Coverage Path Planning for Spraying Drones

Abstract: The pandemic by COVID-19 is causing a devastating effect on the health of global population. There are several efforts to prevent the spread of the virus. Among those efforts, cleaning and disinfecting public areas have become important tasks. In order to contribute in this direction, this paper proposes a coverage path planning algorithm for a spraying drone, a micro aerial vehicle that has mounted a sprayer/sprinkler system, to disinfect areas. In contrast with planners in the state-of-the-art, this proposal presents i) a new sprayer/sprinkler model that fits a more realistic coverage volume to the drop dispersion and ii) a planning algorithm that efficiently restricts the flight to the region of interest avoiding potential collisions in bounded scenes. The drone with the algorithm has been tested in several simulation scenes, showing that the algorithm is effective and covers more areas with respect to other approaches in the literature. Note that the proposal is not limited to disinfection applications, but can be applied to other ones, such as painting or precision agriculture.

• The paper introduces a novel CPP algorithm that enhances spray accuracy and safety by integrating a realistic 3D paraboloid model for droplet dispersion.
• It refines path planning using a modified rotating calipers method to ensure drones operate strictly within designated boundaries.
• Extensive simulations show reduced computation time and superior coverage, supporting real-time onboard implementation in diverse UAV tasks.

Coverage Path Planning for Spraying Drones: An Analytical Overview

The research paper "Coverage Path Planning for Spraying Drones" presents an innovative coverage path planning (CPP) algorithm specifically tailored for spraying drones. The study, conducted by Vazquez-Carmona et al., focuses on the need for automated disinfection in light of the COVID-19 pandemic and its implications for future smart cities. The paper proposes a novel approach that improves the effectiveness of UAVs equipped with sprayer systems by refining their path planning capabilities.

Algorithmic Contributions

The authors introduce a new CPP algorithm that addresses two significant limitations of existing approaches: i) the inadequacy of current drone coverage models which assume a camera-based projection and ii) the assumption that drones can safely navigate outside target regions, which is often infeasible in low-altitude spraying tasks. The proposed method involves:

1. Sprinkler Modeling: The introduction of a 3D paraboloid model fitting the realistic dispersion of droplets. This model assumes an altitude-bound operation, crucial for maintaining disinfection efficacy.
2. Efficient Path Planning: Utilization of a modified rotating calipers path planner to ensure that the UAV’s path remains strictly within the boundaries of the region of interest (ROI). This prevents potential collision with surrounding structures and enhances safety and efficiency.
3. Computation Time Reduction: By integrating heuristic improvements and modifying previous algorithms, the proposed solution achieves reduced computation times, facilitating onboard real-time implementations.

Experimental Validation

The paper details extensive simulated experiments where the proposed method is benchmarked against two state-of-the-art approaches. The algorithm demonstrates superior coverage capabilities, addressing diverse polygonal shapes without compromising safety through extraneous off-border flights. The computation time is reported as significantly lower compared to comparator methods, illustrating the algorithm's suitability for onboard processing.

Practical Implications and Future Directions

The implications of this research are profound, especially in scenarios requiring large-scale urban disinfection. The flexibility of the model extends beyond pandemic response to include applications in precision agriculture and urban maintenance tasks like painting. The ability to customize paths based on realistic spraying footprints may enhance operational effectiveness across various UAV applications.

For future work, the authors highlight extending the method to non-convex regions and pursuing real-world validation, advancing towards deployment in heterogeneous and complex urban terrains. Continued refinement in real-time collision avoidance, as well as adaptation to varying environmental conditions, would also be prudent.

Theoretical Contributions

Theoretically, this paper contributes to the CPP domain by refining UAV path constraints and incorporating stochastic elements into deterministic models. By addressing the practical challenges associated with UAV planning in closed environments, the authors provide a framework that could be expanded to enhance robotic autonomy in various sectors.

In conclusion, this paper represents a step forward in UAV-based automation for urban applications, offering a methodologically sound approach to improve operational efficiency and safety in aerial disinfection and other related tasks.