Beyond Coverage Path Planning: Can UAV Swarms Perfect Scattered Regions Inspections?

TL;DR

The paper tackles the challenge of inspecting multiple scattered Regions of Interest (ROIs) with UAVs under strict energy constraints. It introduces Fast Inspection of Scattered Regions (FISR) and the two-stage mUDAI framework, which first selects a single high-quality viewpoint per ROI via dual-annealing and then solves a time-aware multi-UAV Vehicle Routing Problem to generate energy-feasible trajectories. The approach balances data resolution and mission duration, offering objective variants (MCO and BCO) to tailor captures and demonstrating significant gains in efficiency and data quality over traditional CPP and single-UAV methods. Real-world deployments and simulations show faster missions, reduced extraneous data, and robust terrain-aware altitude adjustments, with open-source implementations and data to encourage adoption. The work advances UAV RS tasks involving non-connected ROIs, with practical impact for security, agriculture, and emergency response applications.

Abstract

Unmanned Aerial Vehicles (UAVs) have revolutionized inspection tasks by offering a safer, more efficient, and flexible alternative to traditional methods. However, battery limitations often constrain their effectiveness, necessitating the development of optimized flight paths and data collection techniques. While existing approaches like coverage path planning (CPP) ensure comprehensive data collection, they can be inefficient, especially when inspecting multiple non connected Regions of Interest (ROIs). This paper introduces the Fast Inspection of Scattered Regions (FISR) problem and proposes a novel solution, the multi UAV Disjoint Areas Inspection (mUDAI) method. The introduced approach implements a two fold optimization procedure, for calculating the best image capturing positions and the most efficient UAV trajectories, balancing data resolution and operational time, minimizing redundant data collection and resource consumption. The mUDAI method is designed to enable rapid, efficient inspections of scattered ROIs, making it ideal for applications such as security infrastructure assessments, agricultural inspections, and emergency site evaluations. A combination of simulated evaluations and real world deployments is used to validate and quantify the method's ability to improve operational efficiency while preserving high quality data capture, demonstrating its effectiveness in real world operations. An open source Python implementation of the mUDAI method can be found on GitHub (https://github.com/soc12/mUDAI) and the collected and processed data from the real world experiments are all hosted on Zenodo (https://zenodo.org/records/13866483). Finally, this online platform (https://sites.google.com/view/mudai-platform/) allows interested readers to interact with the mUDAI method and generate their own multi UAV FISR missions.

Figures (11)

• Figure 1: Fast Inspection of Scattered Regions task. Two UAVs tasked with inspecting 10 disjoint regions: (a) User-defined waypoints allow fast operation but lack path optimization and fair task assignment, while also leading to random ROI captures. (b) CPP planning ensures high-quality captures but introduces superfluous data and long-duration operations. (c) Our mUDAI method balances speed, comprehensiveness, and efficiency, overcoming these hurdles effectively.
• Figure 2: mUDAI methodology splits the FISR problem in two discrete optimization sub-problems - as a first step, given the ROIs and the specifications of the camera used, dual-annealing algorithm is utilized for the calculation of the optimal viewpoints, given the optimization objectives - as a second step, a VRP solver is used for the calculation of the optimal visitation order, respecting the energy constraints of the UAVs.
• Figure 3: The UAV, hovering in a distance from ground -z- captures an image that contains a representative shot of the region $R_i$. The aspect ratio of the captured image $(d_h:d_v)$ depends on the specifications of the sensor (hFOV, vFOV), while the content of the captured image -given the aspect ratio- depends on the position and rotation of the UAV $[x, y, z, \psi]$.
• Figure 4: Illustrative comparison of MCO and BCO optimization objectives. The blue polygon represents the ROI, while the red polygon indicates the camera capture area. In MCO, the objective is to fully capture the defined ROI while minimizing the area outside the ROI. In contrast, BCO strikes a balance between capturing the ROI and reducing coverage of areas outside the ROI, achieving a lower GSD and minimizing extraneous coverage.
• Figure 5: The inclusion of intermediate navigational waypoints in UAV trajectories, enabling transitions at constant altitudes unique to each UAV, ensures safe and collision-free operations. On the left side of the figure, a trajectory directly connecting the viewpoints is shown, which includes intersecting trajectories leading to a potential collision point, whereas on the right side, the introduction of intermediate waypoints has eliminated the presence of potential collision points.