sumo_bridge

ROS 2 package that bridges SUMO (v1.22.0+) traffic simulation to the ADORe stack via libsumo. SUMO vehicles are published as TrafficParticipantSet messages and the ego vehicle state is injected back into SUMO each simulation step.

Requirements

• SUMO 1.22.0 or later
• ROS 2 (Jazzy or later)
• ADORe ROS 2 workspace built and sourced

Getting Started

Run the demo scenario using the provided script, which starts the node with the bundled demo_sumo_bridge.sumocfg:

bash sumo_configs/start_sumo_bridge.sh

Or run the node directly with a custom config:

ros2 run sumo_bridge sumo_bridge --ros-args \
    --param "sumo_config_file:=<absolute_path_to_config>.sumocfg" \
    --param "ego_start_position:=<lat>,<lon>,<psi>"

ROS Parameters

Parameter	Type	Default	Description
sumo_config_file	string	""	Absolute path to the .sumocfg file. Required.
sumo_step_length	double	0.05	Simulation step length in seconds
use_gui	bool	false	Launch sumo-gui instead of headless sumo. Simulation starts immediately without requiring the play button.
gui_settings_file	string	""	Path to a SUMO GUI settings XML file
gui_zoom	double	5000.0	Initial zoom level for sumo-gui (percentage; 5000 ≈ 10 m street level)
gui_follow_ego	bool	true	Whether sumo-gui should track and follow the ego vehicle
ego_start_position	string	""	Ego start position as "lat,lon,psi" (psi in radians CCW from east). Derives utm_zone, utm_letter, and SUMO coordinates automatically. Required for georeferenced scenarios; omit for local coordinate scenarios.
ego_tracking_id	int	0	Tracking ID of the ego vehicle used for GUI follow and color assignment
ego_vehicle_color	string	"255,255,0"	Ego vehicle color as "R,G,B" or "R,G,B,A" with values 0–255
utm_zone	int	32	UTM zone number. Used only when ego_start_position is not set (local coordinate scenarios).
utm_letter	string	"U"	UTM zone letter. Used only when ego_start_position is not set.
use_geo_conversion	bool	true	When false, bypasses UTM/geo conversion and treats all coordinates as SUMO-local. Set to false for synthetic scenarios without real-world georeferencing.
initial_traffic_count	int	0	Number of SUMO traffic vehicles to spawn ahead of the ego at startup
initial_traffic_spacing	double	20.0	Distance in metres between each spawned traffic vehicle

Topics

Topic	Type	Direction	Description
traffic_participants	TrafficParticipantSet	publish	SUMO vehicles converted to ADORe traffic participants
simulated_traffic_participant	TrafficParticipant	subscribe	Ego vehicle state injected into SUMO each step

Launch

Two example scenarios are provided.

OSM scenario (georeferenced)

Uses a real OpenStreetMap network. The ego start and goal positions are specified as lat/lon and all coordinate conversion is handled automatically:

EGO_START      = Position(lat_long=(52.314331, 10.53793), psi=3.14)
EGO_GOAL       = Position(lat_long=(52.31463, 10.55909), psi=0.0)
EGO_VEHICLE_ID = 111

ego_lat, ego_lon, ego_psi = EGO_START.get_lat_long_coordinates()

Node(
    package='sumo_bridge',
    executable='sumo_bridge',
    parameters=[
        {"sumo_config_file":      SUMO_CONFIG_PATH},
        {"use_gui":               True},
        {"gui_settings_file":     GUI_SETTINGS_PATH},
        {"gui_zoom":              5000.0},
        {"gui_follow_ego":        True},
        {"ego_tracking_id":       EGO_VEHICLE_ID},
        {"ego_vehicle_color":     "255,255,0"},
        {"ego_start_position":    f"{ego_lat},{ego_lon},{ego_psi}"},
        {"initial_traffic_count": 3},
        {"initial_traffic_spacing": 20.0},
    ],
)

Synthetic scenario (local coordinates)

Uses a local coordinate network such as circle50m. Set use_geo_conversion to false and supply utm_zone/utm_letter directly:

EGO_START      = Position(xy=(50.0, 0.0), psi=3.14/2)
EGO_VEHICLE_ID = 111
ego_utm        = EGO_START.get_utm_coordinates()

Node(
    package='sumo_bridge',
    executable='sumo_bridge',
    parameters=[
        {"sumo_config_file":   SUMO_CONFIG_PATH},
        {"use_gui":            True},
        {"gui_settings_file":  GUI_SETTINGS_PATH},
        {"gui_zoom":           100.0},
        {"gui_follow_ego":     False},
        {"ego_tracking_id":    EGO_VEHICLE_ID},
        {"ego_vehicle_color":  "255,0,0"},
        {"use_geo_conversion": False},
        {"utm_zone":           ego_utm[2]},
        {"utm_letter":         ego_utm[3]},
    ],
)

Clock Synchronisation

The bridge synchronises SUMO simulation time to the ROS wall clock. Reference times (tSUMO0, tROS0) are captured after the first simulation step completes, ensuring both clocks are anchored at the same point. When use_gui is true, the --start flag is passed to sumo-gui so the simulation begins immediately without requiring the play button.

Coordinate Conventions

Convention	Details
ROS yaw (yaw_angle)	Radians, counter-clockwise from east
SUMO heading	Degrees, clockwise from north
SUMO getPosition	Returns front bumper position; the bridge offsets by half vehicle length to publish/consume centre positions
Geo conversion	convertGeo(lon, lat, toGeo=true/false) — note argument order is longitude first

Initial Traffic Spawning

When initial_traffic_count > 0 the bridge spawns SUMO-controlled vehicles ahead of the ego start position at startup. Spawn positions are walked along the actual lane geometry using the lane shape polyline, so vehicles land on the road regardless of curvature. Each vehicle is given a unique randomised route built by walking the routing graph forward from its spawn edge, weighted toward higher speed-limit roads. Routes are deterministic per vehicle ID across runs.