 320 mm, overall dimensions of 422 mm × 422 mm × 260 mm, a total mass of 2.1 kg, and a maximum flight time of 12 minutes.
2. **Software Design**:
- **Localization Module**: Utilizes the FAST-LIO2 framework, employing an iterative extended Kalman filter and incremental kd-Tree for high-precision real-time state estimation.
- **Mapping Module**: Based on ROG-Map, generates high-resolution local occupancy grid maps through techniques such as unknown grid cell inflation, infinite point ray casting, and incremental frontier updates.
- **Planning and Control Module**: Redesigned the integrated planning and control framework (IPC), adding auxiliary obstacle avoidance flight functions to help pilots avoid static and dynamic obstacles, ensuring safe navigation during inspections.
### Experimental Validation
To evaluate the feasibility of the hardware and software systems, the authors conducted functional tests in non-operational scenarios and, at the invitation of the Geotechnical Engineering Office (GEO), deployed the quadrotor UAV in six real environments, including five flexible rockfall barriers in dense vegetation and one slope affected by a landslide triggered by heavy rain. In all six field tests, the quadrotor UAV successfully completed the designated inspection tasks. Additionally, comparative experiments with the advanced commercial UAV DJI Mavic 3 showed that the quadrotor UAV performed better in dynamic obstacle avoidance and maneuverability in narrow areas, making it suitable for slope inspections in dense vegetation environments.