Technical University of Denmark
5 files

DTU Roadrunners Autonomous Ecocar Race Track Dataset

posted on 2021-03-18, 15:51 authored by Henning Si Høj, Mads Dahl Larsen, Tobias Stougaard Grønne, Michael Teglgaard Christensen, Liguang He, Gustav Ohlendorff Brønd, Alexander Mizrahi-Werner, Lars David Aktor, Jesper Martin Christensen, Oliver Lynggaard Topp, Thomas Passer Jensen
DTU Roadrunners Dataset
This dataset contains data from a Velodyne VLP-16 Puck lidar, a ZED 2 stereo camera, a Swift Navigation Piksi Multi RTK GPS, and odometry, captured on the DTU Roadrunners Dynamo Ecocar.
The data was captured at the race track at Roskilde Racing Center (RRC) in March - July 2020 and at the DTU Autonomous Vehicle Test Track in May 2020.
For more information about the DTU Roadrunners project visit

Rosbag Content

Raw LIDAR point cloud from Velodyne VLP-16 Puck LIDAR

Raw angle from gyroscope used for odometry

Odometry with fields
msg_data->data[0], // x
msg_data->data[1], // y
msg_data->data[2], // z
msg_data->data[3], // orientation
msg_data->data[4], // speed
msg_data->data[5] // acceleration
msg_data->data[6] // driven_distance

SVO File
Stereolabs ZED2 Stereo camera data with IMU.
Images can be viewed using ZED Explorer, extracted with SVO Export, or published to ROS using the ROS wrapper.
See for more information.

Baseline, Position, and Velocity.
"Flags" column has following definition:
0 Invalid
1 Single Point Position (SPP)
2 Differential GNSS (DGNSS)
3 Float RTK
4 Fixed RTK
5 Undefined
6 SBAS Position

Conversion of /velodyne_packets into sensor_msgs/PointCloud2
Install velodyne drivers
$ sudo apt-get install ros-melodic-velodyne
Launch cloud nodelet
$ roslaunch cloud_nodelet_conversion.launch
Play bag file
$ rosbag play X.bag


DTU Blue Dot

Sponsors of DTU Roadrunners