MPCS Students Help Revolutionize the Ranching Industry with Computer Vision
By Katie Knorr
Computer vision technology has potent applications in the ranching industry. “How?” you might ask. Meet OneCup AI, a small software company doing big things for livestock farmers.
OneCup AI has developed BETSY, an artificial intelligence (AI) technology that watches cattle around the clock to track health and growth indicators as well as activity and nutrition levels. Powered by AI and computer vision technology, BETSY can recognize individual animals based on their unique features.
The system receives input through strategically placed cameras, so if an animal is limping, running, feeding, or socializing, BETSY knows. From all this data, BETSY creates a historical record for each animal. The system can also collect data to provide a macro-level overview of a herd’s health.
The impact of this technology is powerful: eliminating the need for ranchers to be everywhere all the time, the system gathers and analyzes rich data to help ranchers manage their livestock and address problems before they become more serious. In other words, BETSY helps ranchers maximize animal welfare while optimizing yield.
Students and graduates of Simon Fraser University’s Master’s in Professional Computer Science (MPCS) have been instrumental in making BETSY come to life.
Having hired MPCS graduates and co-op students for over two years, Mokah Shmigelsky, CEO of OneCup AI, recognizes the impact they have had on the development of BETSY.
“From adding state-of-the-art models, suggestions and output, each student has had an impact on OneCup,” she says. “Their ability to solve problems and come up with innovative solutions has proved essential to move forward our technology in a sustainable way.”
It was her husband Geoff Shmigelsky, CTO of OneCup AI, who discovered the MPCS program as a way to recruit experienced students to help the company build its AI system. “He had a pulse on all academic institutions in the field of visual computing, and SFU was local to our company as well,” says Mokah Shmigelsky.
Today, more than one third of OneCup AI’s team is MPCS powered.
The company initially applied for MPCS talent through Mitacs, a non-profit national research organization that connects Canadian academia, private industry and government by providing funding for innovative projects and internships.
“We saw success with Mitacs at the time and the MPCS program already had an application process in place which made our experience smooth,” says Mokah Shmigelsky. “We were very pleased with the ability to set up interviews through the portal and the quality of all the students we have met throughout the past two years.”
Urvi Chauhan was one such student. She joined OneCup AI as a co-op student in the fall of 2020 and was responsible for creating pipelines to drive BETSY’s development through a path of building, testing, and deploying code, also known as continuous integration and continuous deployment (CI/CD). Chauhan, now a full-time data scientist at OneCup AI, works on training the BETSY system to identify body parts of cattle with a high accuracy rate.
Having seen how ranchers respond to BETSY, Chauhan is confident that the technology will soon become ubiquitous in the ranching industry: “I had a chance to meet ranchers at a couple of events that I attended with the company,” she says. “They were amazed at what BETSY can do and were excited to adopt this technology on their farms.”
Philip Cho, another MPCS graduate, completed a co-op work term at OneCup AI in the summer of 2021 and is now the company’s lead computer vision engineer. He is in charge of researching, training, and integrating the latest computer vision algorithms into the BETSY system.
“There are many things that I like about working at OneCup AI, but I guess my favorite aspect is that I get to research and test out exciting state-of-the-art computer vision technologies,” he says. “It’s something that I’ve always wanted to do, even as a student.”
Founded only three years ago, OneCup AI is just starting its commercialization process. Mokah Shmigelsky already knows that, as the company expands, she wants to continue to strengthen its relationship with the MPCS program.
“We are very impressed with the calibre of ideas and processes offered by the students and alumni of the program, and we hope to continue utilizing their talent and expertise to create the innovative solutions that our customers and researchers are looking for.”