A team of Computing Science students has developed a system for dyke monitoring - cost-effective, self-sufficient and flexible. The "Diek un dat" project group has just presented its results.
Early January 2024: Half of Lower Saxony is flooded, and emergency services in Oldenburg are also battling the masses of water. Dykes threaten to break and have to be reinforced with sandbags.
"Back then, dozens of dyke guards had to regularly walk the dykes to check whether they were holding, with more than 200 people on duty in total," says Imke Korte. Together with seven fellow students in the "Diek un dat" project group, the business informatics student has developed a system that could save dyke walkers a few kilometres during future floods - a measuring device that automatically monitors the stability of dykes. The eight Master's students in Computing Science and Business Informatics recently presented their system at a final presentation.
The result of the one-year study project is a digital, self-sufficient measuring system for dyke monitoring, consisting of a central sensor box (GBox) with integrated power supply, data processing and communication as well as flexible sensor technology and a platform for evaluating and visualising the measurement data. "This device can be used to continuously measure the seepage of the dyke during a flood," explains team member Paul Tepe, who is also studying business informatics. The data is transmitted by radio every 15 minutes and presented clearly on a specially developed internet platform - for example in the form of curves and status reports. "This allows the fire brigade, for example, to see at a glance whether the condition of the dyke is critical," says Tepe.
In order to familiarise themselves with the topic, the project team first met with experts from the Lower Saxony State Office for Fire and Disaster Protection (NLBK), the fire service, the technical relief organisation and the NLWKN, the Lower Saxony State Agency for Water Management, Coastal Defence and Nature Conservation. "We first had to build up a lot of basic knowledge about dykes," says Tepe.
The team learnt that the so-called seepage line - the boundary between dry and water-saturated material in the dyke - plays a central role in dyke stability. Its course within the dyke determines whether the structure is still stable or not. During floods, water emerges at the foot of the dyke on the land side at the level of the seepage line, similar to a spring. "The dyke guards can feel this with their hands, the soil is a bit wobbly there," explains Korte. The team therefore developed ideas on how to monitor the course of the seepage line inside the dyke and on the outside.
The students used two different methods for this. One method consisted of localising the exit of the seepage line using several moisture sensors, which look like small swords and are inserted horizontally into the dyke.
With the second method, the group can record the course of the seepage line within a dyke using level sensors, which are also used for groundwater monitoring. However, in order to install such level sensors, it is necessary to drill into the dyke, which is not necessarily conducive to its stability, especially during a flood. The method could therefore be less suitable for mobile systems that are used in the event of a disaster, but rather for fixed stations. "Basically, you would have to plan for such a measuring system when building a dyke," says Computing Science Professor Dr Andreas Winter, who supervised the project group together with his colleague Professor Dr Oliver Theel.
Various practical tests were particularly exciting for the students, for example on the Osternburg Canal and on the River Hunte in Tungeln. They were even able to test their system in a simulated flood on the grounds of the NLBK fire brigade school in Loy in Rastede: NLBK employees flooded a dry dyke there, which is available for disaster control exercises. The conclusion from the tests: The technology works - the position of the seepage line can be determined using the tested sensors. The automatic monitoring does not make dyke guards superfluous, but can support emergency services in an emergency, for example to continuously monitor known weak points in a dyke.
Not only the students were satisfied with the results, but also the experts from water management and disaster control: The "Diek un dat" team has been invited to present the system in June at Interschutz 2026 in Hanover, the world's most important trade fair for rescue services, fire and disaster control. "It's quite an achievement to get a monitoring system like this up and running within a year," emphasises Andreas Winter.
The team was able to build on the work of previous project groups: In previous years, Computing Science students had developed various measuring stations to monitor environmental data - such as particulate matter pollution in the air or water levels in rivers. As a result, they already knew how best to transmit and process the collected data or which microprocessors are suitable for processing it. The online platform "Guerilla Sensing", which was set up in earlier project groups, also provided a system that could be used to access and visualise the data.
"One of the core ideas behind "Guerilla Sensing" is to develop measuring stations that cost little and can be assembled by yourself with a little technical knowledge," explains Paul Tepe. The components of the dyke monitoring system, for example, are available for significantly less than 1,000 euros in total. In principle, such stations could be used in citizen science projects, for example in measurement campaigns organised by interested citizens. Imke Korte would be involved immediately: "If there were a flood, I would be keen to deploy our boxes and support the emergency teams," she says.
For her and Paul Tepe, the project was one of the highlights of their studies - with lots of new tasks away from the usual seminars, written exams and tests. "We really achieved something," says Tepe. For supervisor Andreas Winter, the added value lies in the diverse experiences that the students gain during the one-year project phase. "The project group conveys many aspects of later professional practice that are difficult to teach explicitly," he says. He repeatedly observes that after a few months, the students "make the project their own" and throw themselves into the various tasks, which range from soldering circuit boards to planning sessions with external project partners.
The topic of the next project group is again to do with water: In the approaching summer semester, Theels and Winter's students will be tasked with monitoring the water quality in the Moorhaus polder, for example. As guerrilla sensing is aimed at expandable environmental sensor technology, further experiments will focus on the digital detection of invasive species such as the Asian hornet.
