The FCD (floating car data) export contains location and speed along with other information for every vehicle in the network at every time step. The output behaves somewhat like a super-accurate high-frequency GPS device for each vehicle. The outputs can be processed further using the TraceExporter tool for adapting frequency, equipment rates, accuracy and data format.

Instantiating within the Simulation#

The simulation generates this output when setting the option --fcd-output <FILE>. <FILE> is the name of the file the output will be written to. Any other file with this name will be overwritten, the destination folder must exist.

By default, fcd-output is enabled for all vehicles and persons in the simulation with output in each simulation step. Using device assignment options or parameters (i.e. --device.fcd.probability 0.25) the set of vehicles which generate fcd output can be reduced. The output period can be set by using option --device.fcd.period <TIME>. To delay output (i.e. until some warm-up time has passed), the option --device.fcd.begin <TIME> may be used.

Generated Output#

The generated XML file looks like this:


  <timestep time="<TIME_STEP>">
      <vehicle id="<VEHICLE_ID>" x="<VEHICLE_POS_X>" y="<VEHICLE_POS_Y>" angle="<VEHICLE_ANGLE>" type="<VEHICLE_TYPE>"

      ... more vehicles ...


  ... next timestep ...

Name Type Description
timestep (simulation) seconds The time step described by the values within this timestep-element
id id The id of the vehicle
type id The name of the vehicle type
speed m/s The speed of the vehicle
angle degree The angle of the vehicle in navigational standard (0-360 degrees, going clockwise with 0 at the 12'o clock position)
x m or longitude The absolute X coordinate of the vehicle (center of front bumper). The value depends on the given geographic projection
y m or latitude The absolute Y coordinate of the vehicle (center of front bumper). The value depends on the given geographic projection
z m The z value of the vehicle (center of front bumper).

Note: This value is only present if the network contains elevation data
pos m The running position of the vehicle measured from the start of the current lane.
lane id The id of the current lane.
slope degree The slope of the vehicle in degrees (equals the slope of the road at the current position)
signals bitset The signal state information (blinkers, etc). Only present when option --fcd-output.signals is set.

When the option --fcd-output.geo is set, the written (x,y)-coordinates will be the lon/lat geo-coordinates.


By default fcd-output returns location values in meter with a precision of 1cm. (changeable by setting option --precision) If you set option --fcd-output.geo the values are lon,lat as decimal values with a precision of 6 decimal places (changeable by setting option --precision.geo)

Person and Container Output#

Any persons or container in the simulation will cause output of he following form:


  <timestep time="<TIME_STEP>">
      <vehicle .../>
      <person id="..." x="..." y="..." angle="..." type="..." speed="..." edge="..." slope="..."/>
      <container id="..." x="..." y="..." angle="..." type="..." speed="..." edge="..." slope="..."/>

  ... next timestep ...


If persons or containers are transported within a vehicle, their respective <person> and <container> elements will be written as child elements of that vehicle.

Filtering / Restricting Output#


The generated output files can become quite large. To write gzipped output files, simply name the output file with an .gz extension.

Restricting the set of vehicles that generate output#

Output can be restricted to specific vehicle types or vehicle ids by controlling the set of vehicles that are equipped with the fcd-device. The following example restricts output to a single vehicle called ego:

--device.fcd.explicit ego

The following example restricts fcd-output to a single flow out of the whole simulation

--device.fcd.probability 0 ...
<flow ...>
   <param key="has.fcd.device" value="true"/>

Restricting the locations#

Output can be restricted to a specific set of edges by loading a list of edges from a file with option --fcd-output.filter-edges.input-file <FILE>. The file format for this is the same as the one when saving selections in netedit:


Restricting the locations by shape#

Output can be restricted to vehicles within a specified area by setting a list of <poly> ids with option --fcd-output.filter-shapes. The polygon shapes must have been loaded from an additional file.

Restricting output by sensor range#

When not all vehicles are equipped with an fcd-device, other vehicles and persons in a radius around the equipped vehicles can be included in the output by setting option --device.fcd.radius to the desired range in m.

Further Options#

  • --fcd-output.geo will toggle output coordinates to WGS84 (for geo-referenced networks)
  • --fcd-output.signals will add signal state information to the output
  • --fcd-output.distance will add kilometrage information to the output
  • --fcd-output.acceleration will add acceleration data to the output (also lateral acceleration when using the sublane model
  • --fcd-output.max-leader-distance FLOAT will add attributes leaderGap, leaderSpeed, leaderID whenever a vehicle has a leader within the given distance. Otherwise, leaderID will be "" and leaderGap, leaderSpeed will be -1.
  • --fcd-output.params KEY1,KEY2,... adds generic parameters to the output (supports device and carfollowmodel parameters as well as arbitrary user-define values)
  • --fcd-output.attributes ATTR1,ATTR2,... restricts written attributes to the given list (to reduce output). The following attributes are special:
    • all: enables all attributes
    • vehicle: add a vehicle attribute to each person (and thereby distinguish riding from walking persons).
    • odometer: write odometer value for each vehicle (distance driven since departure)
    • posLat: write lateral position on lane for each vehicle


In combination with the given geometry of the vehicles (shapes) you can build some nice animations, e.g NASA WorldWind or Google Earth.

Visualization example#

Accelerations versus distances#

Generated with plot_trajectories.py.

All trajectories over time#

Generated with plotXMLAttributes.py.

Generated with plot_trajectories.py.

Selected trajectories over time#

Generated with plotXMLAttributes.py.

Generated with plot_trajectories.py.

FCD based speeds over time#

Generated with plot_trajectories.py.