From Code to Coop

Walking into a poultry house for the first time, you’ll likely see litter, eggs and a plethora of hens and roosters strutting about. What you probably wouldn’t expect to see is a 4-foot-tall humanoid robot with a blue luminescent face moving through the chickens, offering a helping hand.

Researchers at NC State University have zeroed in on a future where this could become an everyday reality. Ramesh Bahadur Bist, assistant professor of biological and agricultural engineering, and his Artificial Intelligence & Robotics (AIR) Lab are developing next-generation solutions to address labor shortages and improve animal welfare, safety and productivity in the poultry industry.

From autonomous egg-collecting robots to intelligent systems that can assess the health of individual birds, Bist’s AIR Lab is cracking into AI-driven farming to create cutting-edge tools that can one day help producers better care for and manage commercial flocks.

“The poultry industry is facing a labor shortage, and, consequently, robotics and AI can help with daily tasks,” says Edgar Oviedo, a poultry science professor who is working with Bist on several AI-related projects.

It’s an effort aimed at finding synergies to support poultry producers and an industry that, in North Carolina alone, generates approximately $40 billion in economic impact annually (roughly 10% of the nation’s poultry supply).

“Agriculture will always be deeply human-centered, but the farms of the future will see humanoid robots working alongside farmers, bringing human-like intelligence and support to everyday farm tasks,” Bist says.

Modernizing Poultry Production

Since joining NC State in the summer of 2025, Bist has spent many hours working in the university’s Chicken Education Unit to pursue this future-focused vision. He, along with three graduate students and one undergraduate student, make up the AIR Lab team.

“Our lab provides interdisciplinary training that integrates engineering, robotics, artificial intelligence and animal systems to build smarter, more sustainable farms,” Bist says.

The graduate students working in the AIR Lab describe the experience as both technically rigorous and purpose-driven. Stephen Afrifa, a Ph.D. student specializing in AI and big data, emphasizes that the lab’s work extends beyond academic research to find real-world applications that make a difference.

“We are building intelligent systems that can directly influence animal welfare, productivity and food production at scale,” he says.

A Helping Hand for Cage-free Systems

In recent years, the poultry industry has shifted from caged-house systems to cage-free systems, leading to more hens laying eggs on the floor rather than in nesting boxes. 

“Floor eggs can account for 2% to 15% of total production in certain environments, and collecting these eggs requires time and labor, and delays can affect product quality,” Bist says. “If you’ve got around 100,000 birds, that could mean anywhere from 2,000 to 15,000 eggs ending up on the floor each day. It takes a lot of time and labor to collect them all.”

But often there aren’t enough workers to gather the eggs or tend to other manual tasks.

“Poultry production requires a lot of labor, and even with automation, there are many issues with automatic scales and methods for collecting eggs, handling birds or processing them,” Oviedo says.

To solve that challenge, the AIR Lab recently added a new member to the team — a humanoid robot dubbed “Howl” that’s the same make and model as the viral dancing robots trending on social media earlier this year. Instead of busting a move, though, the team’s goal is for Howl to use AI and robotic manipulation to detect and collect floor eggs, addressing labor shortages and reducing product loss.

Since this is largely uncharted territory, Bist’s students have to start from scratch to code Howl’s movements and response features using virtual reality manipulation along with other machine learning methods like reinforcement learning and imitation learning.

With a few more months of work, the team hopes to have Howl complete egg-detection and retrieval simulations — and maybe even do a little dance to celebrate.

From Data to Decisions

Bist believes the most exciting aspect of integrating AI technologies into agriculture is the ability to transform raw data into actionable results. One of Bist’s ongoing projects, Poultry Talk, works to address data and knowledge gaps in the poultry industry. 

The Retrieval-Augmented Generation (RAG) model acts like a poultry-specific ChatGPT, referencing information from documents and databases when users enter questions. The goal of this AI-driven model is to provide industry experts, researchers, farmers and students with more detailed information on poultry topics such as nutrition, disease diagnosis and prevention, housing, welfare, production management and precision poultry farming. 

“What makes Poultry Talk particularly useful is that it combines a large language model with a curated knowledge base on poultry science, so responses are grounded in open-source extension publications and peer-reviewed research rather than generic internet information,” Bist says. “This helps reduce misinformation and make AI insights more trustworthy for real-world poultry farming decisions.” 

With the demo up and running, Bist has already sent it to professionals around the country for their feedback. 

“It’s a very good tool, and it will be a game changer,” says Yang Tian, a postdoctoral fellow in the Department of Biological and Agricultural Engineering at the University of Arkansas, who has tested the application. 

Tian believes that with more development, topic-based applications for agriculture will be very useful because they can provide “more specific and more accurate” information. 

The Power of Prediction in Poultry

Along with the shift to cage-free environments, there is a growing emphasis within the poultry industry to monitor the well-being of individual birds. 

“Traditionally, poultry management has operated at the flock level, but with AI we can move toward more individualized monitoring, even within large populations,” Bist says. 

In collaboration with Oviedo, Bist is developing a precision poultry monitoring system. A series of mounted 2D and 3D cameras at NC State’s Chicken Education Unit pair with AI models to digitally assess each bird and rate its health based on feather score (the condition and density of plumage) and body weight. The model aims to detect early signs of stress and disease while supporting the development of precision feeding systems. 

“Modern poultry farms generate enormous volumes of data, like behavioral data, environmental readings and production metrics, yet much of it goes underutilized,” Bist says.

He believes harnessing the power of AI will help identify health-related patterns that would be hard for humans to continuously observe across an entire flock. The precision system under development would support individual monitoring of each bird, allowing conditions and diets to be tweaked to promote every bird’s health and welfare. 

“Integrating this data into a usable platform can enable early detection of issues and support more precise and efficient farming practices,” Afrifa says.

The next era of farming will be built by people, strengthened by intelligent machines and driven by a vision to feed the future.

As these research efforts come to fruition, the AIR Lab remains focused on Bist’s goal to break open the advantages of precision agriculture and whisk that into the hard work of traditional farming. 

“The next era of farming will be built by people, strengthened by intelligent machines and driven by a vision to feed the future,” Bist says.

The combination could create a new path for resilient, future-proof food production.