Mosquito-borne illnesses remain a global health threat, and current control measures often fall short. Pesticides and environmental management are labor-intensive, costly, and damaging to ecosystems. While robotic systems have begun automating mosquito trapping, most lack intelligent decision-making about where to operate.
The missing link has been the ability to focus on places where people—and therefore mosquito activity—are most concentrated.
The new human-first approach for complete coverage path planning (HFA-CCPP) directly addresses this challenge. By guiding mosquito-control robots to prioritize human-dense areas first, the method improves trapping efficiency and reduces transmission risks compared with non-prioritized strategies.
Inside the Dragonfly Robot
To test the system, researchers used the Dragonfly robot, a compact yet sophisticated platform built for autonomous mosquito control. Its differential drive system, powered by two 24-volt brushless DC motors, allows reliable navigation in semi-outdoor environments. A sturdy aluminum-and-3D-printed chassis, supported by three wheels, ensures durability and balance.
Processing power comes from an industrial-grade computer (i7 processor, 16 GB RAM) that fuses data from multiple sensors: a 2D lidar for mapping and navigation, a depth camera for 3D perception, and a VectorNav IMU for orientation. A 24V battery supports 8–10 hours of operation, giving the robot ample runtime for large areas.
But the standout feature is its mosquito-trapping mechanism. Built into the robot’s “neck,” the system uses a combination of cues to mimic humans: a safe 368 nm UV light and a synthetic sweat scent blend of octanol and lactic acid. Drawn by these signals, mosquitoes are then pulled in by a high-speed suction fan (1000 RPM) and captured on sticky tape. This design enables Dragonfly to autonomously patrol and reduce mosquito populations in targeted zones.
Testing the Human-First Strategy
The HFA-CCPP algorithm was evaluated through extensive simulations and real-world experiments. In virtual tests, it directed the robot to prioritize human-dense areas nearly three times faster than the widely used Glasius bio-inspired neural network (GBNN). While it was slightly slower in covering the entire area, the trade-off was negligible compared to its primary goal: quickly reducing mosquito exposure where humans are most at risk. Statistical analysis confirmed the improvement with 99 % confidence.
Field trials brought these results into practice. Using the Robot Operating System (ROS), Dragonfly executed HFA-CCPP-generated routes with strong path accuracy and docking reliability. The robot covered human-dense areas 32.2 % faster than with GBNN, proving the algorithm’s practical value. Most importantly, it ensured that limited battery life was spent protecting people first, making it ideal for use in public spaces like parks or large facilities.
Looking Ahead
This research shows how aligning robotic navigation with human activity can dramatically improve mosquito-control outcomes. By prioritizing zones where mosquitoes and people intersect, the system achieved faster, more effective coverage than conventional methods—an essential advantage for disease prevention.
Future work will focus on integrating real-time human tracking and environmental sensors to make the approach even more adaptive. Beyond mosquito control, the same strategy could inform broader applications in public health and environmental monitoring.
Journal Reference
Wan, A.Y.S., Veerajagadheswar, P., Elara, M.R. et al. Human activity-aware coverage path planning for robot-based mosquito control. Sci Rep 15, 31009 (2025). DOI:10.1038/s41598-025-16114-1
https://www.nature.com/articles/s41598-025-16114-1
Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.