From MATSim to MultiMATSim: Rethinking Traffic Modeling Using the ‘Unite and Conquer’ Approach

In the ever-evolving landscape of road traffic simulations, agent-based models are experiencing unparalleled ad-vancements. Central to this transformation is the vast amount of data they manipulate, essential for their escalating realism and precision. Yet, this wealth of data poses challenges, especially when depicting the intricate motion of vehicles in sprawling metropolises or vast regions. This paper introduces multiMATSim, our novel application of the ‘Unite and Conquer’ strategy tailored specifically for MATSim, a frontrunner among multi-agent road traffic simulators. We detail the mechanics of this revolutionary algorithm, touch upon its key parameters, and present enlightening experimental results. Our analysis underscores the efficacy of our approach and tantalizingly suggests its potential applicability across a spectrum of other multi-agent simulators.