Sign up for our daily Newsletter and stay up to date with all the latest news!

Subscribe I am already a subscriber

You are using software which is blocking our advertisements (adblocker).

As we provide the news for free, we are relying on revenues from our banners. So please disable your adblocker and reload the page to continue using this site.

Click here for a guide on disabling your adblocker.

Sign up for our daily Newsletter and stay up to date with all the latest news!

Subscribe I am already a subscriber

Certified organic greenhouse uses AI to finetune precision robotics

New York-based company Agbotic has been using AI in their operation to improve their efficiency and the quality of the crops they grow. "We finetune precision robotics through data analytics and AI," says John Gaus, founder of the company.

John is a serial tech entrepreneur who describes himself as a passionate environmentalist. "Some time ago, my team and I got a lot of exposure to really bad environmental aspects of farming, which motivated some of us to find solutions to those problems." And when it comes to horti problems that hinder a company's bottom line and sustainability efforts, the culprits are always the same two: "Opex and Capex," says John. "When you look at Opex, energy and labor are always the heavyweights, and you need to aggressively tackle those. We have come up with some novel structures in greenhouse design efficiency, as well as with some interesting tools in the way of precision robotics." For Agbotic, that resulted in lowering the facility capex, while driving labor cost out of it. "Next to that though, you also need to be profitable and create high value-high quality products. In agriculture, the biggest sensitivity in the economic model is the yield, and we can drive that through finetuning the big raw data set we possess."

Do it right or don't do it at all
At the heart of Agbotic, there's quite an easy concept that gets often overlooked, especially by the most tech-intensive space of CEA. "If we cannot grow a product organically, chemical-free, with a regenerative farming model that cannot compete with all the other forms of cultivation, then it's a bad idea," says John. Bluntly put, if you are growing a commodity food product, you must do it in such a way that it can compete with other methods of farming. This aspect feels like has been overlooked by the (early) vertical farming space. "The tech of vertical farming is cool, and the story behind that is cool, but at the end of the day, you have to make food at a competitive cost of production. Currently, everybody knows that the economics of vertical farms hardly add up."

At the same time, economics not adding up do not belong to vertical farms only. "There was a study that analyzed in great detail the economics of vertical grow farm growing lettuce, as well as greenhouses. Well, that study found that all forms of greenhouse farming, either horizontal or vertical, were more expensive and with a bigger carbon footprint than outdoor cultivation." What John is trying to say is that all the greenwashing in the world will never be enough to cover facts. In other words, stating that CEA produces less emissions by default is not true unless specific pain points get addressed properly.

Artificial wisdom
So, how is Agbotic using AI in their greenhouse? "Most AI comes to data analytics," John explains. "For instance, think of when you type on your phone, which is guessing which word may come next. It's a probabilistic analysis. AI is an exceptionally large language processing model which is not like your average software. Rather, AI models run probabilistic analysis on huge scales, and do data analytics, and this allows to increase plant yield and quality. We can effectively do this because we have gathered so much data. After all, knowledge is power, and data is knowledge."

John proceeds to describe how that is translated into their operation. "We have been able to increase yields for some of our crops. It's a matter of finetuning everything about plant biology and environmental control, together with the precision of machines. With robotics you get automation, but if robots can also be very precise and tunable in tiny increments then you get more out of the automation than just their replacement of labor."

A sleeper build
While AI and robots are surely the thing that stands out the most in John's operation, the greenhouse structure too is quite innovative and is an absolute sleeper build. "The greenhouse is tailored to the local climate. This was designed in such a way that maximizes the thermal gains, while simultaneously improving the energy efficiency." From the outside, the greenhouse indeed looks quite low-tech – yet things inside look totally different. Workers inside the greenhouse are wearing t-shirts, while snow is falling outside. "The genius of our greenhouse design lies in its simplicity and efficiency, maintaining warm temperatures inside at a low cost even while it snows outside," Gaus notes.

All of that tech, data/knowledge allows Agbotic to be fully USDA organic, meaning that they grow in soil, and yet in blind tests, the quality of their products always stands out. "We recently started growing spinach, which is an interesting crop because it doesn't go well in hydroponics, and outdoor spinach gets super sprayed. We have been able to grow this in our greenhouse with good unit economics. And needless to say, it absolutely rocked the blind test."

Obviously, John and Agbotic are looking to export their model to other countries and locations. "We are now thinking about nine different projects, some here in the US, and some abroad – like in Ireland, for instance. At the same time, we are not going to own greenhouses everywhere. We are a food company first of all, and we make our own tech to grow that. We are going to provide this tech to other partners to give them the ability to do the same thing we do here."

For more information: