When it comes to digitization, most people think of startups, IT and industry. Yet the digital transformation is radically transforming one of the world’s oldest trades. And for farming, the biggest changes could still be ahead.

Text Thomas Schmelzer Photos Marian Lenhard Illustration Julian Rentzsch

Nineteen years ago, when Andrea Rahn-Farr began working on the dairy farm in Büdingen, Germany, she’d hardly ever heard the word digitization. The first mass-produced mobile phones with high-resolution displays were just sweeping the market, while meanwhile Rahn-Farr spent almost four hours every day herding her cows to be milked. “The work was fairly exhausting,” Rahn-Farr recalls as she slips on her knee-high rubber boots. Now 48 years old, she’s still running the farm, yet aside from the rubber boots, there’s little left of the farming world she started in.

Today, instead of 160 cows, Rahn-Farr and her husband tend 430 dairy cows and 400 calves on their farm. No longer do they need to tediously lure the cows to the machines to be milked; the animals visit the milking robots completely independently. Instead of having to muck out the stalls, they have a mechanical scraper to clean up after the animals every hour. The analogue farm has become a digitized, highly networked agricultural facility, where surveillance cameras monitor the calves’ development and a feeding robot mixes fresh feed for the cows.

As Rahn-Farr walks across the muddy courtyard in her rubber boots, she talks about how the work has been changing over the past years. How she’s trained to become a bit of an IT expert and now, every morning, checks the stall info on her smartphone. Now she has reached the core of the premises. The operation’s most modern barn is located on a knoll where the dairy cows are kept. Three years ago, Rahn-Farr and her husband invested 2.5 million euros in the building. The new bovine accommodations needed to be animal-friendly and comfortable. At the same time, Rahn-Farr wanted the business to run as efficiently as possible.

The result is a state-of-the-art building packed with technology. This high-tech facility has as much in common with the idyllic wooden barns seen in milk advertisements as professional soccer has with after-school tournaments. Every cow here wears a tracker on its leg. Circular receivers on the ceiling collect the data. There are also sensors dangling from the animals’ necks that record typical chewing movements. Using this data, Rahn-Farr knows exactly where every cow is and how much it has eaten, when it has moved and whether it has been lying down for too long. As soon as a value deviates from the average, the system sounds a warning. Even the roller shutters on the sides of the barn automatically open and close depending on the weather. “Our robot barn,” says Rahn-Farr, grinning.





What sounds like science fiction is actually science fact on the farms of today. Cows that wear more sensors than most people are just as common as milking facilities that can evaluate the milk quality fully automatically and farm owners who talk shop about data interfaces. Almost completely unnoticed by the public, digitization has reached an industry that many still think of in bucolic terms with the image of a simple, nature-loving farmer. In one of the world’s oldest and most tradition-conscious sectors, of all places, is where the digital transformation is revolutionizing business models and workflows.After spending centuries trusting their senses, farmers are now relying to a greater extent on sensors, cameras and satellite imagery. While until recently they averaged the data and used their wealth of experience to determine when fields needed to be fertilized or cows impregnated, today they’re increasingly leaving these types of decisions to algorithms and artificial intelligence. Satellites have long been passing high over fields and sending information to farmers about the optimal time for harvesting. Field robots scatter perfectly proportioned amounts of fertilizer based on real-time calculations. And tractors automatically navigate across fields using GPS signals.Four years ago, sensor technology, electronics and software accounted for 30 percent of value creation in agriculture — the share in the automobile industry at that time was just 10 percent. According to a study by Bitkom, Germany’s association of digital and IT companies, one of every two farmers is using digital technology today. Goldman Sachs estimates the market volume for digitization in agriculture at about 240 billion US dollars annually. And the US Department of Labor forecasts that more than half of the jobs in agriculture could be automated.


Rainer Winter from the German Agricultural Society (DLG) has been observing the rapid digitization out on the fields for years. Yet he believes that it’s really only just starting to pick up speed. Many of the technologies being used have been around for some time now, but they long remained too expensive for most farmers. It wasn’t until the past several years that sensors, computing power and software became inexpensive enough to make the leap from scientific laboratories out into the fields. “Since then, we’ve been experiencing a quiet but continuous revolution,” Winter explains.

The farming profession has changed dramatically over the past two decades as a result of these developments, Winter notes. Anyone running a farm today must be an all-in-one manager, taking care of accounting, the fields, animals and technology. Yet at the same time he believes that the real revolution in the industry may still be ahead; with every farmer who invests in digital technology, the mountain of data grows—and with it the chances that the forecasting quality of self-learning algorithms improves. The more accurate the forecasts, the more worthwhile the investment for farmers. Although Winter remains certain that the relationship between the farmer and the animal, as well as the farmer’s experience, will continue to be paramount, he also believes: “Artificial intelligence will become increasingly important for farming huge tracts of land or larger animal herds.” It is in agriculture in particular that it becomes clear how the digital transformation will affect traditional industries, how it can be implemented, what the pitfalls are and where the future is already succeeding in the here and now.

While Rahn-Farr is making her way among the cows in her automated barn, she talks about the reactions she repeat­edly gets to the level of digitization on her farm. “When I show my barn to groups of visitors, they say that they’d prefer to have everything like it used to be.” Many people just assume that traditional husbandry is automatically better for the cows. “But it isn’t true,” Rahn-Farr explains. “The data that I get from the system help me see what every cow really needs.”

Along the way to the other side of the barn, cow number 2514 walks behind her and nudges her again and again. “She could be in heat,” Rahn-Farr says. Animals in heat move more. Later, around noon, Rahn-Farr checks the computer to see if her suspicions about cow number 2514 are correct. And, in fact, the screen shows that warning as well as five others that indicate the cow is in heat. The graphics show the movements of cow 2514 over the past several days. Initially the data curve is constant, then suddenly it rises. Rahn-Farr discovered it herself during the inspection today, she says. “But with the number of cows we have, it’s something that can easily slip by unnoticed.”


As in other industries, the transformation in agriculture is happening gradually. Digital technology supports farmers, but they continue to make the important decisions themselves. In the end, Rahn-Farr herself decides which cows will be impregnated. The digital indicators help her with the decision and increase efficiency. Thanks to the automation on her farm, Rahn-Farr and her employees can spend just 30 hours tending to each cow per year, while it used to be 40. With the same number of employees, Rahn-Farr can look after significantly more cows.




Andrea Rahn-Farr is a farmer — and meanwhile also a bit of an IT expert.
On her farm, sensors, robots and algorithms do a lot of the dirty work for her.


Michael Clasen, a professor for agricultural informatics at the Hannover University of Applied Sciences and Arts, finds that technology’s purely support role will expand over time. Thus far, technological advances in the form of sowing machines, semi-autonomous tractors and smart combine harvesters have replaced farmers’ muscle power in particular. “Now, however, algorithms and artificial intelligence are beginning to trump the farmers’ expertise and know-how in part,” Clasen says.

To get a feeling for the transformation occurring out in the fields, Clasen likes to compare agriculture with small to medium-­sized enterprises: similar turnover, similar numbers of personnel. “If you place these two areas next to one another, agriculture is much further along in digitization,” he explains. It’s even taken on a pioneering role in many areas.



To get a feeling for the transformation occurring out in the fields, Clasen likes to compare agriculture with small to medium-­sized enterprises: similar turnover, similar numbers of personnel. “If you place these two areas next to one another, agriculture is much further along in digitization,” he explains. It’s even taken on a pioneering role in many areas.

If the speed of change remains constant, Clasen can even imagine fields that cultivate themselves with a sort of digital consciousness a decade from now. In the scenario that he describes, the intelligent software recommends to the farmer what she can profitably cultivate out on the fields. After she makes her decision, the fields then take care of themselves using sensors and satellite imagery to decide which fertilizers they need, when they should be harvested and which pests need to be controlled. Autonomous robots carry out the commands out on the field. “The fields could develop a digital consciousness,” Clasen says — a software-supported, digital representation.




Daniel Bohl isn’t quite so far along. In Trams, 25 kilometers south of the Baltic Sea, he is still driving his tractor across the fields himself, towing the fertilizer spreader for nitrogen fertilization. Before him the green and yellow fields stretch off to the horizon, interrupted by the occasional clump of bushes or trees. For the Wariner Pflanzenbau co-operative, Bohl, aged 47, is responsible for 2,950 hectares of land between Schwerin and Wismar, about twice the size of Los Angeles International Airport. The co-operative harvests around 13,000 tons of grain and twice as much corn annually. The fields stretch out 10 kilometers long and 15 kilometers wide.

Eighteen years ago, when Bohl completed his studies and began working at the co-op, he’d also never heard any specifics about the magic word “digitization.” The farmers used to eyeball which field needed which fertilizer at what time. “You just walked up next to the field and someone said that the wheat there in the back was growing quite well and pretty badly here at the front — and then you fertilized accordingly,” Bohl explains. Of course even then people took notes, compared them to their own experience and drew their own conclusions. “But with an area as huge as the one we’re tilling, it’s really difficult to keep track of all of it.”

Today Bohl is supported with sensors and computer chips. Before he got into the tractor this morning, he measured the green color of his winter wheat crop — the factor used to determine the nitrogen requirement of the plants — and loaded that data into his tractor’s onboard computer. Now he’s just keeping the tractor on its path. Sensors on the tractor’s roof scan the wheat stalks in all directions and compare the data with the reference value. Milliseconds later, it sends the results to the attached spreader, which automatically regulates the amount of nitrogen fertilizer being scattered on the field. “It allows us to fertilize much more accurately and also save nitrogen,” Bohl says.




Sensors scan the fields, then the onboard computer uses the data to calculate the optimal amount of fertilizer. The farmer’s task here is that of supervisor.

Two hours later, Bohl is in his office looking at the results of the nitrogen application. His PC monitor shows an outline of the area he drove over. The software has laid blue gradations over it. “The more intense the blue, the more nitrogen that was scattered,” Bohl says. He just spread 4,293 kilograms of fertilizer on 23.5 hectares of field. Yet there are enormous variations in the specific amounts. At some points, the spreader distributed 20 kilograms of fertilizer, while in others it was five times as much. “Even with decades of experience, you can’t recognize and control this as precisely with the naked eye,” Bohl explains.

The principle Bohl used to apply nitrogen works almost exactly the same for the long-term fertilization of the fields with potassium, phosphorous and lime. Bohl doesn’t use his own sensors for this, but instead a soil analysis carried out by a service provider. But he also feeds the provider’s data into his tractor’s computer.

For the corn crop this year, he’s going to try out the harvesting app provided by his seed supplier for the first time. Using satellite imagery, the app evaluates the dry mass of the corn out on the field. Bohl will eventually just receive an electronic message telling him when the optimal time for harvesting has arrived and the harvesting combines should be sent out into the fields.


This great relief for farmer Bohl and dairy farmer Rahn-Farr is also a huge market for the manufacturers of digital systems. In order to secure their position, companies have set up their own, often incompatible standards — to the annoyance of the farmers. “Today there are thousands of systems for thousands of tasks,” Oliver Martin complains. He used to run his own farm and today advises farmers on digitization with his company Farmblick. Martin believes that the challenge of the future will be to network the individual systems and to create more practical standards for software and hardware applications. Only then will scenarios in which artificial intelligence independently manages fields or stables become reality.


Daniel Bohl relies on the data’s wisdom and can access it at any time using his smartphone.
Despite this, he still continues to rely on his experience and intuition.


Back in Büdingen, Rahn-Farr also had to solve these types of problems with standards. Her milking robots and the sensor system come from two different manufacturers. Initially there was radio silence between the two systems. By now, digital communication has been established and Rahn-Farr can imagine an even deeper networking in the future. Along with chips for the necks and feet, there are also sensors that deliver information directly from the cow’s stomach. “Maybe at some point I’ll be walking through the barn wearing augmented-reality goggles and will be able to see the exact parameters for each cow where it stands,” she says.

Even for farmer Bohl, the digitization of his farm is far from over. He plans to use the pinpoint fertilization system also for pesticides in the future. Forecasting systems will follow, once he’s collected more data. Despite all the progress in the field, he’s not worried that digital technology may soon make his profession redundant. “Digital technology is an enormous help,” he says, “but, in the end, we still always have to use our own brains.”