An alum of the North Carolina School of Science and Mathematics and NC State is developing an iPhone-based app that uses 3D scanning, computer vision, and machine learning to help increase cows’ resilience to injury, illness, and environmental factors.
Catherine McVey’s Graham-based startup OsRostrum was named one of seven recipients of NC IDEA’s signature $50K SEED grants in May. [We previously featured SEED winners Chapel Hill-based GameFlo and Raleigh-based Academic Insight Lab/Moxie from the same grant cycle.] OsRostrum was inspired by McVey’s research while at NCSSM and Colorado State on using facial structures as indicators of hormonal balance in cows.
Her concurrent work at the University of California-Davis in livestock behavioral analytics gave her further inspiration to help increase cow resilience.
“I just realized I wanted to be on the front lines of helping farmers solve solutions,” McVey said. “Being out in California with the fires and the drought and seeing the crazy impacts that it was having out there.”
Farmers will use OsRostrum’s app to scan their cows’ hooves with their phone’s camera. The app will then give farmers a report with one-dimensional biometric information on the cow and tell the farmer how traits are distributed across the herd.
The information in these reports will help farmers make management decisions about their herd. For example, if the report shows that a cow is already struggling with bad ankles or curved claws, the farmer may decide what cows to keep, or what animals need a little extra TLC.
McVey said hoof quality isn’t something that is currently routinely measured, but that farmers have expressed a need for improved hoof quality. OsRostrum will equip farmers with the information they need to breed and manage cows with better hooves. Cows with hoof issues might not gain sufficient weight because they have issues accessing the feeding trough, among other potential problems.
“I thought faces would actually be the original product, but I […] heard over and over again [from farmers], ‘that’s really cool, my [cows’] feet suck; fix them, and I’ll pay you,’” said McVey. “It turns out it’s actually really frustrating when [your cows] have Ferrari genetics and flat tires.”
The information will also come at an affordable price by minimizing upfront investments in hardware and compute costs. McVey also plans to pursue cost-offsetting strategies, like making the app free to farmers by generating revenue through partnering with breed associations. Breed associations provide genetic evaluations for farmers’ herds, usually at a price of $15 to $50 per cow.
The breed associations keep track of and add new traits to the evaluations to improve their genetic evaluations, but are struggling to collect detailed hoof phenotypes, which McVey’s market research showed is a current point of emphasis for many farmers.
One of the business models that OsRostrum is exploring is to charge breed associations for the data that farmers generate from the scans, so that the breed associations can include these hoof traits in their evaluations. This model would enable McVey to keep OsRostrum free for farmers. McVey said “this is one of the places where I will fall on my sword,” to ensure that her product doesn’t contribute to the gap between small and large farms.
“I think diversity is the best thing we can have in agriculture,” said McVey. “So, just about every step of our algorithm pipeline, there’s been places where we’ve gone the extra mile to make sure that it works just as well on a 20-cow family farm as a 20,000-cow giant dairy out in Texas.”
Beta coming soon
McVey plans to launch OsRostrum’s beta—which will be free to beta testers—at the end of June. The team is aiming for Q3 of 2024 to officially launch the hoof modules. This would put her ahead of any competition, as McVey believes OsRostrum is the only domestic company focusing on the genetics phenomic side of hooves.
Before launch, McVey said she needs to focus on tweaking some of the computer vision functionalities and tying up loose ends on the front end of the iOS app and on cloud architecture on the back end. She plans to present at conference circuits over the summer so that OsRostrum will be ready to launch for calving season in the fall.
She has funded OsRostrum by bootstrapping and through a startup fellowship she earned after completing her PhD. She has another year of the fellowship funding, and plans to address further funding after OsRostrum’s first commercialization.
McVey is optimistic about OsRostrum’s launch, as several breed associations and farmers have expressed interest.
“A handful of associations were definitely looking forward to trying it out with some of their members this summer,” said McVey. “I’ve been working with farmers around here, and so far it’s been running pretty slick, and they seem impressed with it.”
