‘Smart-device application’ will make recommendations using data that beef producers collect daily on their operations
An expert who plans to use artificial intelligence to help create a mobile app for Alberta’s beef producers says being a successful farmer is like being a good doctor.
“To be a good doctor, to be a good farmer, you have to be able to take in a whole bunch of information, use a lot of experience, and make some good guesses,” said David Wishart, a professor in the Faculty of Science at the University of Alberta.
“And you have to do that consistently if you want to be a profitable farmer or a successful doctor.”
Wishart is collaborating with a team led by Graham Plastow, a fellow professor at the university and chief executive officer of Livestock Gentec, to create the Arm-Chair Ranching app.
In addition to providing information on their smartphones that would otherwise be hard for beef producers to find, the three-year project will use smart technologies to help them better manage their operations.
“The smart-device application will use the reams of data that beef producers collect daily on genetics, feedlots and economics to make optimal herd and animal management recommendations,” said a statement by Alberta Innovates, which is the province’s largest research and innovation agency.
The project received $481,000 in funding from the agency’s Smart Agriculture and Food Digitization and Automation Challenge program.
“(Much) of this data is online, but you have to really know where it is,” said Wishart, who is a professor in the departments of computing science and biological sciences. “And then plus, it has some machine learning on the data so that you can make predictions that farmers and ranchers can use.”
Machine learning is a branch of artificial intelligence in which computers mimic how humans learn, he said.
It could help identify trends in everything from the weather to beef prices, making recommendations to help producers make the best decisions.
“A rancher could say, ‘was this a really dry spring or not?’ We have historic data for just about every locality in Alberta mapped to positions that people can say, ‘yeah, it was a dry spring, but compared to 2012, this is actually pretty wet, or relative to the last 10 years, it’s average,’ and that’s something that might be useful.”
The app could also include current data to allow beef producers to better judge factors such as prices for feed, fertilizer and oil, as well as market futures and the existing capacity at different feedlots.
It could even provide information about how a farmer or rancher’s productivity compares to other cattle operations in their area, said Wishart. It might “show you that your farm’s productivity is 1/10 of your neighbor’s, and maybe you’re doing something wrong here.”
Being a beef producer can be difficult, said Wishart.
“Now in some respects, I think farming is a little bit like where you have to take a lot of information and make an assessment: ‘is this the time to sell, is this the time to breed, is this the time to buy or wait another couple of weeks?’”
The ability to make such decisions is something many beef producers likely already possess to some degree, he said.
“Farmers are probably really good at it — they just have an intuition; others may struggle with it.”
Wishart likens the Arm-Chair Ranching app to how weather forecasting can help people make better plans for the weekend. “It’s not perfectly reliable, but at least sometimes it helps make our decisions a little easier.”
Agriculture is one of the last industries to be digitalized, a situation Wishart said was partly due to its complexity.
“But you know, so is stock trading, and stock traders in the stock market have used data and machine learning to help them for many years.”
Although every stock transaction from the past few decades has been digitally recorded, a fair chunk of agricultural knowledge still consists of what’s in farmers’ heads or in books, he said.
“And without electronic records, it actually makes it really difficult to do or apply machine learning and artificial intelligence, so that’s why a lot of our work right now is really trying to chase down some of that information to compile it,” said Wishart.
“It’s more arduous than we thought, but it’s something that we knew was going to be the challenge, and that’s one of the reasons why farming has been one of the last frontiers.”
Although the Arm-Chair Ranching app is aimed at Alberta’s beef producers, it could eventually be expanded to other provinces as well as different types of farming, he said.