DJI Neo Drones and AI Team Up to Track Turkey Health

Farmers spend a lot of time walking through barns to check on their birds. They look for stress, sickness, injuries, and anything unusual. This work is essential, but it also takes a lot of hours and a lot of staff. Poultry barns can hold thousands of animals, which makes close monitoring difficult. High staff turnover makes things even tougher.

A new study from Penn State University tested a simple question. Can small drones with a camera, like the DJI Neo, combined with artificial intelligence, watch turkeys from above and identify what they are doing on its own? The goal was to reduce manual labor and give farmers a better way to track animal welfare all day without disturbing the flock.

As reported by Phys.org, Researchers used a DJI Neo (one of my favorite drones) a very small drone that weighs only 135 grams. It carries a 4K camera and can fly very safely indoors thanks to his building with integrated protectors.

The team flew the drone over 160 young turkeys from day 5 to day 32 of age. They collected video every day and used that footage to train an AI model to recognize eight key turkey behaviors.

Those behaviors were feeding, drinking, sitting, standing, perching, huddling, wing flapping, and dead. The idea was simple. If the drone can see these behaviors, farmers can spot problems early.

The team used the YOLO computer vision model. YOLO is one of the most popular AI tools for detecting objects and actions in images. They created a dataset of 2388 labeled images and divided it into training, validation, and testing groups. After testing different setups, they found that YOLOv11 large with a 1280 pixel resolution worked the best.

The results were impressive for an early test. The model reached a precision of 0.90 and a recall of 0.87. The final accuracy score, called mAP50, was 0.89. In simple terms, the AI was able to identify turkey behaviors with a high level of confidence.

Drones Inside Barns

Using drones in poultry barns can sound risky, but the study showed that it is safe when done correctly. To avoid stressing the turkeys, the research team played recorded drone sound through a speaker inside the barn for several days before the flights. This helped the birds get used to the noise.

Once flights started, the drone operated at a height of two meters. It moved slowly, covering the full pen during each one minute flight. The camera pointed downward at a 45 degree angle. This gave a wide field of view and made it easier for the AI to see the shape and posture of each turkey.

The drone flew four times a day. Two flights were in the morning and two flights were in the afternoon. Between flights, the drone landed for five minutes. A trained UAV pilot controlled each flight and another person watched for safety.

An action camera inside the pen recorded the turkeys the entire time. This allowed the team to check if the drone caused stress. They found no signs of negative reactions. The turkeys behaved the same with or without the drone overhead. Previous studies in broiler houses also support the idea that small drones do not cause meaningful stress when introduced correctly.

What This Technology Could Mean for Farmers

If drones and AI can monitor bird behavior accurately, farmers could get real time insights into flock health. This includes feed intake, water use, activity levels, and early signs of sickness. It also reduces the need for frequent staff walkthroughs, which lowers labor costs and training time.

Another benefit is coverage. A drone can capture the entire barn in one flight. Fixed security cameras can only see one area at a time. Drones can also adjust their height or angle to get a better view. This flexibility is valuable in large barns with thousands of animals.

The study points to a future where drones fly on schedule, collect video, and AI models instantly analyze behavior patterns. Farmers could receive alerts on their phones if birds are huddling too much, not eating, or showing signs of heat stress. These alerts would help catch problems before they spread.

This is the first study to use a drone combined with a YOLO model to track turkey behavior. The results suggest that commercial adoption could come soon. More testing is needed with older birds, larger barns, and full commercial conditions. Still, the early results are promising.

DroneXL’s Take

Drones are already transforming farming, and now they may change how poultry barns are monitored. A tiny drone like the DJI Neo can gather data that once required hours of walking. When paired with AI, it becomes a powerful welfare tool. The biggest win here is early detection. If farmers can see problems before they grow, they can save birds and reduce costs. This study shows that drones inside barns are not just possible, they are practical. The technology is young, but the potential is huge.

Photo credit: Sciencedirect.com, Wikimedia, Rafael Suarez