MyAnIML’s autonomous edge monitoring system is allowing ranchers to go (or stay) off the grid — addressing a primary barrier preventing them from adopting advanced agricultural technology, explained founder Shekhar Gupta.
The Overland Park-based agtech startup — whose patented technology uses artificial intelligence and machine learning to predict a broad range of total health indicators through analyzing data from cow muzzles — recently announced the commercial availability of the industry’s first AI-powered cattle health platform that requires no WiFi or electrical infrastructure.
MyAnIML’s autonomous edge monitoring system; photo courtesy of Shekhar Gupta, MyAnIML
While traditional AI systems depend on cloud connectivity and fixed power, MyAnIML’s edge-native device performs all complex analysis directly on the hardware, noted Gupta, CEO of MyAnIML, one of Startland News’ Kansas City Startups to Watch in 2022.
“It cuts down on all the friction points that the ranchers have,” he continued. “‘I don’t have power running to my pens; I don’t have WiFi in my ranch.’ All of those friction points are gone.”
“We’re delivering a solution built for how ranches actually operate,” he added, “allowing for targeted treatment rather than the mass use of antibiotics.”
The autonomous edge monitoring system delivers stress-free monitoring with 48-hour early warnings, operates on solar power, and maintains reliability in mud, dust, and extreme weather conditions, he said. It also can be mounted anywhere cattle naturally gather.
“We tested it in the very strong wintertime — in January and February — and then in the summertime — in July and August — in the heat and humidity to make sure that the product works,” he explained. “Because if it’s going to fail, it’s going to fail either in the very, very high wind, extreme cold, and snow or in the summertime when the humidity is so high. But it worked really great.”
Using proprietary deep-learning models trained on more than 700GB of bovine biometric data collected over three years, Gupta said, the My AnIML system has demonstrated USDA-validated accuracy of 99.8 percent in detecting pinkeye and up to 70 percent accuracy in identifying Bovine Respiratory Disease (BRD), the costliest illness in U.S. feedlots.
“This is one thing that I’m very proud of,” he continued, “is that we have over 700 gig of data working in the industry, working with the ranchers. So this is the data that you can’t get on the internet anywhere.”
“Our system is very robust,” Gupta added, noting the company currently monitors about 10,000 head of cattle monthly across Kansas operations and expects to exceed 50,000 head monitored within the next two months. “It is a vertical AI. There is nobody who’s building a vertical AI in cattle health, so we’re literally dominating that area.”
Although the system is trained to identify pinkeye and BRD, he said, ranchers have noticed it detecting nutritional stress and reproductive status.
“It’s actually looking out for further things that we haven’t even trained it in,” Gupta explained, adding that he is working on a peer review paper about this emergent discovery. “It’s training itself in the field and then providing notifications based on that and updating our algorithms. So that’s what is so interesting about this AI part.”
