Precision Livestock/Poultry Farming with Machine Vision Systems



Program 1: Machine Vision for Monitoring Broiler Behaviors and Welfare 

United States is currently the world’s largest broiler producer and the second largest egg producer (e.g., poultry and eggs had a sale value of $40 billion in 2021), but poultry and egg productions are facing grand challenges of animal welfare concerns, food safety issues, and environmental impacts.

The rapid growth rate of broilers is associated with welfare concerns such as leg issues and lameness. Broilers with lameness suffer behavior restrictions, physical discomforts, and impingement of fundamental freedoms. Those welfare concerns have triggered the attention of the general public and the food industry to improve broiler well-being, and well-being evaluation. Animal welfare evaluation is currently performed manually by farm workers daily or occasionally in the poultry houses, which is time consuming, labor intensive, and subject to human errors. This task calls for the design of an automated system that can monitor poultry welfare automatically. 

Sensing technologies, such as ultra-wideband, radio frequency identification, accelerometer, and computer vision-based monitoring, have been and are being adapted and tested for livestock and poultry farming systems to aide well-being evaluation. Computer vision-based phenotyping technologies have been tested efficient in monitoring large animals such as cattle and pigs. However, it is technically challenging to monitor smaller animals such as broiler and layer chickens.

Dr. Lilong Chai’s team in the Department of Poultry Science at the University of Georgia is developing specific imaging and phenotyping technologies (e.g., deep learning models) for monitoring/tracking floor distribution patterns of broiler chickens and individual birds’ moving in different zones of feeding, drinking, and resting. Those imaging and phenotyping technologies will be further innovated to evaluate poultry welfare indicators in commercial broiler houses.

Broiler Floor Distribution
Deep Learning/Phenotyping Technologies for Tracking Individual Chickens.
Tracking Individual Chickens
Deep Learning/Phenotyping Technologies for Tracking Individual Chickens.

Program 2: Precision Management for Cage-free Egg Production

The primary fast food chains and big box grocers (e.g., McDonalds and Walmart) have pledged to source only cage-free eggs by the year 2025. According to the pledges, the market share of cage-free eggs will be increased from 24.5% in December 2020 to 72% by 2025.

In addition, seven states have passed regulations to source only cage-free or free-range eggs by 2025 or earlier. However, cage-free production is not without its own challenges such as high mortality, pecking, and floor eggs.

Dr. Lilong Chai’s team in the Department of Poultry Science at the University of Georgia has been working on the precision management practices for addressing the issues related to heat stress, pecking, and floor eggs in the cage-free system.

Pecking Behaviors
Pecking Behaviors Tracking and Prevention with Machine Vision/Deep Learning.
Thermal Imaging
Thermal Imaging Technology for Animal Heat Stress Evaluation and Management.


Lilong Chai Assistant Professor
Poultry Science