From Drone Footage to Population Estimates: Closing the Gap in Wildlife Monitoring

In a study led by Chaim Chai Elchik, we tackle a key challenge in conservation: how to turn drone data into reliable estimates of wildlife populations. While drones are now widely used to collect imagery, converting that data into meaningful metrics like density and abundance has remained a bottleneck. Our work focuses on linking drone surveys directly to ecological analysis, making monitoring more practical at scale. We developed a workflow that combines drone imagery, onboard telemetry, and AI-based detection and tracking. This allows us to automatically identify animals in video and assign them precise locations on the ground. Tests showed that we can locate animals with metre-level accuracy, which is a step forward compared to traditional aerial surveys. A key advance is that our system doesn’t just detect animals—it also calculates their perpendicular distance to survey transects. This means the outputs can be directly used for distance sampling, one of the most widely used methods for estimating population size. By generating these ready-to-use data automatically, we remove a major processing step that often slows down conservation projects. Overall, this approach helps move wildlife monitoring towards fully automated, scalable workflows. It allows conservation teams to go from drone flights to population estimates much faster, while maintaining accuracy. As drone and AI technologies continue to improve, this kind of integration will be key to monitoring wildlife more effectively across large and complex landscapes.

Key takeaway: Drone workflows can now automatically deliver both animal locations and distance sampling inputs, making large-scale population estimation faster and more practical.

Citation:
Elchik, C. C., Hudson, M., Buckland, S., Longmore, S., & Wich, S. (2026). Using drones for large mammal monitoring. Mammal Review, 56, e70044. https://doi.org/10.1111/mam.70044