SUMMARY: The “One Health” approach emphasizes the link between animal, human, and environmental health. At Purdue’s Department of Animal Sciences, researchers are developing automated technologies to detect livestock illness through subtle behavioral changes before symptoms appear. Computer vision with pen cameras monitors ear posture, facial expression, and social isolation in dairy calves. These tools can detect illness early, support farmer decision-making, and prevent disease spread, improving livestock health, welfare, and biosecurity. By enabling earlier intervention, they may reduce treatment costs, minimize production losses, enhance productivity, protect human health, and reduce antibiotic use in livestock operations. 

Purdue One Health takes a holistic approach to tackling global challenges that impact human, livestock, crop and environmental health through academic programs, research, facility infrastructure and industry partnerships.  Purdue’s One Health transdisciplinary research is led by faculty and scientists from academic units and research facilities across campus, and with stakeholders in industry and agricultural production systems. 

Livestock health is paramount to livestock productivity.  When livestock become sick, their poor health can jeopardize the output of the entire herd.  Often, by the time symptoms manifest, the disease is already spreading.  Disease vectors can also spread into the environment and affect human health and beyond.  For these reasons, One Health is an important strategy to address both food safety and agricultural productivity. 

Livestock behavior can tell us about how animals are feeling. Changes in behavior before clinical signs of illness, referred to as sickness behavior, can be captured and analyzed utilizing technological tools such as ML and AI.  Scientists at Purdue University, in collaboration with other universities, are exploring subtle behavioral changes that could be early indicators of illnesses such as diarrhea and bovine respiratory diseases that are highly prevalent in dairy calves.  

 On-farm technologies are being tested for their ability to detect: 

• decreases in feeding behavior –  a common sickness behavior shown in many species;  
• changes in social behavior – such as avoiding interactions by isolating from the rest of the herd;  
• changes in body language – such as ear position and movement; and
• reduced pleasurable activities – such as grooming, a natural behavior expressed by cattle to scratch and maintain hygiene. 

When animals are falling ill but not yet showing outward clinical symptoms, these behavioral expressions are expected to change. Cameras can monitor and detect these subtle behavior changes better than humans, especially when monitoring large herds. Cameras are versatile, with the possibility to monitor a wide range of behaviors of many animals on farms. Detecting these early sickness behaviors could lead to early non-antimicrobial treatment or nutraceutical support to limit severity or prevent illness.  

Computer vision and AI can transform how livestock health is monitored, enabling earlier detection of behavioral changes that indicate emerging illness. By integrating these innovative tools within One Health frameworks, we can enhance animal health, productivity, and welfare by reducing disease spread and antimicrobial use. Ultimately, these approaches strengthen the resilience and sustainability of livestock production systems, supporting food security and protecting human and environmental health. 

Researchers: Heather W. Neave & Gary R. Burniske 

Return to the 2025 GAP ReportTM Partner Stories