Forge4Flame: An intuitive dashboard for designing GPU agent-based models to simulate infectious disease spread

Agent-based models are computational models that simulate the dynamic interactions, behaviours, and communication protocols among agents in a shared environment. The use of such models in the field of epidemiology has attracted much attention, allowing the evaluation of the effectiveness of possible interventions and vaccination strategies. However, setting up these environments typically requires a manual and technical process that can be both time-consuming and complex. To address this challenge, we introduce Forge4Flame , a novel and user-friendly dashboard that simplifies the definition of agent-based models for FLAME GPU 2. Our goal is to make this modelling framework more accessible to a broader audience of researchers and public health professionals. Specifically, the tool streamlines model design, execution, and analysis by automatically generating the required FLAME GPU 2 code and incorporating valuable visualisation and post-processing features. Moreover, the integration of two different levels of population model was explored, allowing a detailed analysis of disease dynamics. This shows the tool’s potential to enhance both the accessibility and scalability of agent-based models through Docker and Slurm for efficient distributed computing on high-performance computing systems. Finally, the effectiveness of this tool is demonstrated through a case study that investigates the COVID-19 emergency in a generic Italian middle school. • F4F simplifies Agents Based Models design, easing disease spread modelling. • Intuitive dashboard with advanced visualisation and post-analysis. • Integration of compartmental models and Agents Based Models for what-if analysis. • Fast and reproducible simulation via Docker, Slurm and High Performance Computing.