MRCD: Mobile Robot Campus Dataset

Welcome to the MRCD dataset for outdoor mobile wheeled robotics! This ROS2 Humble dataset is designed to support algorithm development and benchmarking for localization, navigation and perception in real-world outdoor environments. The dataset features a variety of challenging outdoor sequences for outdoor mobile robots that have been recorded at the campus of Hamburg University of Technology. The data was gathered with our self-built robot Laura.

Highlights

  • Our dataset features high-quality sensor data—including:
  • 📷 HD720 30 FPS stereo frontal camera streams
  • ☁️ High resolution visual frontal pointcloud
  • 📷 HD720 30 FPS mono depth + colored ground-facing camera streams
  • 📏 3D spinning LiDAR
  • ⬆️ High quality IMUs (raw accelerometer, gyroscope and magnetometer data)
  • 🛞 Odometry of our wheeled robot
  • 📍 Highly accurate GPS-measurements
  • 👣 Embedded discrete and continuous external ground truth
  • Comprehensive bags, including all sensor modalities and recordings
  • Additional lightweight bags, excluding visual data
  • High resolution large scale survey grade prior map of the campus environment
  • Docker Images for Humble SOTA SLAM algorithms.

Notes:

We are open to contributions to our dataset. Please feel free to raise an issue or open a discussion on our Github 🙃

Publication:

For more information, please find our publication below. If you use MRCD or reference our work, we kindly ask that you cite it as follows:

@article{mrcd2025,
  title={MRCD: Mobile Robot Campus Dataset for Evaluating SLAM Algorithms on Wheeled Robots},
  author={Doe, John},
  journal={arXiv},
  year={2025}
}

Arxiv Supplementary Material

This work was funded by the Federal Ministry for Digital and Transport Affairs