A cloud-based demonstrator was developed for the evaluation of non-target analysis results, which has been so far labour intensive. Specialized public water supply laboratories can use it to detect known and even unknown trace substances in water with high sensitivity and to automatically evaluate the analysis data within and across laboratories.
Surface waters such as rivers, lakes or dams are an important resource for public drinking water supply. Currently, around a third of drinking water in Germany is produced from surface water, bank filtrate or surface-influenced groundwater. At the same time, surface waters are constantly exposed to various contaminations by micropollutants. Typical sources include industrial production, municipal sewage treatment plants as well as diffuse substance inputs from runoff, erosion or agriculturally used areas. There are also sources from known as well as unknown damages such as accidents or leaks. Of the total trace substances released into the water, only a fraction is currently noted, recognized and identified. Therefore, trace substances often are unnoticed and cannot be assigned to a source or only with considerable instrumental and time-consuming effort. More comprehensive knowledge about the deposition and distribution of trace substances in surface waters supports focused protection measures for drinking water resources and the aquatic environment.
In this project, a demonstrator for a cloud-based software was developed to evaluate results from non-target analysis of water samples efficiently. This sensitive analytical method is based on high-resolution mass spectrometry and allows to detect known and even unknown trace substances in water. In the cloud, analytical results of each laboratory can be processed largely automatically using suitable algorithms. Laboratories of the public drinking water supply from different regions also have the option of sharing and merging their analytical results. The collective human intelligence from the water supply and the artificial intelligence during data processing in the cloud provides additional value in the identification of trace substances and their sources of entry.