Microscopic traffic simulation plays an increasingly important role in developing and evaluating intelligent traffic-control strategies. Nevertheless, comparing new controllers remains difficult. Studies often use project-specific baseline implementations, different simulation settings, or only a single experimental run, making results difficult to reproduce and compare fairly.
ETH Zurich researchers, as part of their work with FEDORA, introduced sumoITScontrol, an open-source and extensible Python framework containing a curated collection of established traffic controllers implemented for SUMO (1) through the TraCI interface. The framework covers both urban and freeway traffic-management applications.
For freeway control, the collection includes widely used ramp-metering approaches such as ALINEA, PI-ALINEA, METALINE, and a HERO-inspired coordinated controller. For urban networks, it provides different implementations of Max-Pressure signal control, alongside adaptive strategies inspired by the principles of SCOOT and SCATS. The controllers are modular and configurable, allowing researchers to apply consistent baselines across different simulation scenarios.
However, the contribution goes beyond software implementation. The paper also highlights the importance of rigorous experimental design in stochastic microscopic simulations. Variations in departure times, driving behaviour, lane changes, and route choices can noticeably affect performance indicators such as travel time, delay, and throughput. Consequently, a small apparent improvement may reflect random variation rather than a genuine advantage.
The framework therefore promotes repeated simulation experiments, systematic controller calibration, variance-aware reporting, statistical hypothesis testing, and transparent benchmarking procedures.
By combining reusable controller implementations with methodological guidance, sumoITScontrol aims to reduce duplicated development effort and strengthen reproducibility, comparability, and experimental standards in SUMO-based intelligent transportation systems research.
Read a preprint version of their article here and access the source code for sumoITS control here
(1) for more about SUMO and other tools developed within the FEDORA framework, see here
















