EPISERVE

Epidemic Simulation & Data Service Platform (Platform for epidemic simulation and data services: The Architecture of Preparedness – Integration of Mobility, Social Dynamics and Infection Data)

Project content and objectives

Objective: To design, validate and implement a sustainable real-time epidemic simulation platform that integrates heterogeneous data streams and multiscale modeling to enable timely, evidence-based public health decision-making.

Building on infrastructure technologies developed in previous large-scale projects, EPISERVE integrates traditional biological monitoring data with novel social and behavioral data streams. This enables the models to depict not only the dynamics of virus transmission, but also the behavioral reactions of the population.

Project structure

AP1: Multi-virus adaptation. Modular, pathogen-specific components customize transmission characteristics, incubation times, immune protection and severity profiles, enabling comparative scenario analyses across influenza, RSV and emerging respiratory threats. Pandemics are characterized both by biological transmission dynamics and by social interaction processes. The new models integrate sentiment analysis of social media data (e.g. X and Telegram) with survey-based indicators to quantify anxiety, fatigue and polarization and map their impact on compliance and behavioral adaptation within the epidemic simulation.

AP2: The MATSim-EpiSim core. From “crisis code” to sustainable software engineering. The core of the MATSim-EpiSim framework is being professionalized. This includes a comprehensive code audit and the implementation of AI-driven pipelines that automatically calibrate the model based on incoming RKI data, thus eliminating manual work steps.

WP3: Hybrids & AI models. On the way to near real-time scenario testing. The MATSim-EpiSim framework is being extended, and hybrid and AI-based surrogate models are being developed that approximate high-precision simulations with a fraction of the computational effort. This enables rapid evaluation of intervention strategies, severity predictions and capacity planning, while maintaining mechanistic interpretability and epidemiologic consistency.

WP4: Data, model & simulation platform. The platform automates the process from data collection to modeling to decision making, minimizing manual bottlenecks between data collection and decision support in public health.

WP5: Turning data into action. From calculation to decision support and dissemination. Interactive dashboards: Web-based dashboards allow public health stakeholders to explore epidemiologic indicators, scenario simulations and severity predictions in near real-time. Building on infrastructure technologies developed in previous large-scale projects, EPISERVE integrates traditional biological monitoring data with novel social and behavioral data streams. This enables the models to depict not only the viral transmission dynamics, but also the behavioral reactions of the population.

Project managers and partners

Institution	Partners involved
Zuse Institute Berlin (ZIB) Berlin, Germany	PD Dr. Tim Conrad, Natasa Djurdjevac-Conrad, Christof Schütte
Technical University of Berlin (TUBI) Berlin, Germany	Prof. Dr. Kai Nagel
Freie Universität Berlin (FUB) Berlin, Germany	Prof. Dr. Max v. Kleist
Robert Koch Institute (RKI) Berlin, Germany	PD Dr. Thorsten Wolff, Djin-Ye Oh
German Aerospace Center (DLR) Cologne, Germany	Dr. Martin Kühn
Associated partner: Wroclaw University (Wroclaw University of Technology, WUST) Wroclaw, Poland	Prof. Dr. Tyll Krüger