As 70 percent of drinking water is obtained from groundwater, protection against excessive nitrate levels is of great importance. AI-supported systems can better predict the distribution of nitrate in groundwater and therefore offer intelligent decision-making support for groundwater protection measures.

Use of machine learning algorithms

In the project, the use cases regionalization, monitoring of measured values and monitoring network optimization were considered. Regionalization is used to spatially predict the distribution of nitrate in groundwater. In intelligent measurement monitoring, large quantities of measurement data are examined for deviations using regression and classification algorithms (anomaly detection). Measuring network optimization aims to reduce redundant measuring points and identify missing locations. This can save costs for technology and personnel, but also identify locations for missing measuring points. Various machine learning algorithms were used for these applications, such as random forest methods, deep learning with LSTM networks and methods of geostatistics and mathematical optimization.

Integration of AI processes in data analyses

In addition to the specific use cases, a methodological and technical framework was created to combine AI processes with environmental informatics technologies in the field of water. The AI lighthouse project has made various contributions to the design, development and initial testing of this framework. In particular, it also laid the technological foundation for the analysis extension of the disy Cadenza data analysis platform. This extends the analysis options by supporting the embedding of scientific diagrams and results of AI-based algorithms, such as classification, prediction or clustering in dashboards, among other things.

Presentation of the project results at the networking meeting

The BMUV is inviting all project implementers of the AI lighthouses to the fourth networking meeting in Berlin on June 11, 2024. Disy will represent the research consortium of the AI Lighthouse NiMo and present the project with its goals, approaches and results. Due to the continued high demand for research into AI applications for natural climate protection, the BMUV has stepped up and launched a third call for funding in the spring.

The AI lighthouses of the BMUV funding initiative

The "AI Lighthouses for the Environment, Climate, Nature and Resources" funding initiative from the Federal Ministry for the Environment, Nature Conservation, Nuclear Safety and Consumer Protection (BMUV) was launched in 2019 to support projects that use artificial intelligence to tackle environmental challenges. In two funding rounds, 53 AI lighthouses have been supported so far with a total volume of around 70 million euros. The NiMo project, coordinated by Disy Informationssysteme GmbH, was funded with around 2.5 million euros and successfully implemented in collaboration with the Karlsruhe Institute of Technology (KIT), the Fraunhofer Institute of Optronics, System Technologies and Image Exploitation IOSB and the DVGW Water Technology Center in the period from September 2020 to December 2023.