An Ontario start-up is working on a vision system technology its founders say will ultimately help apple growers produce more food.
Vivid Machines is developing the Vivid X system that will automate predicting and managing apple yield and quality, and eventually provide early detection of pests, diseases, and nutrient deficiencies.
Why it matters: Precisely monitoring high value fruit can help head off challenges before they become a problem.
Co-founders Jenny Lemieux, who grew up on a farm in Ilderton and holds a Masters in Artificial Intelligence Management, and Jonathan Binas, a physicist with a PhD in brain-inspired computing, first met when they were participants in a program called Entrepreneur First.
With their joint interests in information technology, machine learning and food production, developing technology solutions for agriculture seemed a natural fit.
After conversations with approximately 200 growers, agronomists, staff from the Ontario Ministry of Agriculture, Food and Rural Affairs and Vineland Research and Innovation Centre, Lemieux and Binas learned that while there is a lot of post-harvest technology in fruit growing, most of the production work was still manual.
“One of the main challenges expressed was that they are still manually measuring and counting bugs, blossoms and apples, resulting in them losing marketable yield,” Lemieux says.
Vivid Machines developed a small spectral imaging sensor that detects wavelength into the near infrared - which can distinguish green apples from green leaves. Given the variability in the apple industry, they’re focused on building a system that is easily adaptable, inexpensive and simple to set up.
The current prototype includes a tractor or golf cart-mountable camera that can move at tractor speed, look at trees and detect diseases and pests, count blossom clusters and detect and measure apples. The system doesn’t require a separate pass through the orchard but can scan trees as growers are performing regular activities.
The goal is to have a commercial version by next spring that can be tested in up to 10 orchards next season.
“We are at the information-gathering stage now for the system. We have to feed it data, so it knows how to interpret the photos and make it more accurate as it learns,” she says. “The more data it gets, the more accurate it can be.”
Helping Lemieux and Binas with that are three growers in Ontario and one in British Columbia who have volunteered to serve as research farms during the system’s development.
Gerbe Botden is Orchard Manager at Blue Mountain Fruit Company near Thornbury, which supplies most of Canada’s major retailers. That means Botden is always asked to provide estimates on crop quality, quantity and size, a process currently done manually and by tree block.
“If you can make this an automated process and you come in and capture data on the farm, it’s going to be immensely helpful to us,” Botden says. “Data drives decision-making on our farm and in the future, we hope we can dive into more difficult aspects like nutrient content or deficiencies or pest and disease pressure, which can help us make better decisions through the season.”
Riley Bruce, orchard manager at Chudleigh’s Farm in Milton, is also helping Vivid Machines with data and farm access and sees a lot of potential to help boost crop quality, particularly in being able to predict disease problems before they strike.
“In the future, they should be able to tell with the camera what disease is coming next week, which will be better than our trained eye. The leaves can tell you what’s up,” Bruce says. “When you see the disease on the tree, it’s too late so this can open up our window a bit and we can treat before it shows up, which will increase our quality.”
Lemieux says the sweet spot for the technology is orchards in the 100 to 300 acre range, and she’s hopeful the technology will be able to lift grower profitability by as much as $1,000 an acre by getting more fruit into the quality and size distribution retailers want.
Vivid Machines was recently named one of 24 semi-finalists in the federal government’s Food Waste Reduction Challenge. Grand prize winners will be announced in 2023.