Problems of precision farming and how to live with them

    I publish reflections at the request of my colleague, who is not on the hub



    Hi, my name is Seva Genin. I have been engaged in GIS and agricultural chemistry for many years, and now I am involved in the development of OneSoil - a free platform for farmers around the world with AI inside.

    The theme for the habr is a bit unusual, but precision farming is technology-based, and so I decided to reflect on its problems.

    We often hear about the disadvantages of precision farming. Dissatisfied farmers, wary agricultural consultants giving advice for money, prudent financial analysts - all point to the same pain points.

    A lot of useless


    - Yes, sometimes technologies appear that farmers do not need. Such "technology for technology."

    In my opinion, one of the reasons is that many startups have a poor understanding of the specifics of agriculture. For example, developers saw how technology is applied in one country - and try to use it in another. But if the production of a mobile phone in the USA and Vietnam is approximately one process, then in order to grow a rice crop in these two countries, two different technological paths must be followed. Each country in the world has its own specifics of fields, soils, climate, and farming traditions - a great number of factors must be taken into account when developing new tools.

    To make an agrotechnical startup, you need to understand how farmers live and what they think. We practically lived in the fields for three seasons: we left home at 4 in the morning to return to midnight - and the next day to start all over again. Now we go to the fields a little less often, but every day we communicate with our farmer friends and constantly test ideas in terms of their applicability. At the same time, I personally know several agricultural companies whose founders have never been in the fields.

    How to live with it?


    Still do not use unnecessary applications and devices. The lack of demand for this kind of development will lead to the fact that meaningless startups will leave the market, and useful for farmers will remain.

    You can look at the question more broadly and decide for yourself what precision farming is and where you need to implement it. Now, under this term even trackers on tractors that measure fuel level have begun to understand. From my point of view, precision farming is a farm management system that involves the differentiated (with a variable rate in different parts of the field) application of seeds, fertilizers and pesticides. Based on this definition, if the fields are flat, with the same yield indicators, then precise farming does not need to be introduced at all. And the problem of useless applications is getting much smaller.

    Too complicated


    - Yes, it is, often applications and equipment are complex. I myself find it difficult to figure out the interface of some programs, let alone farmers, who often have to appoint an individual to digitize their farm. There is one more difficulty - psychological. When several generations of your family cultivated the fields in a certain way and at the same time harvested, it can be difficult to decide on something new. Farmers are quite conservative, but it seems to me that we are on the verge of change. These are the new requirements of the market - technologies increase the efficiency of the farm and make it more competitive.

    How to live with it?


    Have patience. It is important to understand that all new technologies are initially complex and accessible to a small number of people. Talk that something needs to be done with the heterogeneity of the field began in the 1930s. Here, for example, is a textbook of the University of Illinois in 1929 , in which the author discusses how to build a map of soil acidity. But active precision farming began to develop only in the 2000s, when the quality of satellite images, the computing power of computers, and data processing methods made it possible to identify heterogeneous areas of the field and learn to manage them.

    We at OneSoil, for example, use Sentinel-2 satellite images with a resolution of 10 meters - the device itself was launched only in 2015. We found the boundaries of all fields in Europe and the USA with the help of a neural network - while the concept of neural networks itself began to develop only in the 2000s, partly due to the appearance of graphic processors of sufficient power.

    But it is worth considering several more factors. First of all, farming is really difficult. A plant is a living organism, which is influenced by a huge number of factors. Secondly, farming is still not very fashionable among young people, which means that not many people are interested in and developing these technologies. For a long time, agriculture was considered the lot of peasants, while the world's population has been actively moving to cities since the end of the 19th century. But agriculture is becoming more and more technological and attractive. For example, according to official US statistics, the number of farmers under 35 in the country is growing. In addition, agriculture can be very profitable, because the situation will change.

    Too expensive


    - Unfortunately yes. Tractors, seeders, sprinklers and other equipment with on-board computers cost hundreds of thousands of dollars. Field analysis in commercial online platforms - from $ 2 per hectare, and agrochemical soil analysis - from $ 10−20 for one sample. I think that application developers and technology manufacturers feel that their products are unique, which means they will be bought anyway. On the other hand, the point is the high price of the development itself. Research, design, testing, implementation - there are no beaten paths in precision farming, which is why the price of the final product increases.

    How to live with it?


    Wait and use free apps. As competition among agricultural companies grows, applications and equipment will become more accessible, but right now the farmer needs to invest a considerable amount in the modernization of his farm. We decided that we would make free applications for farmers. Our web platform and mobile application help to get information on the fields - borders, crops, vegetation index, weather forecast, productivity zones for calculating fertilizers in a couple of minutes. We are for affordable solutions for everyone.

    I want to show that technology can be understood. They should benefit everyone: not only large farms that have money for modernization, but also medium and small farms.

    Lack of standards


    - Sometimes they talk about poor synchronization between technology and applications from different manufacturers. This problem still exists, but is gradually disappearing. The turning point, from my point of view, was the emergence in 2001 of the ISOBUS protocol, which allows the exchange of data between the tractor and seeders, sprayers and other equipment. Today, this protocol is supported by more than 200 companies and institutions .

    How to live with it?


    Manufacturers of hardware and software are gradually realizing that products that are closed on themselves have no future. API services appear that allow you to work with data in different formats. For example, in 2016, Trimble and John Deere announced the integration of their applications - this greatly simplifies the life of farmers who have the equipment of both companies.

    But there is another problem - the incompetence of equipment dealers. Their task is to sell their goods, dealers do not care how the farmer will then use this equipment. Sometimes dealers cannot even explain how the seeder or tractor works, and often farmers have no one to ask their questions. I hope that the situation will change soon. Probably, equipment manufacturers will open up something like support centers that will accompany farmers in the process of introducing precision farming - and then the need for equipment dealers will disappear.

    All this is just noise


    - It is often said that there is too much noise around precision farming. After all, agricultural technologies have been developing all the time, what's new? Like, now everyone says “artificial intelligence”, “machine learning”, “neural networks”, but at the same time many applications work poorly, the data are inaccurate, and the benefits for farmers are not obvious. From my point of view, there really is some noise. As one of the indicators, there is a constant growth of investments in this sphere: last year's report of the AgFunder fund says that investments in AgTech for the year grew by 43% and amounted to almost $ 17 billion.

    How to live with it?


    As always, when hype goes into decline, some companies will disappear, and the rest will learn how to make simple and necessary applications. If you are engaged in precision farming - seize the moment. Build a reputation, look for investments and make useful applications that survive hype. If you are a farmer, carefully monitor the technologies and learn to understand them in order to distinguish fashionable from useful. If you want to farm in the 21st century, you will have to do it anyway, so it's better sooner rather than later.

    Also popular now: