leaders In header

Data Driven Farming

“The World Food Programme recently forecast that the pandemic could push around 130 million more people into a state of chronic hunger by the end of 2020, due to a range of problems exacerbated by the virus, including barriers to cross-border trade, a limit in supply of labor, and general uncertainties along the entire food supply chain.”

By: Adi Gaskell, Contributor


The coronavirus pandemic triggered the first reversal in global poverty levels in a generation, with food insecurity a major part of that.  Indeed, the World Food Programme recently forecast that the pandemic could push around 130 million more people into a state of chronic hunger by the end of 2020, due to a range of problems exacerbated by the virus, including barriers to cross-border trade, a limit in supply of labor, and general uncertainties along the entire food supply chain.

Data is playing a crucial role in alleviating some of these challenges, with the Covid-19 Earth Observation Dashboard being launched earlier this year by NASA, ESA and the Japanese Aerospace Exploration Agency (JAXA) to bring together satellite data to help monitor the impact of Covid-19 on agricultural production.

A recent study analyzed the harvest of various winter cereals in Spain and highlighted the value of the dashboard.  Winter cereals are mainly grown in the regions of Andalucia, Aragon, Castilla-la-Mancha, and Castilla and Leon.  The satellite data allowed for these harvests to be tracked in near-real time across the entire country.

When analyzing the data with the help of machine learning, they found that the harvesting season in 2020 started in mid-June, which is considerably later than usual for winter cereals in Spain.  It’s a situation the team believes is partly explained by Covid-19, and partly by the weather, with droughts affecting much of Europe during 2019, resulting in early harvests last year.

Data from JAXA satellites have also been fruitfully used to monitor rice fields in California.  The data is providing vital information about the phenology of the rice, including when it was planted, when it will mature, and when it is eventually harvested.  The data allows us to see that in many parts of California, rice has been planted earlier than in the last two years.  The researchers believe this allows us to better understand how nature-related disruptions to farming can affect trade policy and consumer demand.

“Satellite indicators demonstrate the capabilities of monitoring the planting, growth and harvest of staple crops such as cereals and rice at national scales,” they say. “These data are vital in providing timely and transparent information on agricultural production during the COVID-19 outbreak and recovery.”

Smart farming

The project underlines the gusto with which the farming industry has adopted data and analytics in recent years.  Hands Free Hectare is a project setup between tech startup Precision Decisions and Harper Adams University.  It uses a range of autonomous vehicles and drones to automate much of the farming process, from planting through to harvesting.

An early prototype was able to successfully yield nearly 5 tons of barley, and illustrated how entire crops can be grown autonomously without requiring any human input.  This pilot utilized an Iseki tractor to spray, drill and roll the fields, with a Sampo combine harvester then used to tend to the crops.

A drone fleet then provided ongoing aerial support to provide data from multispectral and RGP sensors, while a Scout vehicle provided video footage at ground level.  The vehicle was also capable of taking physical samples that could be analyzed by agronomists to understand what chemical support the crops needed, and when they should be harvested.

The open source project used technology that was already on the market, which the team believe highlights the accessible possibilities for such an approach to work at scale, with the whole project coming in at just under £200,000.

“This project aimed to prove that there’s no technological reason why a field can’t be farmed without humans working the land directly now and we’ve done that. We set-out to identify the opportunities for farming and to prove that it’s possible to autonomously farm the land, and that’s been the great success of the project,” the team says.

Tracking growth

Researchers from Skoltech have shown the potential when data and AI come together to help bring about precision farming.  Their research utilizes a combination of machine learning and computer vision to monitor and optimize crop maintenance.

A 3D camera begins by measuring the growth of the plant to establish a link between the biomass of the plant and the expansion of the surface area of the leaves.  A second camera then captures the increase in leaf area to allow a plant growth model to be built.  In total, the researchers processed around 10,000 images of each plant as it grew, with the aim to develop a recommendation engine for use in fields and greenhouses around the world.

“The key strength of our method is that it’s enough to get 3-D images of each plant species only once. Then you can predict the biomass growth in greenhouses using the simplest cameras. This helps create much simpler and cheaper prediction, monitoring and optimization systems for greenhouses and artificial life support systems,” the researchers say.

Making data open

The potential of data to transform farming is underlined by these examples, but there are strong moves to ensure that any data that is captured is made as openly available as possible.  A paper published by the Open Data Institute, GODAN, Open Data for Development (OD4D) Network and the Open Data Charter makes the case for open data in farming.

The paper identifies fourteen key types of data that it believes will be key for governments to make available, including soil and hydrology data, market data, and land use data.  Each data type is accompanied by use cases that illustrate how open data can be used to enhance food security and sustainability.

The challenge for the open data movement is the growing commercial interest in agricultural data from companies such as John Deere, whose hardware as a service business model is often predicated on the value of the data generated by their machines.  The potential of using data to make farming smarter and more efficient is clear, however, even if a sustainable business model is not yet.

If you loved this article–our ‘Future of’ series will explore new topics released weekly every Thursday–stay tuned!
Strategic Media Partner