Model-based analysis of multi-UAV path planning for surveying postdisaster building damage

Area of Science:

• Disaster Management
• Robotics and Automation
• Geospatial Analysis

Background:

• Accurate and timely damage assessment is crucial for effective emergency response post-disaster.
• Satellite imagery provides a broad overview, but detailed assessment of damaged buildings requires higher resolution data.
• Unmanned Aerial Vehicles (UAVs) and 3D reconstruction offer enhanced capabilities for detailed damage evaluation.

Purpose of the Study:

• To develop and evaluate a multi-UAV coverage path planning method for 3D reconstruction of post-disaster damaged buildings.
• To improve the quality and detail of 3D models generated for damage assessment.
• To optimize UAV flight paths for efficient data acquisition in disaster scenarios.

Main Methods:

• A multi-UAV coverage path planning methodology was developed using NetLogo3D and tested in Unity3D.
• The method generates and filters camera locations around damaged buildings, then clusters them using K-means or Fuzzy C-means.
• Route optimization was performed as a multiple traveling salesman problem, with path corrections for obstacle avoidance and efficiency.

Main Results:

• The proposed method generates optimized camera locations and flight paths for multiple UAVs.
• The algorithm successfully balances flight distance and time for efficient data collection.
• Examination of texture resolution shows the proposed method significantly outperforms conventional nadir-looking overhead flight methods in 3D mapping quality.

Conclusions:

• The developed multi-UAV path planning method enhances the 3D reconstruction quality of post-disaster damaged buildings.
• This approach provides more detailed and accurate data for emergency responders.
• The optimized path planning contributes to more efficient and effective disaster damage assessment.