Development of a state-wide wastewater surveillance in Thuringia using mobility data and artificial intelligence ("AMELAG")

Background and motivation

The detection of hygienically relevant microorganisms and viruses in wastewater, combined with epidemiological methods (known as "wastewater-based epidemiology"), has the potential to establish effective and cost-efficient surveillance procedures in public health protection, as corroborated by recent national and international studies on detecting SARS-CoV-2 in wastewater. Crucial prerequisites for reliable wastewater surveillance in public health protection are (i) cost-effective, sustainable concepts for collecting wastewater samples and (ii) suitable data analysis strategies, and (iii) the consideration of the population mobility. Therefore, AI-driven analyses of mobility and wastewater data may significantly contribute to rendering wastewater surveillance more reliable and wastewater sampling strategies more adaptive to current and future infection events.

Project goals and expected results

The goal of this research project is the development of a centralized AI model, to be trained with data combined from different sources, including mobility data, SARS-CoV-2 data collected from wastewater treatment plants, environmental data, as well as data originating from domain-specific engineering knowledge. The AI model builds upon established AI techniques and architectures, such as the Generative Pre-trained Transformer (GPT), to allow for predictive statements about SARS-CoV-2 in wastewater treatment plants. In a further step, the centralized AI model, in the form of reduced partial models, will be embedded into wireless sensor nodes installed in wastewater treatment plants. The sensor nodes will be equipped with IoT capabilities, enabling seamless communication and data integration from external sources. Finally, in collaboration with the project partners, the concept developed in this project is intended to be adapted to other pathogens, such as influenza. As a result, it is expected that an innovative methodology for wastewater surveillance and AI-based prediction of infectious events can be provided, entailing a reduction of wastewater samples required for data analysis. It is further expected to contribute to early responses to infectious disease outbreaks, strengthening the public health protection.

Project partners

  • Bauhaus University Weimar (Coordinator)
  • University Hospital Jena
  • Hamburg University of Technology
  • Hamm-Lippstadt University of Applied Sciences
  • SMA Development GmbH
  • KOWUG GmbH


Professor Dr. Kay Smarsly
Hamburg University of Technology
Institute of Digital and Autonomous Construction
Blohmstraße 15
21079 Hamburg
Email: kay.smarsly(at)tuhh(dot)de