By Soham NandiReviewed by Lexie CornerApr 23 2025In a recent article published in British Ecological Society, researchers introduced WildWing, an open-source drone system developed to autonomously track and film group-living animals in natural environments.
Image Credit: LMIMAGES/Shutterstock.com
Designed to address the limitations of manual piloting, the system enables standardized, high-quality 4K video collection at optimized angles and distances. Field trials with zebras, giraffes, and horses demonstrated WildWing’s effectiveness for behavioral data collection.
With a total cost of around $650, it offers an accessible, scalable solution for wildlife monitoring and computer vision-based ecological research.
Background
Drones have significantly expanded the capabilities of behavioral ecology by offering aerial viewpoints that are difficult to achieve with ground-based observation. They allow researchers to monitor entire animal groups in real-time, capturing complex interactions across varied and often challenging terrain.
Although drones have been used successfully to study species from zebras to elephants, current workflows typically depend on manual piloting. This introduces several limitations, including inconsistent camera angles, difficulty tracking multiple animals, and challenges adapting to fast-changing field conditions. As a result, data quality often varies, limiting its usefulness for detailed behavioral analysis.
Most existing autonomous drone systems are optimized for single targets or animals fitted with tracking tags. These approaches are not well suited to studying untagged wildlife groups in natural settings, where group dynamics, habitat structure, and species-specific behaviors require more adaptive tools.
To address this gap, the WildWing system was developed as a low-cost, open-source autonomous drone platform. It automates aerial data collection while optimizing flight paths and camera positioning to reduce disturbance. By integrating ecological insight with autonomous control, WildWing provides a scalable, consistent method for gathering high-quality footage of group-living animals in the wild.
System Design and Workflow
WildWing has three main components: a Parrot Anafi drone, custom control software, and a GPU-equipped laptop. The system uses SoftwarePilot to interface with the drone, enabling onboard video processing and real-time autonomous navigation. The drone captures 4K video at 30 frames per second and uses incremental adjustments to maintain optimal filming distance without disturbing the animals.
Animal detection and tracking are performed using the YOLOv5 computer vision algorithm. The system identifies herd centroids to guide drone movements and includes customizable settings to adapt to different species and habitats. Its modular design allows integration of updated models and hardware, making it adaptable for various ecological research needs.
Field Validation
WildWing was tested at The Wilds conservation center in Ohio, where it autonomously tracked Grevy’s zebras, giraffes, and Przewalski’s horses. After initial manual flights to evaluate animal sensitivity, the system collected behavioral footage while adhering to FAA guidelines. Zebras and horses were tracked as cohesive herds, while giraffes were monitored as pairs.
Key metrics included approach time (duration until first detection) and usable frames (clear behavioral footage). Przewalski's horses exhibited longer approach times, primarily due to greater initial distances from the drone. However, once tracking was established, the system maintained nearly 100 % usable frames across all species, demonstrating consistent video quality during active monitoring.
The drone also recorded detailed telemetry data, including position, altitude, and command logs, which enabled precise post-flight analysis. Supplementary scripts allowed researchers to calculate metrics such as flight speed and heading, and to reconstruct full mission trajectories. These results confirm that WildWing performs reliably in semi-wild environments and supports the standardized collection of high-resolution aerial data for behavioral studies.
Discussion and Future Directions
The WildWing system demonstrated strong performance during field trials at The Wilds conservation center. It successfully tracked Grevy’s zebras, giraffes, and Przewalski’s horses with minimal disruption to the animals.
However, several factors affected system performance. Detection accuracy varied by species, with YOLOv5 models performing less reliably for less-represented species like Przewalski’s horses.
Environmental conditions, including high temperatures, also influenced battery life, while complex habitats introduced challenges not present in open pasture environments, where the system performed most reliably. Incorporating obstacle avoidance would be necessary for deployment in more vegetated or uneven terrain.
Future directions include integrating swarm robotics for multi-view tracking, improving computer vision models with expanded training datasets, and developing quieter drones to minimize disturbance. Edge computing and integration with sensor networks (such as camera traps and global positioning system (GPS) tags) could enhance ecosystem-wide monitoring.
While WildWing shows strong potential for standardized wildlife data collection, further refinements are needed to adapt to diverse habitats, species behaviors, and environmental challenges—key steps toward scalable, autonomous ecological research.
Download your PDF copy now!
Conclusion
WildWing offers an accessible, autonomous tool for collecting high-resolution behavioral data on group-living animals. Its open-source framework and low cost make it suitable for a wide range of research settings, while its modular design supports ongoing improvements.
By automating aerial tracking and minimizing disturbance, WildWing addresses key barriers in behavioral ecology and conservation science. Continued development in computer vision, robotics, and sensor integration will further expand its applicability to diverse habitats and species, supporting scalable and standardized ecological research.
Journal Reference
Kline, J., Zhong, A., Irizarry, K., Stewart, C. V., Stewart, C., Rubenstein, D. I., & Berger‐Wolf, T. (2025). WildWing: An open‐source, autonomous, and affordable UAS for animal behaviour video monitoring. Methods in Ecology and Evolution. DOI:10.1111/2041-210x.70018. https://besjournals.onlinelibrary.wiley.com/doi/10.1111/2041-210X.70018; https://news.osu.edu/how-a-new-drone-system-may-transform-next-gen-ecology-research/
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.