tahamidhossain/2dsumo

2D Dubai SUMO Simulation Dataset

Real traffic simulation of the Al Wasl / Dubai World Trade Centre corridor (approx. 25.202–25.228° N, 55.248–55.282° E).

Files

File	Description	Size
dubai_al_wasl.osm	Source OSM XML exported via Overpass API (65k+ elements)	6.3 MB
dubai_al_wasl.net.xml	SUMO road network (junctions, lanes, signals)	27 KB
polygons.xml	Building + water + commercial footprints	2.6 MB
trips.trips.xml	Randomized traffic demand (600 veh/h, seed=42)	55 KB
routes.rou.xml	Shortest-path routes assigned by duarouter	73 KB
dubai_al_wasl.sumocfg	SUMO configuration (entry point)	765 B

Usage

# Install SUMO (Ubuntu)
sudo add-apt-repository -y ppa:sumo/stable
sudo apt-get update && sudo apt-get install -y sumo sumo-tools sumo-doc

# Headless
sumo -c dubai_al_wasl.sumocfg --no-warnings

# GUI
sumo-gui -c dubai_al_wasl.sumocfg

# Monitor Trackio dashboard
python3 scripts/monitor.py --project 2dsumo --run default
trackio show --project 2dsumo

Build pipeline

OSM (65k+ elements)
  ↓ curl + Overpass API + User-Agent header
  ↓ netconvert --osm.layer-elevation 5 --junctions.join --tls.default-type static
Net (_sumo/dubai_al_wasl.net.xml)
  ↓ polyconvert + osmPolyconvert.typ.xml
Polygons (_polyconvert/polygons.xml)
  ↓ python3 /usr/share/sumo/tools/randomTrips.py --seed 42 --period 6.0
Trips (_sumo/trips.trips.xml)
  ↓ duarouter --routing-algorithm astar --remove-loops
Routes (_sumo/routes.rou.xml)
  ↓
.sumocfg (entry point for sumo-gui)

Build details

• Epoch range: 0 – 1800 s (30 minutes simulated, verified headless)
• Seed: 42
• Inter-departure period: 6.0 s (~600 veh/h)
• Min trip distance: 80 m
• Fringe factor: 5
• Class mix (default): passenger, bus, motorcycle, truck
• TLS strategy: Static, default-type set during --tls.default-type static in netconvert; individual phases inheriting SUMO defaults
• Routing algorithm: A* (--routing-algorithm astar)

Source

Companion code, CI, docs: thtahamid/2dsumo

License

ODbL (from OpenStreetMap).

Generated by ML Intern

This model repository was generated by ML Intern, an agent for machine learning research and development on the Hugging Face Hub.

• Try ML Intern: https://smolagents-ml-intern.hf.space
• Source code: https://github.com/huggingface/ml-intern