“No sector has a greater impact on climate change and biodiversity than agriculture”

“No sector has a greater impact on climate change and biodiversity than agriculture”

Series: The societal impact of UM research

12-06-2024 · Interview

In today’s world, the societal impact of research findings seems more important than getting published in an academic journal like Nature or The Lancet. What impact has research conducted at UM had in recent years? This week: how UM computer scientists are developing software for farmers.

A camera system that can detect pests or diseases in fields. GPS trackers that monitor livestock in remote areas. A water-saving irrigation system that uses sensors to measure local rainfall, temperature, soil moisture and crop water needs. All these are examples of smart farming, the use of data, modern technologies and artificial intelligence (AI) in agriculture.

“The main goal is to make agriculture more efficient, and therefore more environmentally friendly”, explains data scientist Christopher Brewster. “There’s probably no sector that has a greater impact on climate change and biodiversity loss. The IT world claims that computers are the solution. They can help reduce fertiliser, pesticide or water use, for example. It’s better for the environment and economically attractive for farmers. It also meets the growing desire of governments to monitor agriculture and of farmers to automate their processes due to the farming labour shortage.”

But how can you best use technology in agriculture? This is the question Brewster, who also works as a researcher at the Netherlands Organisation for Applied Scientific Research (TNO), is looking to answer at the Faculty of Science and Engineering (FSE), where he holds an endowed chair in the Application of Emerging Technologies. As its title suggests, his work goes beyond mere theory. “I’m particularly interested in solving practical problems.”

Potato fields

Take OpenAgri, an EU project involving several universities and organisations, with Maastricht University taking the lead – and Brewster as the project leader. “We develop software for farmers interested in digitalisation, or IT professionals interested in agriculture. We’re not working towards a final product, but developing separate components for people to use independently, like codes to relay local weather forecasts or calculate how well a crop is growing. This makes digitalisation more accessible. Rather than having to start from scratch, you can build your own program using these existing components.”

The translation into practice is already underway. Five pilot projects were launched this year, each involving several farms in the participating European countries. In Greece, for example, a computer system using high-resolution cameras to automatically detect fungi in vineyards is tested. And Belgian farms are using AI to analyse drone images for pests in potato fields and simultaneously select the right pesticides.

The cloud

“These projects will run for three harvest seasons, and more pilots will follow. Meanwhile, we analyse what works best in practice and meets the needs of farmers. What data should be collected, how and when? And should you process the data on a local computer or online in the cloud? The latter requires an internet connection, which uses more energy, but might produce better results due to greater computing power. The question then becomes: what is the best result that’s still energy efficient?”

Ultimately, the aim is to develop software that gains widespread adoption. “That’s why we want to make everything open source [freely available for modification and redistribution]. I don’t really believe in intellectual property. This way, we hope to create a community of users.”

Utopianism

The approach is similar to that of another EU project Brewster is involved in, DiTECT, which is currently in its final phase. “But it has a different focus from OpenAgri, namely food safety. The aim was to develop systems to collect and immediately analyse food quality data in various ways throughout the production process, from harvest to transportation to the supermarket. This way, we can increasingly prevent people from falling ill or dying.”

The DiTECT project has produced a lot of useful software, says Brewster. Is it now widely used in practice? “Unfortunately, it’s hard to say. The downside of open source is that you don’t know who’s using it.”

Brewster is careful not to be overly optimistic when it comes to impact. “I’m aware of the danger of technological utopianism, the idea that technology will solve all our problems. People have been quick to hype smart farming, but so far, there’s little evidence that it’s actually more efficient or environmentally friendly. Much research still needs to be done on that.”
 

This is the final article in this series on societal impact of UM research. You can read all previous articles here.

Photo: Ellen Oosterhof

Categories: news_top, Science
Tags: impact,smart farming,agriculture,it,programming,data,farms,ai,digitalisation,fse,science

Add Response

Click here for our privacy statement.

Since January 2022, Observant only publishes comments of people whose name is known to the editors.