Could a hacker attack bring agriculture to a standstill and endanger the basis of our food supply? Even if that still sounds like a futuristic horror scenario today: The more digitalization advances, the more it is necessary to look at its flip side. In his current Humboldt reloaded project, economics student Philipp Reister is therefore looking at an approach to make smart farming resistant to the failure of Internet and communication structures. And he's doing it all from home under Covid-19 conditions...
A few weeks ago, the parcel arrived at Philipp Reister's home. The contents: a Raspberry Pi 4 mini-computer and sensors for plant cultivation, which can be used to measure air temperature and the humidity of air and soil.
Gardening is one of the hobbies of the Hohenheim economics student. He gets his passion from his grandfather, whom he helped in the garden when he was a boy. Today he nurtures and cares for his own small tomato colony. This is also where he wants to try out the supplied sensors.
On the other hand, Philipp would not exactly call himself a computer nerd. And that is precisely what qualifies him for his task in the interdisciplinary Humboldt reloaded project: after all, the farmers who will later be using the mini-computer are not IT experts per se. Philipp wants to put himself in their shoes as much as possible.
Opportunities and risks of digitalization
More and more farmers are already relying on intelligent agricultural machinery. With the help of sensors and data from the cloud, they can apply exactly the amount of fertilizer, pesticides, or water for irrigation that is actually needed. This is good for the yield and for the environment.
The potential of digitalization in agriculture is huge. In the future, field robots could even take on more and more tasks autonomously. But if dependence on data increases, there are also risks: What, for example, can be done in the event of a power failure or internet connection failure?
"If all data is hosted centrally, the system is very vulnerable. And it's also a lucrative target for hackers," Philipp explained. "Of course, corresponding Internet failures are devastating in any industry. But our food depends on agriculture. Society cannot really accept this risk."
Decentralized data storage on the farm box
This is where the mini-computer comes into play, which Philipp is currently testing in his home network. The idea: In order to arm themselves against short or even longer-term failures, farmers should in future store the most important data decentrally on their own farm box, which could continue to be operated with solar power if worst comes to worst.
"The Humboldt reloaded project is concerned with the question of what such technology must be like in order for farmers to actually use it. In the long term, it provides more security, but in the short term it means additional effort," Philipp stated.
He receives support not only from his supervisors Evelyn Reinmuth and Sebastian Bökle in Hohenheim, but also from Daniel Eberz, a Hohenheim alumnus who now works in the Rural Service Centre Rhine-Hesse-Nahe-Hunsrück in Bad Kreuznach.
This is where a major project on operational data management and resilient networking in agriculture is coordinated ("GeoBox Infrastructure"). Among the cooperation partners are the TU Darmstadt and the German Research Center for Artificial Intelligence. The project is funded by the Federal Ministry of Food and Agriculture.
Research-based learning under Covid-19 conditions
"Distance doesn't really play a big role at the moment. After all, due to Covid-19, the Humboldt Project has to be completely digital anyway," said Daniel Eberz. "We regularly exchange information via video conference and chat. And I must say that I am really impressed by how hard Philipp is working. The subject is not without its problems and one must not forget that he is still working on his Bachelor's degree."
When organizing the collaboration, the team relies on SCRUM, a framework for agile project management that originally comes from software development.
Evelyn Reinmuth explained the basic idea as follows: "It's all about breaking down a complex topic into work packages that are as small and manageable as possible, clarifying responsibilities, providing regular feedback, and repeatedly determining the next steps together. On the one hand, this creates flexibility to make adjustments and, on the other, provides the necessary support for independent work. We use GitLab to assign tasks, manage document versions and as a ticket system."
Philipp finds the working method very accommodating: "If you have 2-3 smaller tasks instead of one large one, it's easier to tackle them. But most of all, I was surprised by the intensive feedback and the great working atmosphere: We were all immediately on a first-name basis and I was treated on an equal footing right from the start. So it's not difficult to ask questions when you're stuck."
In the meantime, Philipp has decided to take a closer look at the topic from an economic perspective in his Bachelor's thesis. Among other things, he plans to conduct a survey of district offices and machinery rings. The focus will be on measures to avoid digitalization risks in agricultural primary production.
His conclusion on Humboldt reloaded: "I have never done anything nearly as interesting in my studies!
