ADAPTI-M
Next Generation Modeling: Adaptive System for Supporting Public Health Decision-Making in Respiratory Infections Pandemics
Project content and objectives
The German Epidemic Microsimulation System (GEMS), which was developed in the OptimAgent project of the first MONID funding phase, is an agent-based model for simulating respiratory infections. It maps the German population georeferenced and enables the evaluation of non-pharmaceutical interventions (NPI). In the ADAPTI-M project, GEMS is being further developed technically and organizationally: It integrates droplet and aerosol transmission, complex mobility, adherence behavior and new virus variants. Calibration is automated using reinforcement learning and surrogate models. Generative AI supports the development and optimization of NPI strategies based on multiple criteria such as case numbers, school closures or mental stress. A decision framework developed together with decision-makers ensures practical application. The approach is evaluated through stress tests and use cases.
Project structure
SP0 – Management
SP1 – GEMS extension and dynamization
SP2 – Parameterization and calibration
SP3 – Intelligent NPI strategy development and optimization
SP4 – Decision framework
SP5 – Stress tests, evaluation, use cases for demonstration purposes
Project managers and partners
Friedrich Schiller University Jena – Manja Marz
University of Leipzig – Markus Scholz
Münster University- Bernd Hellingrath & Johannes Ponge
Robert Koch Institute – Frank Sandmann
University of Rostock – Adelinde Uhrmacher
University of Trier – Jan Pablo Burgard
University of Lübeck – André Calero Valdez & Alexander Kuhlmann
Martin Luther University Halle-Wittenberg – Rafael Mikolajczyk, Matthias Müller-Hannemann
Associated: Tyll Krüger(Wroclaw University of Science and Technology), Mirjam Kretzschmar (University Medical Center Utrecht), Uwe Siebert & Beate Jahn (UMIT Tirol)