CAAIN
CAAIN
Vivid Machines, Toronto, ON
Tall Grass Ventures, Calgary, AB
Gibson Orchards, Newcastle, ON
Blue Mountain Fruit, Thornbury, ON
We recently had the pleasure of sitting down with Vivid Machines’ CEO and co-founder, Jenny Lemieux. She graciously and patiently answered our questions and explained what Vivid does and how it came to be.
CAAIN: Thanks for taking the time to speak with us today, Jenny. We’ve been looking forward to learning more about the work in which Vivid Machines is engaged.
Jenny Lemieux: It’s my pleasure. I’m happy to be here.
CAAIN: So, you are the CEO of Vivid Machines, a Toronto-based agtech startup. Correct?
JL: That’s right.
CAAIN: We like to tell our project partner’s story before diving into the specifics of the technology being developed. What was it that set young Jenny Lemieux on the path to becoming the chief executive officer of a promising agtech startup?
JL: The swimming pool in our Ilderton, Ontario, farm.
CAAIN: I’m sorry. What was that?
Jenny laughs and explains.
JL: I grew up in London. But both sides of the family were rooted in farming—pun very much intended—and we moved to a farm near Ilderton when I was a teenager. Our property had the only pool for miles around and I was a strong swimmer, so I launched the Lemieux Swim School. The money I made paid for my university studies in Chemical Engineering.
CAAIN: Oh…gotcha. So, you were destined to be an entrepreneur who made a big splash!
JL: I’ll pretend you didn’t say that.
CAAIN: Thanks. Much appreciated. I get that a lot.
JL: After graduation, I spent three-and-a-half years working for a large multi-national insurance company doing risk management in huge industrial plants. My first boss was a tremendous mentor who taught me so much about staff and team management. But the work itself wasn’t feeding my creativity, so I went back to school to study Industrial Design. After graduation, I began working on every-day things like dishes and furniture. I also had the opportunity to develop competencies in product management, which was an intersection of various skills required to take a product to market—basically it was all about how to generate revenue from business assets, another set of skills on which I rely at Vivid.
CAAIN: I can see how that would be relevant in your current role. What led you from industrial design and product management to found an agtech startup? There doesn’t seem to be a connection.
JL: I wanted to make a difference, and I felt that wouldn’t happen working for someone else, so I signed up to participate in “Entrepreneur First,” a 12-week program that matches pairs of would-be founders and has them collaborate in developing ambitious, venture-backable start-ups. Right off the bat, I was paired with Jonathan Binas, a brilliant physicist who had worked on everything from AI and neuroscience, to electronics, including brain-inspired learning algorithms and novel hardware for neural network computation.
CAAIN: We I didn’t understand any of that, but I’m guessing what you’re really saying is that Jonathan is a genius.
JL: Exactly. We hit it off right away, and four years later we have a thriving business, a team of more than 10 employees, and clients in Ontario, Nova Scotia, Quebec, New York, Massachusetts, Michigan, Washington State, Spain, and New Zealand. The reason we clicked is that we both wanted to find a sector where we could make a difference.
CAAIN: Okay, but why orchards?
JL: We started by talking to farmers about what kind of tech they would find useful. Nothing jumped out at us. Then someone suggested we look into orchards, and the owners and managers we contacted enthusiastically invited us to visit them. We took them up on their offers and learned about the issue of crop load management.
CAAIN: What is crop load management?
JL: Essentially, it refers to ensuring that each tree yields the optimal number of the right size of fruits in the current crop year, while also being set up for future success. Put simply, have you ever noticed how when you go to a grocery store produce section, the apples and pears are pretty much consistent in terms of size, shape, and colour?
CAAIN: I’ve never thought much about the appearance of fruit, but now that you mention it, yeah. That’s interesting.
JL: Exactly. Crop load management is the process that generates target fruit size, whether on trays or in bags. Our tech can “read” trees and predict crop features for each geolocated tree. There can be literally thousands upon thousands of trees in a single operation.
CAAIN: Wow! That’s remarkable.
JL: It really is. Jonathan and the team developed a camera that can see, record, and capture a range of data in the RGB spectrum, as well as in near IR, at tractor speed, which is about 7km/h. This as a significant advance. The Vivid takes videos of each plant; technology geolocates them and attaches a range of meta data, including fruit variety. This information is then fed into our computer vision models. Currently we can predict things such as the number and size of apples on a tree and extrapolate yield for a farm for the year. And we aren’t resting on our laurels. We are researching disease detection in collaboration with Vineland Research and Innovation Centre and Cornell University.
CAAIN: When you said Jonathan was smart, you were really understating the size of his big brain.
Jenny flashes a huge grin.
JL: Yes. I absolutely was.
CAAIN: So, Jenny, it sounds as if you guys are rocking. Why did you apply for ISED/CAAIN funding? How has our support helped?
JL: The funding has been invaluable. First, it’s helping with tree vigor assessment. For now, we can predict a tree’s yield based on fruit quantity and growth rates. But that’s not good enough. We’re working on prediction based on the number of buds on a tree to help direct pruning. Buds appear well before the blossoms, so it gives the farmer more time to take appropriate action—i.e., how much pruning is required to optimise yield. But this is very difficult. We continue to build on our computer vision and machine learning models to broaden our predictive capacity. CAAIN’s contribution is allowing us to hire more people to build more datasets, which in turn will allow us to detect the relevant factors while the trees are dormant. That would, in turn, enable us to predict a tree’s crop load far enough in advance to allow the orchard manager to prune dormant branches for optimal crop load per tree, making for a healthier operation. We’re going to be able to generate historical data and compare year-over-year results. Before we were awarded the CAAIN funding, we were unable to start our analysis until April. Now we’ll be able to provide useful predictions in the winter well before the trees “wake up.” We’re also able to factor in other parameters, such as tree size. Basically, it comes down to having been able to hire more highly qualified personnel, which has us generating revenue after only three+ years.
CAAIN: Jenny, to say that we’re impressed would be an understatement. Congratulations on your well-earned success.
JL: Thanks. I can’t tell you how much we appreciate the support we are receiving from ISED and CAAIN.
CAAIN: It’s a privilege and a pleasure, Jenny. Thank you so much for taking time from what is surely a very busy schedule. Vivid Machines is engaged in fascinating R&D.
JL: You’re very welcome.
CAAIN Contribution
$820,164
Total Project Value
$2,443,006
Jenny Lemieux
Co-Founder and CEO
Vivid Machines
jenny@vivid-machines.com