Le premier magazine de l'industrie agroalimentaire française, Végétable, a publié cet article concernant nos efforts de fournir l'intelligence artificielle aux producteurs de serre.
[Citation: no 364 / mars 2019 • vegetable.fr •p. 71]
Cultilene [de Saint-Gobain] annonce son partenariat avec la start-up canadienne Motorleaf pour la commercialisation du service de prévision de récolte haute technologie de celle-ci. L’intelligence artificielle permet aux producteurs d’utiliser leurs données de serre pour contrôler, prédire et améliorer les conditions de croissance et les processus opérationnels.
L’innovation réside dans la construction d’algorithmes d’apprentissage automatique personnalisés permettant de prédire les niveaux de production futurs dans les serres. En plus de réduire les risques liés aux rendements de récolte hebdomadaires imprévisibles, cette technologie permet aux producteurs de rationaliser leurs opérations, de réduire leurs coûts et de conserver leur expertise en matière d’agriculture grâce à des logiciels intelligents.
Les serres de tomates et de poivrons du monde entier utilisent déjà les prévisions de récolte automatisées de Motorleaf : leur technologie d’automatisation artificielle a permis de réduire les erreurs de prévision de 50 à 70 % par rapport aux prévisions actuelles basées sur des calculs manuels des cultures.
Growers used to ask, "does it work?" Now they ask, "how does it work?"
The organizers of the leading international produce trade show, Fruit Logistica 2019, selected Motorleaf as one of 20 most impactful new agriculture technology companies to present their innovations on stage. We grabbed the attention of many in attendance, including several media outlets.
Here you can watch one of our interviews summarizing our AI automation services for greenhouses we displayed at the event, between journalist Chris White (CW) of Fruit Net Media International and Dr Jason Behrmann (JB) of Motorleaf. If you prefer to read, below you will find a transcript of the interview.
CW: Hi, good morning. Welcome to the media studio at Fruit Logistica. I’m Chris White from Fruit Net Media and I’m joined this morning by Dr Jason Behrmann of Motorleaf. You’re a Canadian company, Jason, a newly established start-up running for a couple of years; and your business is all about artificial intelligence and glasshouse vegetables, is that right?
JB: Yes, that’s true. So, our company Motorleaf—I oversee marketing and communications—the company was established in 2016 and we specialized in artificial-intelligence-enabled automation for the production of tomatoes and peppers in greenhouses.
CW: So what does that mean: ’Artificial-intelligence automation of production’? I’m intrigued.
JB: Well, instead of using robots and other forms of machinery, you can use very intelligent software to automate important tasks in greenhouse farming. So, our primary service right now is doing highly-accurate yield prediction, meaning forecasting the exact amount of tomatoes and peppers a greenhouse will produce weeks in advance on a week by week basis.
CW: And I can kind of see the advantages, but tell me what they are for this application of this technology for the grower.
JB: Well, using big data of the growing conditions within the greenhouse we are able to develop incredibly smart software that can predict within a few percentage points of the actual yield a greenhouse will produce on a week-by-week basis. So, it’s the first time in history that a farmer will know just how much tomatoes or peppers they have so they can better plan their labour; they can better plan promotions and marketing initiatives well in advance; and they can better plan ideal pricing for their produce; and they will know whether or not they’re going to under- or over-produce to meet contractual agreements with their buyers so they can better plan if, you know, something goes wrong and make sure they keep that really great relationship with their buyers.
CW: Now, the big advantage you have, of course, is that you’re growing in a controlled environment so you know what the inputs are, you can literally predict day-by-day, and determine day-by-day the conditions for growing are going to be. That’s presumably why this works, is it?
JB: It’s one of the first times in history, now, with all the different types of technologies and machinery and also different types of sensors, we have a wealth of data available now where we can gain a fundamental understanding of the ins-and-outs of greenhouse farming and that’s what we can now use to develop cutting-edge automation—artificial intelligence technology. Before, we did not have access to this data and also now with all the data we have, a lot of people are scratching their heads, especially the farmers, saying, “ok, well what can we do with it?” Well, the ‘what’ is artificial intelligence.
CW: And that’s taking big data, essentially, and condensing it down and giving people the kind of information they need in terms of being able to predict what is coming in terms of production.
JB: Yes. So, the services we offer is a turn-key, full-suite of services where we will develop custom-made technology for each greenhouse and each variety of tomato or pepper they produce. So, we handle all the ‘mess’, let’s just say, all the complex computer technologies, and then we will be able to convey that information to them in terms of weekly yield simply be communicating through any kind of digital format; it could be as simple as an email. And the additional services we are developing now are automated disease scouting; so, we can conduct are very detailed analysis of the growing conditions within a greenhouse and identify zones that are experiencing environmental stress, like extreme heat or dryness, and that makes plants vulnerable to disease. Armed with that information we can tell growers, “hey, look out in that area, intervene early using fewer pesticides and sacrificing fewer crops to avoid the problem.” Another product we are developing and will be on the market soon is detailed analysis of growing conditions that cause slight imperfections in tomatoes and peppers. So, right now, about six to 12 per cent of a harvest cycle is lost due to small imperfections in the fruit, such as skin cracking—
CW: And ‘loss’ means lost money as well, of course.
JB: Absolutely lost money, and also contributes to food waste, which is a monumental problem that many people are aware of now and there are new regulations throughout Europe, for example, that are trying to quell this problem. And with our data analysis we can identify factors that will cause those small imperfections and inform growers what they can do to avoid the problem altogether, increase their profits and also avoid crop loss.
CW: Sounds like a win-win. Now, your are based in Quebec, in Canada, and yet you are working with growers all around the world now.
JB: We are based in Montreal, Canada, which is an international hub for the development of artificial intelligence technology. And we currently have clients in Japan, Tunisia, Europe, Canada and the United States; and I’m very excited to announce that last week we just established a new partnership with Cultilene. Cultilene is one of the world’s largest suppliers of glass substrates and substances for greenhouse growers, and though this partnership we’re going to introduce our artificial intelligence technology to their client base, which is in, oh, spans over 50 countries. So, we believe we can introduce AI into greenhouse farming throughout the world within a short period of time.
CW: So, it sounds like the future of farming very much is about AI.
JB: The future of many industries is going to be heavily disrupted by this technology. It’s really ground-breaking. Farming is just one sector, and with the growing slew of services that we offer greenhouse growers, I am very confident that we can make a significant, positive impact in this one industry.
CW: And you use the word ‘disruption’. Disruption always sounds to me very negative. Sounds what you’re talking about is extraordinarily positive.
JB: ‘Disruption’ is a very positive term. It’s providing new ways of doing old forms of industry and making current industries far more efficient and far more profitable and accessible. And artificial intelligence checks all those boxes.
Cw: Excellent. Thank you and good luck.
JB: Thank you very much.