As the holiday season approaches, many of us eagerly anticipate a hearty turkey dinner, complete with all the festive trimmings. But here’s the unsettling truth: rising costs and labor shortages are making it increasingly difficult for turkey farmers to keep this beloved tradition affordable for families. Enter a groundbreaking study from Pennsylvania State University, where researchers are leveraging drones and artificial intelligence (AI) to revolutionize turkey farming—not only cutting costs for farmers but also improving the health and well-being of the birds themselves. And this is the part most people miss: this technology could be the key to meeting the growing global demand for animal protein without simply raising more animals.
Enrico Casella, a Penn State assistant professor of data science for animal systems, emphasizes the urgency of this innovation. “We need to focus on raising the animals we already have more efficiently and productively,” he explains. “That’s where this work fits in.” Casella leads the drone-enabled monitoring study, which uses small drones equipped with 360-degree cameras to observe turkey behavior inside poultry houses. These drones capture footage of hundreds of birds engaging in activities like feeding, drinking, perching, and even wing flapping. The video data is then fed into an AI program called YOLO (You Only Look Once) for analysis.
But here’s where it gets controversial: while the drones are designed for social media and selfies, their use in farming raises questions about animal welfare and the ethical implications of technology in agriculture. Casella notes that the team initially worried the drones might stress the young turkeys. However, they discovered that piloting the drones in a smooth, cinematic mode minimized disturbance. “It’s not the drone’s appearance that bothers them, but the sound and wind from the propellers,” he explains. This finding highlights the importance of refining technology to ensure it benefits both farmers and animals.
The study tracked eight common turkey behaviors, annotating over 19,000 individual activities to train the YOLO model. One critical data point was mortality rates among young birds, which, if detected early, can prevent the spread of pathogens. The YOLO model’s simplicity made it accessible for the interdisciplinary team, many of whom are animal scientists learning to code. Casella believes that with more computational expertise, even more efficient models could be developed, potentially incorporating historical data on individual birds’ behaviors.
After successful trials at the Penn State research farm, the team tested the technology in a large commercial poultry house, with promising results. “The animals adapted even better than we expected,” Casella says. Now, the team is experimenting with larger drones equipped with thermal cameras, though they face challenges like indoor GPS signal loss. These advancements could open new possibilities for studying bird behavior in greater detail.
But here’s the bigger question: Can this technology truly address the agricultural labor shortage, which sees a 60% annual turnover rate? Casella argues that drones and AI can free workers from repetitive tasks like monitoring flocks, which traditionally require hours of walking through massive poultry houses. “This technology allows for more frequent and efficient checks, freeing up time for other critical tasks,” he says. By improving monitoring, it also enhances animal welfare, ensuring healthier and happier birds.
As we look to the future, this fusion of drones and AI in agriculture raises thought-provoking questions. Could this be the sustainable solution to feeding a growing population? Or does it risk dehumanizing an already struggling industry? We’d love to hear your thoughts—share your opinions in the comments below!
