Data-Driven Decision-Making: Early Detection of Flood Risks

Artificial intelligence calculates, based on sensor data, whether and when a flood is imminent. The Netilion Flood Monitoring system by Endress+Hauser keeps those responsible consistently informed about local risks.

Floods result from prolonged, widespread rainfall, short-term intense downpours, or snowmelt during the winter and spring seasons. Mountainous regions are particularly at risk, as are landscapes characterized by valleys or gorges. Often, it is the smaller water bodies and streams, which are minimally monitored, that can quickly transform into dangerous torrents during heavy rainfall.

When it rains, some of the water is absorbed by the soil, contributing to the replenishment of groundwater. Another portion of the water is stored in the soil or evaporates, while the remainder flows across the surface into water bodies. The amount of water that runs off depends on soil properties, the degree of surface sealing, and the topography. These factors can lead to rapid surface runoff and, in the worst cases, flooding, requiring swift action.

Andreas Graf is the mayor of the municipality of Lenzkirch, where the Netilion Flood Monitoring by Endress+Hauser is in use.

The early warning system, Netilion Flood Monitoring, contributes to the early detection and mitigation of such events. This system utilizes sensors and artificial intelligence to predict flood hazards and enable timely countermeasures. The system was developed by Endress+Hauser in cooperation with Okeanos Smart Data Solutions, a startup originating from the Ruhr University Bochum.

Precise Measurements with Sensors

The objective of the early warning system is to alert decision-makers such as authorities, local representatives, and emergency services in advance of impending floods. “Our goal with this system is to enable users to accurately assess the flood risk for their areas and assist them in making targeted decisions”, explains Florian Falger, Market Manager at Endress+Hauser. The affected parties—from mayors and firefighters to municipal yard workers and engineering office employees—receive timely information via smartphones or computers about water developments in their areas and any critical values reached.

The basis for these warnings are local measurements collected around streams and their surroundings. Water level gauges, heavy rain sensors, and soil moisture sensors are deployed to build a detailed understanding of the respective area. These data are transmitted to Endress+Hauser's cloud platform, Netilion, where they are analyzed and processed by artificial intelligence. Taking into account additional information such as weather forecasts and terrain conditions, the AI predicts whether and when a flood might occur and identifies potential causes. The results are communicated clearly and distinctly to users, so they do not have to interpret the data themselves.

The sensors from Endress+Hauser are installed within a day. Extensive construction measures and permits are not necessary.

The continuous operation of the system and the ongoing collection of data lead to deeper insights and improved predictions. “The great advantage of artificial intelligence is that it continuously optimizes itself. The algorithm learns over time and thus understands the area more precisely”, emphasizes Benjamin Mewes, founder of Okeanos. “Digitalization not only enables quicker decisions but also long-term improvements in flood protection strategies.” With the accumulated knowledge, particularly vulnerable areas can be more specifically secured.

Quick Installation and Eligible for Funding

Once the locations of the measuring stations are determined, the Netilion Flood Monitoring early warning system can be installed within a day. Most sensors in the system are battery-operated and do not require special infrastructure, such as alterations to the water bodies, creating cross-sections, or erecting gauge houses. Thus, extensive construction measures and the need for permits for installation are avoided. The sensors can be easily mounted on existing structures such as pipelines, bridges, underpasses, walls, or roads. Furthermore, the system is designed to integrate existing measurement stations, like state water level gauges, into its analysis.

From the beginning, immediately after the sensor installation, the flood protection system provides reliable data. Moreover, the system's funding is eligible for subsidies: municipalities can apply for state grants to enhance the effectiveness of their flood protection.

Successful Deployment in the Black Forest

In Lenzkirch, Netilion Flood Monitoring is already in use. The community in the Black Forest was last surprised by a flood in 2018, which caused damage amounting to several hundred thousand euros. The village is located in a basin surrounded by mountains and hills. Additionally, two rivers run through the town. “In the case of heavy rain, the danger is great and the lead time is short”, says Andreas Graf. The mayor of Lenzkirch recalls the 2018 flood: “At that time, the water rose incredibly fast, due to a storm coinciding with snowmelt.”

Innovative early warning systems can help minimize the consequences of floods.

Today, various measuring stations are distributed around Lenzkirch: Six soil moisture sensors are embedded in the ground, revealing whether the soil can still absorb rainwater or is already saturated. At the roof gable of the municipal yard and at another location, a rainfall sensor is installed, measuring the intensity of the rain. And at a total of nine points along the rivers and streams of the area, radar water level gauges are installed, showing the water levels.

Members of the Lenzkirch Volunteer Fire Brigade are convinced of the benefits of digital monitoring options. “When critical flood levels were reached in February 2021, we were able to monitor the water levels from our control center using Flood Monitoring, without having to drive to each location. This not only saves valuable time but also increases safety for the emergency personnel”, says Thomas Raufer, overall fire brigade commander.

Technology alone—and thus early warnings of impending floods—is not enough to prevent disasters. Here, the municipalities are called upon to establish specific crisis concepts with concrete actions. “For us, technological support is very welcome”, says Mayor Andreas Graf. “The system cannot prevent a flood, but it gives us valuable time through early warning. And in a flood, every minute counts.”

A Good Example of the Benefits of IIoT

“Netilion Flood Monitoring is an example of how we at Endress+Hauser create concrete benefits for users through the Internet of Things (IoT)”, says Florian Falger. “We achieve this here by combining measured values, a cloud platform, and an AI”, Falger continues. “All three ingredients of this combination result from our own development and research: We developed the sensors for water levels and soil moisture, as well as the cloud platform Netilion ourselves—and thanks to the scalability of our cloud architecture, we can integrate the AI algorithm developed by Okeanos into our digital ecosystem.”

The research effort for such innovations is considerable. At Endress+Hauser, a medium-sized company that has been developing and producing measurement technologies for industrial process engineering for nearly 70 years, 1,200 of the more than 16,000 employees worldwide are dedicated to research and development. The newly founded Innovation Lab in the Level+Pressure business area serves as a breeding ground for new approaches and business models. To specifically promote innovations, the innovation management has been directly linked to the executive board.