Scientific and technological advances are opening up new possibilities for farmers around the world. The networked digital farm of the future is already making agriculture more efficient and sustainable today.
The face of agriculture has changed: tractors and combines are highly automated and crammed with a vast array of sensors. And already, unmanned aerial vehicles (UAVs) and even orbital satellites are becoming indispensable. The delivery of highly detailed field observations straight to a farmer’s tablet computer will soon become a standard in agriculture. Certainly the capabilities exist now. “Current commercial satellite imaging allows us to analyze a single patch of land at a resolution of just 30 centimeters, with UAV’s and manned aircraft providing an even finer scale”, explains Michael Schlemmer, Project Development Manager Field Phenomics at Bayer. “This means we can now diagnose the condition of a crop and nearly distinguish between single plants all the way from space.” Specialists around the world are looking into the many different possibilities digital farming has to offer. And in many places the new systems are already supporting agronomists out on the fields.
We can now diagnose the condition of a crop and nearly distinguish between single plants all the way from space.
Starting in 2014’s growing season for example, new varieties of soy were tested throughout the American Midwest. Several farmers participated in the study to compare new varieties with existing ones. “Thanks to remote sensing techniques we were able to take stock of the crop’s development much more efficiently”, says Marshall Beatty, a Regional Agronomist with Bayer USA. The trial fields were analyzed with an airborne multispectral sensor mounted on a UAV. The resulting infrared images, for example, immediately showed which areas in the field needed more attention.
“Long before the stress factors that endanger plants can be seen in the visible spectrum, they appear in the near-infrared”, explains Beatty. This is because remote sensing technology is different from what we typically use. In Google Earth, for example, we see true color images that are similar to digital photographs. Airborne imaging sensors and sensors orbiting Earth on a satellite however, measure the planet’s surface reflected and emitted radiation in a wide range of wavelengths or “bands”. While the human eye can only perceive light in the wavelengths of the colors red, green and blue, remote-sensing tools can capture additional ranges, for instance near- and short-wave infrared.
These non-visible bands can reveal a wealth of information about the condition of crops such as their overall vitality – a measure of their chlorophyll content and structure. “The more chlorophyll a plant contains the more blue and red it absorbs. The reflected green light is what we will see”, says Beatty. Since healthy plants have higher chlorophyll content and vigor, they also produce more plant material. This results in an increase in near infrared reflectance. It is this increase that is associated with the red color of the infrared images. As a result, during the 2014 soy trial the farmers and crop specialists knew exactly what part of the field to focus on, even before they or a land-based machine had even gotten close to it.
From Precision Farming…
Nonetheless, land-based agricultural machines are also adding to the growing data pool. Most tractors, combines and sprayers roaming today’s fields are also tracked via satellite positioning systems (GPS) to monitor and store data on their individual operations. They can potentially be controlled remotely without even a driver on board. These high-tech machines are equipped with various sensors that collect data on the crops’ health during the growing season and its final harvested yield, as well as the field’s soil composition and topography.
Long before the stress factors that endanger plants can be seen in the visible spectrum, they appear in the near-infrared.
This data is used to generate detailed field maps and also stored in databases for retrieval at a later time. “Thanks to GPS, we’ve been able to pursue precision agriculture for quite some time now”, says Schlemmer. “So today, we can add this historical data to the equation along with real time data acquisition for better field level situational awareness.” Integrating high precision in-season data with data from past seasons is bringing an entirely new dimension to modern agriculture. It gives farmers the possibility to foresee the upcoming harvest’s yield, manage a variety of inputs, and lets them react to certain changes earlier to prevent potential losses. Another way to factor in historical data is by integrating information on pest occurrences. This even allows farmer’s to foresee the spread of diseases or pinpoint them at a very early stage due to already known spread patterns. “The next big step is combining all these data sets with a system that produces and manages the information smartly in a more automated fashion”, says Schlemmer. And having all of it accessible on a mobile interface, such as a tablet computer, would even allow farmers to make their vital decisions directly out in the field.
… to Decision Farming
Several start-ups around the globe are already trying their luck at harnessing and making sense of these gigantic data masses to tackle the challenges of food security and prepare agriculture to be more efficient for the future. One such start-up is green spin, a company based in Wuerzburg, Germany. “We believe that every branch of agriculture can be made more profitable if key information is made available at the right time”, says Clemens Delatrée, the company’s CEO. “Today a great deal of data already exists, and if it is used effectively, it can facilitate decision processes in agricultural management”, he add.
By integrating every aspect of a farm’s processes, such management tools can really facilitate a farmer’s life. They could eventually span all production areas on a farm, from crop growing to cattle and pig breeding. They might even work in unison with partner applications from the entire agro business: a seed company can give suggestions on the suitability of a crop variety to the matching plot of land, another company delivers highly reliable weather forecasts and yet another one supplies details on the soil composition on the farmer’s field. These systems will also integrate data that is generated by satellites and agricultural machines out on the field. “By deducing different parameters from satellite-images and observing them over a season, as well as combining them with weather and soil data, it is possible to precisely calculate the biomass and the yield of a field”, explains Delatrée.
At the Smallest Scale
All this information amounts to very large quantities of data and requires not only storage but also a lot of computing power. “The hardware already exists but it is not in everybody’s laptop”, explains Schlemmer. And while data management remains a major challenge, the efforts of companies like green spin are an important step forward. “The purpose of Digital Farming is to adjust farming practices to the smallest necessary scale. The scale we have reached today is already improving yields and making agriculture a little more sustainable. In the long run, Digital Farming will completely revolutionize agricultural technology – bit by bit”, Schlemmer concludes.
In the long run, Digital Farming will completely revolutionize agricultural technology – bit by bit.
The Networked Farm – More Data for Efficient Processes
Collecting Data Throughout the Season
There are thousands of different soil types around the world. But even in a region or a single field, a soil’s quality can vary greatly. The more a farmer knows about his soils, the better he can decide which varieties to sow in a certain area to gain optimal yields.
As the crop grows, sensors on the tractor can detect the health of the plants by measuring their photosynthetic activity. Similar measurements can also be done from space via unmanned aerial vehicles (UAVs) or stellites to cover a larger area while still delivering high resolution.
At the season’s end, the combines harvesting the crop have built in yield monitors that record fluctuations and automatically generate a detailed yield map of each field. This data can be used in the next season to further optimize cultivation.
“Transparency and Trust will accompany success”
Why does data security also play a role in agriculture?
Modern agriculture collects many crop parameters, for example during the growth and harvesting period. This process data is unique for every farm – and therefore worthy of protection. Based on this data it is possible to draw conclusions about how a company conducts its business efficiently and productively. It is understandable that every farmer wants to keep control of this data in order to manage and optimize his processes independently.
Who ensures protection of the farmers’ sensitive data?
First of all, the manufacturer of the equipment and the systems that can detect and manage these parameters. However, there are currently no established standards. In the medium term, the market will regulate what farmers need and demand from data security. And ultimately, the farmer as an entrepreneur is himself responsible for protecting his data. Nevertheless, it has become difficult for the layman to keep track of and to ensure adequate data protection. Thus, service providers are also required to provide appropriate technologies.
What benefits do farmers have if their data is anonymously stored in databases?
The benefit comes from the offered service, for example from the data analysis that the farmer receives. Only large amounts of data allow deriving statistically relevant trends and correlations. If there is a shift in climate, for example, it may be useful to use data from climatically similar regions – data that show observed effects on plant growth or certain diseases. So a farmer can also benefit from the information provided by many other individuals.
What else can be done in this area?
First, we still have much work to do concerning the combination of individual data sources and the evaluation of different parameters. There is also much to be done in terms of data protection. Among other things, we are working on a way to better control data usage. One can, for example, provide access to the data only for a certain period, or limit access to only a portion of the measured values. In other words, data usage can be described and controlled in much more detail. This is an important field of research, which enhances trust. Implementing transparency and trust in data usage will certainly accompany the mid- and long-term success of such systems.