Welcome to the TUD-SUMO Wiki!

This is the documentation for the TUD-SUMO package, a research-oriented wrapper for SUMO[1], developed for the DAIMoND lab at the Technische Universiteit Delft (TUD), the Netherlands.
The main goal of TUD-SUMO is to act as a simplified framework for microscopic traffic simulation that allows researchers and students to focus on the important aspects of their projects; their own work, instead of simulation code. TUD-SUMO provides an easy and standardised way to simulate a wide range of scenarios whilst facilitating complex interactions. Resulting data can then be saved, summarised and visualised with minimal code.
More information on "Simulation of Urban MObility" (SUMO) can be found in the SUMO documentation, here: sumo.dlr.de/docs/
The main features of TUD-SUMO include:
- Automatic and standardised data collection.
- Simplified interface to interact with and control the simulation in complex ways.
- Traffic signal control logic.
- Extendable controllers already implemented (ramp metering, route guidance and variable speed limits).
- A weather & event system with dynamic or scheduled incidents.
- Plotting functions for a wide range of applications.
- Videos for recording the network or specific vehicles during the simulation.
- And more in the future! ...

Links
- Simulation of Urban MObility (SUMO) documentation: sumo.dlr.de/docs
- TUD-SUMO source code (GitHub): DAIMoNDLab/tud-sumo
- TUD-SUMO PyPI distribution: project/tud-sumo
- TUD-SUMO example (GitHub): DAIMoNDLab/tud-sumo-examples
Latest Version
The Latest version of TUD-SUMO is v3.3.2, and was released on 10/03/2026. All previous versions and their change notes can be found on GitHub or PyPI. This documentation was last updated on 10/03/2026.
The change notes for v3.3.1 and v3.3.2 are:
Floating Car Data, Closing Roads & Fixes (v3.3.1)
- Added
Simulation.[open|close]_road()to indefinitely open/close road. - Added
Simulation.save_fc_data()to save all floating car data (vehicle position & speed). - Added ability to smooth data in space-time diagrams using gaussian filtering. Requires tuning of
gf_sigmaparameter. - Changed
individual_vehicle_datato more accuratefc_data(floating car data). - Changed default simulation data output format from '.JSON' to '.pkl'.
- Corrected docstring for
Simulation.cause_incident(). Incident vehicles stop on the following edge, not current one. - Fixed
positionparameter not being used inSimulation.cause_incident(). - Fixed reduced speed on incorrect edge during an incident.
- Added missing grid to VSL plots.
- Added SciPy as a new dependency.
Code Restructuring (v3.3.2)
- Restructured and reordered functions in the main Simulation class for maintainability and readability.
- Moved
print_summary()andprint_sim_data_struct()to new helpers module.
Contact
TUD-SUMO is developed in the DAIMoND lab of TU Delft. For any questions or feedback, please contact Callum Evans at c.evans@tudelft.nl. Bug reports can be created in the GitHub repository: github.com/DAIMoNDLab/tud-sumo.
Acknowledgements
TUD-SUMO is part of the research under the project "AI in Network Management," funded by Rijkswaterstaat, grant agreement nr. 31179439, under the label of ITS Edulab.
Citations
- "Microscopic Traffic Simulation using SUMO"; Pablo Alvarez Lopez, Michael Behrisch, Laura Bieker-Walz, Jakob Erdmann, Yun-Pang Flötteröd, Robert Hilbrich, Leonhard Lücken, Johannes Rummel, Peter Wagner, and Evamarie Wießner. IEEE Intelligent Transportation Systems Conference (ITSC), 2018.