Articles & stories

The Digital Farmer 

San Francisco-based the Climate Corporation is reinventing farming. Sensors and robots ensure larger harvests with the use of fewer resources. This will help both large farmers in the US and small-scale farmers in Africa.

For thousands of farmers in the Midwestern United States, the day begins not by looking out their window, but rather by looking at their smartphone. There they can see their fields listed like a music playlist: weather data for every square kilometer, the current growth stage of their corn or soy plants, and an interface that shows in green, yellow or red where the farmer should plow, fertilize, spray or harvest today. As the farmer sits in his tractor, data from the cloud controls where seeds are planted, fertilizer spread or pesticides sprayed, and then this information is immediately stored – row by row – in the network.

Sensors: The Eyes & Ears of Farmers

In the data-driven world of agriculture, an extensive network of sensors on the ground and in the sky are becoming the electronic eyes and ears of farmers so they can increase income and reduce costs. Experts predict a "green data revolution" that will ensure that the population, soon to reach nine billion, is fed – using fewer chemicals and other resources, such as fuel and water.

One pioneer of algorithmic agriculture is the Climate Corporation., which is based in San Francisco, far from the endless grain fields of Iowa and Indiana. It was founded by technology specialists who wanted to bring user-friendly design and the progress made in data analysis and machine learning to the farm. The company, which was founded in 2006 and acquired by agricultural giant Monsanto for 930 million US dollars two years ago, has built a finely-woven network of weather data for the entire US and continuously enhances it with services and products. "Farmers have, until recently, relied mainly on their experience and observations," says Jeff Hamlin, who has been with the company from the start, and today holds the title Director of Customer Success. "But it's impossible to remember the details of so many variables." the Climate Corporation. wants to process this flood of data for each parcel so farmers can make better decisions faster. In doing so, the company is going far beyond the way technology has previously been used in farming, which saw utility vehicles connected more precisely using GPS, but which meant data was stored in onboard computers or on USB sticks rather than being networked intelligently.

Simulations for Every Field in the US

Originally, the company only collected weather data to protect companies in a wide range of industries from unexpected events, such as rain, drought or frost. But Hamlin soon recognized that the business in agricultural data had much more potential. Today, the Climate Corporation. combines measurements from around two dozen public and private sources – from the National Weather Service to satellite images – and uses them to calculate some 10,000 simulations daily for each individual field in the US.

The free version allows farmers to keep an eye on more than 20 million hectares of land, which corresponds to about one-third of all corn and soybean fields in the US. Some of them use a fee-based premium version, which provides more detailed information and decision-making support, such as optimizing the use of nitrogen fertilizer or monitoring the status of a field with aerial photographs displaying 55-meter grids. "This allows them to recognize problems that they wouldn't even see as they drove past their fields," says Hamlin, who cannot, however, provide specific figures for increases in yields or reductions in cost.

Connecting Hardware and Software

This vision is nevertheless ambitious. Rather than simply analyzing sensor data, the Climate Corporation. also wants to operate the hardware in the field. A product called Precision Planting already controls the amount of seeds or chemicals used in 43 countries, from Argentina to Zambia. "When hardware and software work together," says Climate's Head of Data Science Erik Andrejko, "seeds and seedlings can be deployed as precisely as if they were being printed on the landscape." Thus, the data from each field not only determines what is planted or harvested when and where – it also improves calculation models, which are used to optimize weather forecasts and processes. Using the same feedback concept, technology companies like Google continuously refine search methods and speech recognition. The Climate Corporation. is constantly expanding its portfolio so it can gather more and a wider range of data. Recent acquisitions include the Chicago-based startup 640Labs, which turns every utility vehicle into a networked data collector, and Missouri-based Solum, which analyzes soil samples right in the field.

Is Progress the Solution?

Data-driven agriculture is still in its infancy. Information sources and technologies vary widely from country to country and are not necessarily suited for small operations as they have historically emerged in Europe or Africa. In addition, not all standard equipment is compatible with this networked feedback cycle, and new devices would be – in contrast to accessing the latest data on a mobile phone – cost-prohibitive for small farmers in poor countries. Providers such as the Climate Corporation. have therefore united to form the Open Ag Data Alliance to create uniform standards.

However, experts such as environmental scientist Jesse Ausubel of Rockefeller University are convinced that technological progress has already solved the problem of how to feed a constantly growing population. "There is still a lot of flexibility for increasing yields. We will soon witness a re-greening of the earth," says the researcher. He points to statistics, according to which, at least in the US, harvests have increased non-stop since 1940, while the amount of land farmed and input factors, such as chemicals and water, have fallen. "These trends are not a US phenomenon – they'll be repeated by many countries around the world in the coming decades." The UN World Food Programme is also focusing on the growing importance of finely connected agricultural data and has, among other things, set up an early warning system for famines, called FEWS, as well as a network of automatic weather sensors in countries such as Uganda.

Drones Provide Better Data

IIT specialist Erik Andrejko believes data-driven agriculture will establish itself around the world – from satellite-controlled harvesters in the Midwest to subsistence farmers in Africa who can access weather data and market prices on their mobile phones. What components are still missing to bring us closer to this goal? "I would like clear rules for the use of drones because we'll continue to get a greater amount and more precise data with them," says Andrejko. "And I believe robots will play a big role in the future." While the farmer studies his fields on his smartphone, the field workers of tomorrow will be able to independently pinpoint those patches where weeds are growing and remove or spray them in a targeted manner.