Web-based SUMO city simulation
A browser SUMO simulation of a city I built with SmartTransportation Lab, used mainly to validate a graph neural network for route-aware ETA against live trips and logged prediction error.
This is a web-based SUMO simulation of a city, hosted as the live demo at SmartTransportation Lab. You pick start and destination on the road network, watch vehicles move through the scene, and the tester logs each trip against a model’s ETA so error can be inspected while the simulation is running.
I used it mainly to validate a GNN for route-aware traffic graphs: the same microsimulation that generates labeled dynamic graphs also drives interactive trips, inference, and MAE / RMSE / MAPE on completed journeys. That evaluation loop is the web testing environment described in the NiDS 2026 paper Route-Aware Dynamic Traffic Graphs for ETA: Dataset Generation and Interactive Model Testing.
Related publication: /publications/nids-2026-route-aware-traffic-graphs
Live demo: https://smarttransportationlab.com/sim-demo