Ecorobotix and Maya Join Forces to Combine Ultra-High Precision with AI-Driven Data Intelligence

On April 23, Swiss company Ecorobotix, specializing in precision spraying in agriculture, announced the acquisition of Belgian Maya, creator of a digital assistant powered by AI that analyzes data from green spaces to facilitate their management and maintenance. While each plans to continue working on their ongoing projects, they hope to create a bridge between precision tools in the fields and decision-making based on real-time data analysis. A complete digital solution they envision will change the lives of farmers and land managers through the precision of the data utilized. In this joint interview, Steve Tanner, co-founder and CTO of Ecorobotix, and Valentine Godin, founder and CEO of Maya, discuss their companies, their joint projects, as well as their ambitions and vision for the future of agritech.

You recently decided to join forces, why?

Valentine Godin, Maya: There is an alignment in terms of shared values and visions. At Ecorobotix, they have always been interested in digital agronomy. We focus on grass. Both our companies are growing. We are both based in Europe, which makes things much easier. We speak French. The collaboration happened quite quickly and naturally. For us, the idea is to keep the product focused on grass and to share technology to have an impact in agriculture. And, above all, to find the synergies we will have with the ALBA™ machine (Editor’s note: Ecorobotix machine for maintaining grass with minimal chemical inputs) to have an even greater impact.

Maya and Ecorobotix target different clientele, but your activities address common challenges in sustainable management of green spaces. Who are your respective clients and what expectations do they share regarding efficiency, input reduction, and environmental performance?

Valentine Godin, Maya: At Maya, we do everything related to grass. We started in golf. Our main client is the turf manager. In agriculture, the goal is to produce food. In terms of grass, it’s agronomy, but for sports. The aim is the sports surface. In golf, the turf manager is a technical worker who oversees the entire site. Beyond golf, we cover all grassed areas: football, cricket, professional stadiums. And thirdly, all municipalities and green spaces where there are also significant needs: schools, cities with many sports areas to manage. We support them in that.

Depending on the geographical location, the ambition and the reason why our clients work with us differ. In southern countries, water issues are much more significant than phytosanitary problems. In Switzerland, France, Belgium, or the Netherlands, economic issues are very strong. In more northern regions, there are often performance issues related to play and field quality. Sometimes, it’s a bit of everything. So, we adapt. But there is always this desire to do better with less and with greater environmental awareness. In Europe, it is really very important, whether in terms of water, phytosanitary products, or biodiversity, to take care of the environment. The market demand is heading that way. And the laws are evolving in that direction too.

Steve Tanner, Ecorobotix: Our clients are primarily farmers who either own their farm, so the family farm, but you need to have a large farm to afford a machine like ours. In terms of size, you need several hundred hectares for the machine to be cost-effective. Or service providers (or CUMA in France), farmers who, since their farm may not have been large enough, decided: I will offer a service to third parties, to fellow farmers, with the machine. Or even larger professional companies, huge industrial farms, which can buy one, two, three, or five machines.

On which crops is Ecorobotix deployed and what barriers still hinder you from entering other markets?

ST: Vegetable crops represent our main activity, which we call agricultural. And we have green spaces that are very different. It’s not a consumable product, nor agriculture, but the management of green spaces, mainly golf courses. These are two very different verticals in terms of products, marketing, end clients, and the reasons they use our machines.

In agriculture, we have several segments. We have vegetable crops. About 25 vegetables for which we offer algorithms, classifiers, based on market values per hectare. Vegetables represent the highest value per hectare. Then we have row crops like potatoes and beets, which have less value where we are starting to be present. Today, in agriculture, we have 80% vegetables and 20% for the rest, the row crops. The third niche where we are not yet present is large crops, mainly cereals, like corn, but also oilseeds, rapeseed, sunflower. And legumes. For example, soybeans. The classification is done by decreasing value per hectare. (In large crops) what holds us back is the width of the machine and the cost of technology in general, still too high compared to the economic gain the machine brings to the farmer. Even if there is a direct environmental gain, the farmer will not buy a machine just because it is environmentally friendly. The cost must be at least equal, if not better than existing technologies. Our dream is to be able to serve all markets because we see the potential for product savings, thus a positive environmental impact. But we will not be able to enter these markets in the short term.

Do you plan to design wider machines?

ST: Yes, in fact, we are already communicating about a wider machine, but still for vegetables and row crops. We will launch it on the market next year. But for cereals, it will be much longer; we do not know when.

Is Maya's model, designed for green spaces, transferable to field agriculture?

VG: The issues are very similar. Some things are duplicable, or at least transferable to agriculture. But the objectives are not the same. The sectors are structured very differently. The audiences are not quite identical. We need to think about how to do it intelligently, keeping two things in mind. First, digital precision agriculture is more advanced today, more structured than in grass where we started almost alone. The sector is quite nascent, while in agriculture, smart farming, agtech, there are many players. So, the approach is different. This is where digital agronomy is very important. Secondly, we also need to consider the changes happening today in computing. The technological capabilities of today will not be the same as tomorrow. We need to keep that in mind to ensure we meet the sector at the right time, support farmers, and respond to their needs. Ecorobotix knows farmers' issues perfectly. We know grass. How are we going to put that together? My dream has always been to bring Maya into agriculture. For anyone working in agronomy, it’s still the impact we want to work on in the long term, which is much more societal.

ST: We dreamed of having a digital tool one day to advise the client on more than just ideas, but we faced many difficulties. We saw it as a mountain. We were impressed by how Maya managed to integrate all aspects of managing a green space. Exactly what we wanted to do in agriculture. For us, it was also a way to move much faster into agronomic consulting based on Maya's experience. With the hope that the data from the ALBA™ machine could eventually feed the decisions of the Maya software for golf management. And, in parallel, the same thing in agriculture, that our ARA™ machines could feed these software solutions. Today, we have slightly better technology on ALBA™, because golf is more demanding, but next year, we will have that same precision on the agricultural machine.

What common projects do you plan to work on in agronomy?

ST: Our machines can acquire very high-quality images to differentiate one type of grass from another, which can sometimes be very difficult to distinguish by eye. With this level of precision, they can provide high-quality information that we have not yet really exploited. Our plan for the coming years is to exploit the information that the machines use to make their decisions. It’s not a big effort to take this information and upload it to the cloud, process it to extract many insights. We cannot say too much, but we know we can do a lot, and today, there is no machine with equivalent capacity. The only ones that come close are the smart sprayers, but their image quality is much lower, so they manage to provide less precise information. That’s what we offer in addition compared to competitors, both in agriculture and in software.

VG: The idea is really there. The ALBA™ machine has eyes on the ground with unmatched precision today. And Maya, beyond being an interface, is the engine behind it, the contextualization and the power to gather data. Data for data doesn’t give you a decision. But structured and contextualized data does. So tomorrow, we will work with ALBA™, but we have 17 other sensors we are working on, different types of data sources. Contextualizing and structuring that within an infrastructure is what we offer to the client. The client today is not particularly into technology, but we will enable them to access this technology, to contextualize it, to have the answers, the advice, the support they need to make the right decisions. But also according to their needs because in different clubs, sports fields, or schools, there are different standards and thus different needs.

More concretely, what are their needs?

VG: The goal in grass is always to have the most suitable field for sports. In golf, it’s for the ball to roll as well as possible and to manage thousands of golfers per day, in terms of compaction and wear. In football, athletes work all day, and the need is for the grass to remain perfect, not to turn into a muddy field. Each time, we maintain a plant in service of a sport, and our goal is to do it with less environmental impact. As Steve said, sustainability is also, unfortunately or fortunately, financial sustainability. So with fewer resources, less environmental impact, and we also hope for more predictability, because one of the major problems of climate change is that yesterday's rules are not tomorrow's rules. There are many uncertainties, which, in sports seasons, is not easy to manage, especially with issues of diseases, herbicides, invasive plants, and weeds. In agriculture, it’s more about food, but we definitely want to work together to provide more context.

You mentioned diseases, is that a theme you could work on?

VG: Disease is at the heart of our business; it’s the reason for our existence from the very beginning. For example, in stadiums, they install lights to help the grass withstand. This is new information we know how to capture. It will help us, tomorrow, to make even more precise disease predictions, because the more context we have, the better we can assist.

ST: Disease management in agriculture is quite complex. There are many ways to combat diseases. We intervene at several points, not necessarily at the prophylactic level. That is to say: we detect a disease, so we need to treat it. Our technologies allow us to intervene in two other ways regarding diseases. The first is not to weaken the crop. If we “spray” the crop with a selective herbicide, it is not entirely selective; it will always have an impact on the plants, and often, it weakens the crop plant, making it more susceptible to diseases. Our technology allows us to push back the threshold of diseases. And then, another way to combat diseases is to maintain a certain floral diversity, unlike grass, which must remain a monoculture. In agriculture, it’s still different. We want a monoculture, but the functionality of the weed can be positive. It can reduce disease pressure, which manifests more in monocultures. When they are not monocultures, plants have beneficial effects on each other. Some plants can have a repellent effect or slow down certain infections. We are developing this gradually. We are not very strong in this yet, but it’s the third of our three “Rs”: reduce the quantities of chemicals, replace synthetic chemicals with natural chemistry, and regenerate. One way to regenerate nature is to allow floral diversity.

How do your machines act on floral diversity?

ST: By recognizing plants, we can choose to leave some plants, some weeds of that species, because they are not that harmful. They have floral benefits for insects. They can also have benefits for fixing nitrogen in the soil, for slowing down diseases, and soil erosion by wind. Our machines can do that. That’s why we can indirectly combat diseases. We have two ways to combat diseases: non-phytotoxicity and the diversity we can manage. Non-phytotoxicity is already taken into account because we “spray” the crops marginally. They grow better. It’s also very different from one crop to another. Our flagship crop is onions. It’s certain they grow much better without the phytotoxicity brought by a standard herbicide application. We have clients who are starting to do crop associations, but we don’t yet have enough hindsight to say that it really slows down diseases. It’s a bit of a medium-term vision.

Regarding the preventive and prophylactic aspect, we will apply a fungicidal agent just on the crop, not on the weeds, before the disease arrives. Generally, when we detect a disease, it’s too late. We need to intervene beforehand. That’s the whole secret of a recommendation software like Maya. We don’t have software in agriculture. So the farmer uses other platforms to know when to apply their fungicide. However, with our machine, they can apply up to ten times less because they only apply it to the plants that need protection.

How did you adapt Ecorobotix's Plant-by-Plant™ technology to your turf machine (ALBA™) where the grass blades are very close together?

ST: Technologically, we were already working in pastures, which produce grass for livestock. With ARA™, we detect weeds that are extremely small. We are not more precise with ALBA™. It’s the same vision system. In that context, we already knew how to detect a plant among many others. The green carpet, we already mastered. For golf, the greens, it’s about acting a bit more finely. Then, the difficulty was to properly annotate the images, to train the model. To work with people who have the skills and experience. It’s a market we are developing, particularly in the United States. Because we find global specialists, universities where there are even PhDs in grass, with whom we work to gain knowledge on how to train our classifiers. We need to be able to say: in this area, it’s a different species of grass than that one, so it needs to be “sprayed.” To the eye, we think they are the same. It’s easier in an agricultural setting where a beet is very different from a lamb's quarters or a grass.

What do you think is the future of precision spraying? Are we heading towards even more precision, to the millimeter, for example? Will other product mixes appear?

ST: It makes less sense to go below a certain level of precision because we are already doing enough. There will be a limit. We are not there yet, but we think we will get there in 3-4 years. That will be sufficient to do everything that is possible. It won’t make sense to be more precise. Very high precision allows us to change the chemistry. And that’s a disruption in agriculture because the chemistry that is needed consists of very simple, natural molecules - vinegar, pelargonic acid, natural agents - and thus accessible to everyone. There is no longer the chemical lobby that controls the selective molecule that required 200 million dollars to develop and is monopolistic. Or the herbicide/GM resistant pairs. We no longer need that. We believe the future of precision spraying is bright. It’s a change that will happen. We think that in 10 years, most spraying, herbicide applications, will be done with ultra-precise machines. And other machines will no longer make sense because the products will be too expensive, too complex, or may have been banned...

Will the use of simple and natural molecules lead to fewer harmful products in the fields?

ST: Cjust because a molecule is natural doesn’t mean it’s not harmful. For example, an essential oil can be a poison. The simplest agents are acidic agents. They attack the leaf and kill the plant by lack of foliage. They leave almost no residues behind. Obviously, ideally, they should only be applied to the leaf. They can acidify the soil, but that can be compensated by a base input in a previous treatment, or in the fertilizer input. Adding a bit more pH and compensating. We need to choose chemistry that degrades very well, usually natural chemistries. After that, vinegar, we can also synthesize it. Just because a molecule has a natural origin doesn’t mean we won’t synthesize it. For example, pelargonic acid comes from geranium. If everyone starts spraying pelargonic acid, we won’t have enough geranium on Earth to supply it. It’s not a problem to produce it synthetically in a reactor since it’s a natural molecule. After that, we are trying to reduce doses to the maximum. It’s like milk. If you put a cubic meter of milk in a river, you kill all the fish. Yet, we drink milk. Toxicity is very relative. We must always seek to reduce doses to the maximum. But we are heading towards major changes in chemistry.

Are you considering innovation that would go in a completely different direction than precision spraying?

ST: On this subject, we cannot say much more. We are looking at other areas of agtech. But today, there is so much to do in spraying that we are focused on that.

If today robots replace pesticides, is it a springboard for regenerative agriculture?

VG: For golf, it’s monoculture. These are often spaces of 50-60 hectares. I’m mainly talking about European golf. Today, there is between 30 and 50% of sports fields, depending on the countries. The remaining 50 to 70%, or the vast majority, are natural spaces. Orchards, trees, which are green spaces to manage. The monoculture we are talking about occupies 30% of the land but occupies 70% of their maintenance considerations because it’s their sports field. The rest is us, and the big work the sector is doing today is to renaturalize the 70% as much as possible (excluding sports fields). Today, we see more and more natural roughs. The vast majority are working on course designs. They have ecological corridors and ensure that between the green space on the left and right of the course, there is an ecological pathway where insects, birds, and wildlife - there are often deer on the courses - can live. Once 70% of the course is green, and ecological corridors are ensured within the landscape, there is already a much more positive impact on biodiversity.

ST: Yes, but it’s one element in the chain. In agriculture, it’s clear that we want to reduce pollution, but when we talk about regenerating soils and biodiversity, it’s very complex, and it’s not just a robotic technology that will allow it. The big problem we have today is the size of the fields. We have monocultures. In one square kilometer, there is only one species. Imagine the insect; if it doesn’t find food, will it be able to cross that space? One answer would be not to make huge fields but to fragment, to have diversity, to associate crops with smaller plots; that’s the most important. Also, hedgerow corridors, reintroducing diversity in landscape management. Because without that, we can have the best machine and be deluded, but we won’t be able to solve the problem. After that, if we have a smaller plot, to reduce the pressure of monoculture on biodiversity, then our machines can act. And reducing chemistry is indeed a way to achieve that because it is harmful to biodiversity.

Would one of the solutions be to eliminate chemistry?

ST: We can very well, in my opinion, have a system that is very little harmful to nature that retains chemistry if we use much less of it. I don’t think we will need to completely eliminate chemistry because it is still practical. And in nature, there is also chemistry. It’s not: chemistry is bad because it’s chemistry. No, chemistry is bad because we use toxic chemistry and we use too much of it. But if we manage to reduce doses and the chemistry is no longer toxic because it is of natural origin, it’s no longer a problem. We try not to be too dogmatic; otherwise, we risk going into nonsense. For example, saying “we will do everything with lasers,” but ultimately, the laser will kill insects, life in the soil, because by burning the soil, we can have a negative impact. Of course, when we do agriculture, we have a negative impact on biodiversity. But it must be minimal. One way to achieve that is to reduce doses and switch to natural products. That’s a bit our approach.

The approach is that we also need to remain realistic. Not everyone is ready to pay for organic food. So we need to have pragmatic and inexpensive solutions. We must first tackle industrial agriculture and provide economically viable solutions to help it improve and reduce its environmental impact. But we remain in industrial agriculture. We do not see how it could disappear. It will disappear if it is replaced by economically similar agriculture. Because ultimately, the consumer is not ready for that change.

Do you think robots will one day manage farms?

ST: I will talk about the machine, and I will let Valentine talk about the software. Yes, we are moving towards the automation of the tractor and the machine behind it. That will be very good because sitting in a tractor for eight hours, following a line, going back and forth, is not very interesting. It makes sense that it will be replaced by autonomous machines. It’s a matter of years, but there will always be someone to bring the tractor to the field and repair it. The farmer will spend more time doing something more interesting than driving a tractor. Driving will be reduced.

VG: It’s a bit the same answer. Completely replacing, no. Changing jobs, yes. The jobs today, whether in sports or agriculture, are changing. The job of a farmer or the job of a turf or golf manager 20 years ago is not the same as today. The goal is for them to focus on their added value. Everything that is of lower added value can be automated, replaced. We see it in the sports world; we can really reduce the mental load they have, which today is very intense. There is a generational change coming, and it also allows for training and education. But in no case does it replace. It’s a bit like the robotic mowers that are coming into golf. Does that mean you will no longer have a team? No, but you will be able to do things differently, focus on other things that you weren’t doing but that have a more significant impact on your operation. We will redefine certain jobs, but we are certainly not replacing them. We facilitate, we optimize.

Ecorobotix announced it has invested 50 million dollars in the United States and is assembling its ARA™ machine in Kansas. Is the American market currently the most interesting for technologies like yours?

ST: For agriculture, there is a need to increase efficiency in the American market, stronger than in Europe. Particularly regarding labor, which is becoming increasingly difficult to find. This is also a European problem, but in the United States, in certain regions, it is very difficult to find people. And Mexicans will not want to do it for decades. Here too, the Poles who used to come to weed are gone. It’s very difficult to find them. There is a global labor problem. But in the United States, there is an appetite for technologies that increase efficiency. Today, it’s the market that is driving forward. Not necessarily in terms of precision. Europe may be more advanced in precision because we have smaller fields, so we have optimized yield per hectare more. However, in terms of costs, it’s the United States that need to improve their margins. We also benefit from being the first. There are American companies that are active, like Carbon Robotics, in lasers. There are global leaders in the United States in the agtech field. Carbon Robotics is not a direct competitor; they are really in the organic space. The laser remains a niche for very small weeds, for ultra-high precision. We are broader, so we are not really competitive.

VG: In terms of numbers, in the sports sector, over 50% of the market is in the US. Just by volume, the bulk of the market is there. For any player in the sector, the primary market is the US. We have the United States, the United Kingdom, and then the rest of the world. The digital maturity in the United States is much more advanced than in Europe. We have an audience that is much more prepared. In Europe, we are still educating a lot. I don’t want to say we are evangelizing, but in a way, yes, because it’s still very new. While in the United States, they need it since they already have 7, 8, 9 data sources, but they don’t know what to do with them. We provide them with the solution to transform data into decisions. That’s why there is an appetite in the market, and the opportunity to work with Ecorobotix gives us access to a market. With five people in Belgium, it’s more complicated to achieve that.

You mentioned Carbon Robotics… There is also the See & Spray™ technology from Blue River Technologies (John Deere), less precise than yours, but they have that data integration aspect. Do you aim to play on their field?

ST: For now, we are insignificant in terms of size. They are such large companies. The spot spray technology has not demonstrated universal acceptance among farmers because it has limited precision. So that’s also the unique aspect of our proposition. We come with something different. But today, we are confined to vegetables. And that’s not a market where John Deere is, so we are not competitors.

Do you have international development projects, other than the United States?

ST: We have been eyeing South America for a few years. We had machines on demonstration in Argentina and Brazil. That will be the next step for us. South America will be a market with much less value, crops with less added value. And a lower technical maturity for Brazil compared to the United States.

VG: We do not do hardware, but software. So, it’s easier to grow internationally once the schedule works. We want to consolidate Europe and the UK. The United States and Australia are the two big markets where today there is appetite, there are ongoing conversations, so I hope a client signs this summer in Australia.

Currently, which country are you most present in?

VG: The largest countries are France, Belgium, Spain, and the United Kingdom. These are the four countries where we have a good customer base. Then, in Europe, we have Switzerland, Italy, Portugal, and Greece.

ST: For us, it’s the United States. And then the second is the Netherlands.

Ecorobotix is a Swiss company founded in 2014 and based in Yverdon-les-Bains. It currently employs around 350 people, including 50 based in the United States, 35 in France, and 4 in Slovakia. Since the acquisition of Maya last April, it also has 5 people in Belgium and one person in southern Europe. Its ultra-high precision agricultural sprayers ARA™ and turf ALBA™ allow targeting weeds in fields without damaging crops and maintaining grass with minimal chemical inputs. They promise up to 95% reduction in the use and costs of chemical products while allowing the precise application of selective and non-selective products.

Maya is a Belgian company, founded in 2022, specializing in the optimized management of terrains (golf courses and stadiums) and green spaces (parks and campuses). By leveraging data provided by clients and developing a specialized AI assistant, it provides an agronomic analysis tool to terrain managers to assist them in their daily decision-making.