Applying AI in Cattle Operations

AI can be stacked on top of existing technology to create new tools for producers.

Each new technology incorporated into agriculture promises increased yields, reduced labor and improved sustainability. These expectations often clashed with the messy reality of low profit margins that inhibited adoption, sensitive equipment placed in changeable weather conditions and low (or no) cellular service in rural areas.

Combining this existing technology with artificial intelligence (AI) may offer a new path to achieving the goal of feeding a growing global population with fewer resources. According to the World Economic Forum, organizations are incorporating AI into predictions for crop yields, water resources, and risk of pest infestations and drought. Fewer companies are using AI to help individual producers make decisions.

One company that is putting AI-powered technology at the farm level is myAnIML, an animal agtech startup based in Overland Park, Kansas. It is using proprietary facial recognition to evaluate cattle for early signs of illness. This has the potential to identify sick animals faster, which could reduce severity of illness and lower death losses.

Advances in artificial intelligence and equipment are making results more sophisticated and user friendly. Yet, the agricultural industry has unique barrier to adoption.

“My background is very deeply in technology,” said Shekhar Gupta, Founder and CEO of MyAnIML. “What I’ve seen over so many years is that there are many companies that have machine learning processes in place and call it AI. Machine learning is what I feed in comes out the other end. With our technology, we’ve seen it evolve to produce results that I didn’t teach the machine.”

The AI technology in the MyAnIML system is already picking up indications of animals that might be in heat or deliver calves, which is an outcome Gupta and his team did not train the system to do.

Rapid illness identification

Incorporating AI will take money and effort – just like any other technological expansion into a new industry. Ranchers have traditionally had a small profit margin and are skeptical of new investments not yet proven to pay off.

“With these small profit margins, it’s important to demonstrate how to save or generate revenue,” Gupta explained. “If a cow gets sick more than once, that profit margin goes away. It goes into providing that treatment, the labor cost for bringing in the cattle and lost productivity.”

Additional benefits to earlier illness identification could include isolating sick animals before disease is spread, which can help keep the rest of the herd healthy. Plus, fewer animals would need to be treated with antibiotics, thereby reducing costs for the producer.

MyAnIML designed a camera that could be placed at high traffic areas like milking parlors, feed bunks or waterers. Images are analyzed based on beads and ridges in the animal’s muzzle that indicate illness. Cattle are identified by taking a picture of their existing ear tag, and a report is sent to the producer.

In a collaborative study with the USDA, the MyAnIML technology accurately predicted Infectious Bovine Keratoconjunctivitis (IBK), or bovine pinkeye, 99.4% of the time. In addition, disease was predicted several days before veterinarians were able to detect symptoms.

Breed differences have not changed the results, but Gupta found the age of the animal is a significant difference.

“If you think about it, muzzles are like our fingerprints,” he said. “A child’s palm lines are not as clearly visible as they are in an adult palm. For calves 700 pounds or less, the muzzle is not very defined as compared to mature cows.”

Learning from the recent past

Gupta spent years honing the best technological combinations to achieve this goal.

“When I started my company, we looked at drones to see if we can grab a picture of the cow’s face,” he recalled. “It had to be a professional grade drone, and someone had to be capable of launching it and meeting FAA (Federal Aviation Administration) requirements. Plus, cows tend to run away when there’s a drone flying around, which made it hard to get a picture of their face.”

The lack of electricity around common cattle gathering areas also was an obstacle particular to agriculture. The current system solves that problem by using solar power. The equipment also needed to be resistant to curious animals that could damage expensive investments.

“Cows are curious animals. Once, I was installing a camera, and a cow came in and started chewing on my arm because she was curious. With the high cost of initial investment, we have to consider the cost and ongoing maintenance – some of those are potential drawbacks in any industry.”

References:

1 University of Chicago (2024). What is (AI) Artificial Intelligence? Accessed Aug. 30, 2024. Available at: meng.uic.edu/news-stories/ai-artificial-intelligence-what-is-the-definition-of-ai-and-how-does-ai-work

2 Markets and Markets. (2023) Artificial Intelligence in Agriculture Market by Technology. Accessed Aug. 30, 2024. Available at: marketsandmarkets.com/Market-Reports/ai-in-agriculture-market-159957009.html

Key Points:

• Artificial intelligence (AI) represents a branch of computer science that aims to create machines capable of performing tasks that typically require human intelligence.1
• AI subsets include machine learning, where machines learn patterns and can make decisions based on data.1
• The AI in Agriculture Market is projected to grow from $1.7 billion in 2023 to $4.7 billion by 2028.2