How to deal with
precision farming issues

— Yes, farmers often do not need startup developments. It’s kind of "technology for the sake of technology" situation.

It seems to me that many startups have a poor understanding of agricultural specifics. For example, developers know how technology is being applied in one country — and then simply try to transfer it to another country. However, if the production of, say, a mobile phone in the USA and Vietnam can be built in more or less the same way, to harvest rice in these two countries, one needs to go down two different technological paths. Each country in the world has its own specificity of fields, soils, climate, farming traditions, etc., so when developing new AgTech tools, a huge number of factors must be taken into account.

To create an agrotechnical startup, you need to think like a farmer. Together with Slava Mazai, the co-founder of OneSoil, we practically lived in the fields for three seasons: we used to leave home at 4 am to return by midnight — and start all over again the next day. Now we go to the fields less often, but every day we communicate with fellow farmers and constantly test ideas from the "useful/useless" point of view. At the same time, I personally know a few AgTech startups whose founders have never even set foot in a field.

I think that another way to approach this problem is to decide for yourself what precision farming is and where you need to implement it. The thing is that even fuel sensors for tractors are sometimes considered to be precision farming. I disagree with that. From my point of view, precision farming is a farm management system which implies the variable-rate application of seeds, fertilizers, and pesticides. Using this definition, if the fields are flat with the same yield values, there is no need for precision farming at all. So the problem of useless apps becomes less important.

— For sure, new applications and technology are too complicated. Even for me, it is difficult to figure out the way some AgTech programs work. In such situations, farmers often have to appoint a specialist to digitize the farm. There is also a psychological barrier. When several generations of your family cultivated fields in a certain way and managed to harvest crops, introducing new approaches may be difficult. Farmers tend to be quite conservative, but it seems that we are on the threshold of change. Such are the new requirements of the market — technologies increase farm efficiency and make it more competitive.

At OneSoil, we use Sentinel-2 images with a resolution of 10 meters from a satellite that was launched only in 2015. Using a neural network allowed us to find the boundaries of all the fields in Europe and the USA. Remarkably, the very concept of neural networks began to develop only in the 2000s, which is connected the appearance of powerful graphics processors. There are also several other factors.

A plant is a living organism and is influenced by a huge number of factors. Secondly, farming is still not very popular amongst youth, therefore not that many people are keen on developing new technologies. For a long time, agriculture was considered to be the work of peasants while the population of the planet has been actively moving to cities since the end of the 19th century. However, agriculture is becoming more technologically developed and attractive for the youth. For instance, the number of USA farmers under 35 years is increasing according to official statistics. Moreover, farming can be very profitable when managed well. So I think the general situation will change.

— Unfortunately, this is the sad truth. An average farmer can’t afford new applications and technology. Tractors, sprayers, seeders and other equipment with onboard computers can cost hundreds of thousands of dollars. Field analysis on commercial online platforms starts from $ 2 per hectare, and to conduct an agrochemical analysis of the soil is around $ 10−20 per sample.

I think that software developers and equipment manufacturers know that their products are unique, which means that farmers will buy their products anyway. On the other hand, the process of developing precision farming tools is costly. Research, testing, refinement, implementation, marketing, etc, — for all these steps there are no simple and predictable paths yet. Again, that’s because the whole sphere is very new.

We want to show that technology can be friendly and should benefit not only large farms but also those who don't have hundreds of thousands of dollars for new equipment and applications. To fully understand our business model, please read the article on our blog. In brief, we will provide different services to large companies. Our products for farmers will remain free forever.

— In my opinion, this problem is gradually disappearing. The turning point was the appearance of the ISOBUS protocol in 2001, which allowed for the exchange of data between tractors and seeders, sprayers, etc. Today, this protocol is supported by more than 200 companies and institutions.

Another major problem is the incompetence of equipment dealers. Their main goal is to sell machinery, dealers do not care how the farmer will use it. Sometimes dealers cannot even explain how a seeder or a tractor works, and often farmers don’t know whom to ask their questions to. I hope that the situation will change soon. I dream of equipment manufacturers creating some kind of support center that will guide farmers every step of the way in the process of digitizing their farm. Then the need for dealers will disappear.

— Sometimes there is too much noise around precision farming: "agrotechnology was evolving all the time, what's the big deal now?" Everyone is using the terms "artificial intelligence", "machine learning", "neural networks", but many applications do not work very well, the data is inaccurate, and the benefits for farmers are not obvious. From my point of view, there is a certain amount of hype indeed. The constant growth of investment in this area is just one indicator: the AgFunder report for 2018 states that investment in AgTech increased by 43% and amounted to almost $ 17 billion in comparison to 2017.