PICK OF THE CROP: Machine manufacturer John Deere makes precision agriculture solutions.

Farmers have always sought ways to ensure success in an activity plagued by risks and uncertainties. The ancient peoples of the Nile Valley used the position of the stars and the phases of the moon to determine when crops should be sown and harvests reaped. Planting by the signs still has its adherents today. The Farmers' Almanac, an annual periodical that provides astronomical data and astrological advice to American agriculturalists, published its 200th edition last year.

Most commercial growers have moved on to other methods, however. And increasingly, those methods involve big data, advanced analytics and the internet of things.

Picking the right day to put seeds into the ground is only one of many decisions that determine the ultimate success of a crop. To get food onto our shelves and money in their pockets, farmers must juggle a host of variables, seeking to maximize yields while minimizing the cost of inputs such as water, fertilizers and pesticides. It's a tricky balancing act. Give plants too little and they won't produce, but excess inputs don't just drive up production costs, they can also depress output and lead to lasting damage to soils and the wider environment.

To compound these challenges, the right combination of inputs can vary significantly over time, depending on the weather and the presence or absence of pests and diseases. And within the same crop, variations in local soil and shade conditions, or differences between plants, can mean that the right inputs for one individual are too much or too little for another.

DOWN TO EARTH: Photographs from space can reveal information about soil condition.

Farming from space

Like their ancestors, today’s farmers are turning to the sky for solutions to these challenges. But they rely on constellations of satellites rather than stars. High-resolution satellite imagery and GPS are key enabling technologies for the development of precision agriculture, a new approach that allows crop inputs to be tuned to match the needs of specific locations and even individual plants. 

Photographs taken from space using multispectral imaging techniques can reveal a lot of useful information about the condition of crops or soils, from coverage density to canopy moisture levels. The latest generation of commercial imaging satellites, like DigitalGlobe’s World-View-3, launched in 2014, can produce images down to a resolution of a third of a meter. That’s sufficient to differentiate between crop rows, spot rocks and tree stumps in meadowland or count individual cotton bushes in a plantation. High data transmission rates mean today’s satellites also work faster than their predecessors, while advances in automated image processing are reducing the time and cost required for the creation of images. That matters in agriculture, where farmers need to respond quickly to the signs of developing problems.

Back on the ground, tractors and other farm machines equipped with GPS receivers can be guided precisely and repeatably. That allows inputs to be delivered more accurately, for example by preventing overlaps when sprayers pass up and down a field. Integrating geolocation data into the control of implements can ensure that pesticides or fertilizers aren’t released close to water courses or ensure that furrows are cut at precisely the right depth and angle to maximize irrigation efficiency.

It is the combination of these technologies that provides the largest benefits, however. With the aid of data from satellite images and other survey tools such as drones, farmers can adopt variable rate application techniques, setting machines to automatically adjust the quantity of inputs applied according to the needs of specific parts of the crop. Agricultural technology provider Trimble says that variable rate application can reduce crop inputs by 10 percent, while the combination of accurate land forming and precise irrigation control can cut water requirements by as much as 30 percent.

Precision agriculture 

The commercial potential of precision agriculture is attracting interest and investment from many sides. Some of today’s major players come from the technology world: Trimble was founded in 1978 by three former Hewlett Packard employees who wanted to develop commercial applications for the then-new GPS system, for example. But big names from the agricultural equipment world are also heavily involved.

John Deere, the long-established U.S. manufacturer of agricultural machines, acquired GPS technology company NavCom in 1999 and has gone on to develop a comprehensive range of software and hardware solutions for precision agriculture. The company also offers remote condition systems and services for its machines, as well as software products designed to help farmers track assets, schedule activities and generally keep their businesses under control.

Across the Atlantic, 365FarmNet, a subsidiary of German agricultural equipment maker Claas, has developed an online platform that allows farmers to integrate and manage data from multiple sources. The system, which integrates with equipment from a number of manufacturers, is designed to simplify and streamline precision agriculture activities. Claas says that the platform will allow farmers to use data from vehicle-mounted crop sensors or satellite imagery to program variable rate fertilizer application systems with just a few mouse clicks, for example.

Agriculture is proving to be fertile ground for further innovation. Drone manufacturers are developing products that can provide more detailed images of crop condition than satellites, for example. The largest models can even carry tanks of fertilizers or pesticides for selective spraying. Research company Global Market Insights believes the market for agricultural drones will exceed $1 billion by 2024, with more than 200,000 units in operation worldwide.

The sector is also exploring the potential of artificial intelligence technologies. In September 2017, John Deere announced the $305 million acquisition of California-based Blue River Technology, which has developed computer vision and machine learning technology for agricultural spraying equipment that can identify weeds growing within a crop and target them with herbicides. Even in one of the world’s earliest professions, change is afoot. —  Jonathan Ward

Published: January 2018

Images: Visual Services-East Moline/ 2015 Deere & Company; 365farmnet.com