AI-supported drone monitoring of meadow birds: from field-tested system to wider use
An important practical milestone has been reached in the Nestdrone project. From the user’s perspective, the complete system now works as intended: from detecting a possible meadow bird nest in the field to recording its precise location in the BFVW registration environment.
The project team reports that more than 900 nests have been detected in 2026, with over 25 active users. The system is already being used by field practitioners and researchers in the Netherlands and internationally.
What does meadow bird monitoring with drones involve?
Meadow bird nests can be difficult to find because they are well hidden in grassland. Searching large areas on foot takes time and may cause unnecessary disturbance. Drone monitoring helps volunteers cover fields efficiently and focus their attention on locations where a nest may be present.
During a flight, the drone follows a planned route and collects two types of images:
Infrared images show differences in temperature and can help reveal the heat signature of a breeding bird or nest among the vegetation.
RGB images provide a normal, detailed view of the field, allowing a possible detection to be checked in its surroundings.
Artificial intelligence analyses both image streams and marks locations that may contain a nest. Combining thermal and visual information is valuable because vegetation, weather, ground temperature and viewing conditions can all influence what can be seen in an individual image.
RTK positioning then connects each detection to highly accurate coordinates. The results are processed through the cloud and transferred directly to the BFVW registration system. A trained volunteer can review the detection, verify it in the field where necessary and use the information to discuss suitable protection measures with the farmer or land manager.
The technology therefore does not replace the experience of farmers and volunteers. It helps them see more, search more efficiently and make field knowledge available more quickly.
The role of BFVW in FARMBIRD
The results from Nestdrone provide a strong starting point for further development within FARMBIRD. In this Interreg North-West Europe project, the Bond Friese VogelWachten—BFVW, or Frisian Bird Watch Association—will lead the drone-based work for meadow bird protection.
BFVW brings together practical field knowledge, an established volunteer network and experience with digital registration. Its monitoring activities cover more than 130,000 hectares in Fryslân, while approximately 40,000 nests are registered and monitored through its system each year. This means that new technology can be tested in an extensive, real-world monitoring network rather than only under controlled research conditions.
Within FARMBIRD, BFVW will:
test and evaluate new drone and AI developments under different field conditions;
connect drone detections with the existing BFVW registration app and future information tools;
collect data across Fryslân, including different landscapes, habitats and soil types;
share its drone, registration and field experience with partners in other North-West European regions;
train volunteers to use the new monitoring technologies safely and effectively;
develop online tools through which volunteers can learn from one another and exchange experiences;
help translate monitoring results into practical information for farmers, conservation groups and agricultural collectives.
BFVW will also contribute to wider pilots on nest and chick survival, risks such as predation and agricultural activities, and the monitoring of habitat conditions. The longer-term ambition is to develop clear and replicable working methods that can be adapted to different regions and species, including meadow birds such as the Black-tailed Godwit, Northern Lapwing and Eurasian Curlew.
The next step: making the system scalable
The focus will now shift from demonstrating that the system works to ensuring that it remains reliable when more people, regions and organisations begin using it.
The main development priorities are:
supporting more users and simultaneous drone operations;
continuing to train the AI with images from different species, landscapes, seasons and weather conditions;
reducing false detections and making results easier for users to review;
strengthening the cloud and ICT infrastructure;
establishing consistent protocols for flying, validation, data quality and registration;
making training and support available to a broader group of volunteers and field professionals.
Scaling up is not simply a technical challenge. The system must remain practical for the people using it and provide information that supports action during the breeding season. By connecting accurate nest locations with the knowledge of farmers, volunteers, researchers and conservationists, the technology can contribute to better-informed decisions in nature management and agriculture.
The system works in the field. The next task is to make it robust, accessible and future-proof, so that more people can use it to protect meadow birds and their habitats.
We Fly Together.
FARMBIRD is co-funded by the European Union through the Interreg North-West Europe programme.