This readme file was generated on [2025-06-27] by [Sean Maroofi] # GENERAL OVERVIEW Title of Dataset: MRCD - Mobile Robot Campus Dataset for Evaluating SLAM Algorithms on Wheeled Robots ## DESCRIPTION An outdoor mobile robot data set for the development of localization, mapping, and navigation algorithms. It features: - 8 diverse data sequences across the TUHH campus environment - additional continuous ground truth references for evaluation - ready-to-use docker images of 7 sota SLAM algorithms - 1cm resolution pointcloud of the campus ## AUTHOR INFORMATION Name: Justin Ziegenbein ORCID: 0000-0003-0969-5737 Institution: Hamburg University of Technology Institute of Logistics Engineering City, country: Hamburg, Germany Email: justin.ziegenbein@tuhh.de Name: Noel Blunder ORCID: 0000-0001-6979-3991 Institution: Hamburg University of Technology Institute of Logistics Engineering City, country: Hamburg, Germany Email: noel.blunder@tuhh.de Name: Sean Maroofi ORCID: 0009-0008-8504-8404 Institution: Hamburg University of Technology Institute of Logistics Engineering City, country: Hamburg, Germany Email: sean.maroofi@tuhh.de Name: Marko Thiel ORCID: 0000-0002-7249-1203 Institution: Hamburg University of Technology Institute of Logistics Engineering City, country: Hamburg, Germany Email: marko.thiel@tuhh.de Name: Thien-Minh Nguyen ORCID: 0000-0003-1315-0967 Institution: Nanyang Technological University School of Electrical and Electronic Engineering City, country: Singapore, Singapore Email: thienminh.nguyen@ntu.edu.sg Name: Hendrik Wilhelm Rose ORCID: Institution: Hamburg University of Technology Institute of Logistics Engineering City, country: Hamburg, Germany Email: hendrik.wilhelm.rose@tuhh.de Name: Philipp Braun ORCID: 0000-0003-4105-9650 Institution: Hamburg University of Technology Institute of Logistics Engineering City, country: Hamburg, Germany Email: philipp.braun@tuhh.de Name: Carlos Jahn ORCID: 0000-0002-5409-0748 Institution: Hamburg University of Technology Institute of Logistics Engineering City, country: Hamburg, Germany Email: carlos.jahn@tuhh.de ## FUNDING Funding source: German Federal Ministry for Digital and Transport Research project: "Smart Control Center for Automated Transport Robots and Buses in the City of Lauenburg/Elbe - TaBuLa-LOGplus" ## SHARING INFORMATION DOI: https://doi.org/10.15480/882.15125 Handle: https://hdl.handle.net/11420/55496 License: Attribution 4.0 International (CC BY 4.0) https://creativecommons.org/licenses/by/4.0/ # DATASET INFORMATION Version: 1.0 Version changes: 2025-06-27 - Initial Upload Location of data collection: Campus of Hamburg University of Technology (Am Schwarzenberg-Campus 1, 21073 Hamburg, Germany) Area captured: 559 m x 247 m Time of data collection: August 2022 - December 2024 ## Content & filesizes: Sequences: alley_fast - alley_fast_compressed.zip - alley_fast_nocamera_compressed.zip alley_loop - alley_loop_compressed.zip - alley_loop_nocamera_compressed.zip grove_clockwise - grove_clockwise_compressed.zip - grove_clockwise_nocamera_compressed.zip grove_counterclockwise - grove_counterclockwise_compressed.zip - grove_counterclockwise_nocamera_compressed.zip town_clockwise - town_clockwise_compressed.zip - town_clockwise_nocamera_compressed.zip town_counterclockwise - town_counterclockwise_compressed.zip - town_counterlockwise_nocamera_compressed.zip town_courtyard - town_courtyard_compressed.zip - town_courtyard_nocamera_compressed.zip town_trees - town_trees_compressed.zip - town_trees_nocamera_compressed.zip Ground Truth: alley_fast - alley_fast_gt_sampled_10Hz.csv - alley_fast_gt_spline_log.csv alley_loop - alley_loop_gt_sampled_10Hz.csv - alley_loop_gt_spline_log.csv grove_clockwise - grove_clockwise_gt_sampled_10Hz.csv - grove_clockwise_gt_spline_log.csv grove_counterclockwise - grove_counterclockwise_gt_sampled_10Hz.csv - grove_counterclockwise_gt_spline_log.csv town_clockwise - town_clockwise_gt_sampled_10Hz.csv - town_clockwise_gt_spline_log.csv town_counterclockwise - town_counterclockwise_gt_sampled_10Hz.csv - town_counterclockwise_gt_spline_log.csv town_courtyard - town_courtyard_gt_sampled_10Hz.csv - town_courtyard_gt_spline_log.csv town_trees - town_trees_gt_sampled_10Hz.csv - town_trees_gt_spline_log.csv Docker Images Google's cartographer - mrcd_cartographer_container.tar SLAM evaluaiton - mrcd_evaluation_container.tar FAST LIO - mrcd_fastlio_container.tar NVIDIA ISSAC ROS SLAM - mrcd_isaacrosvislam_container.tar Open VINS - mrcd_openvins_container.tar ORB3 SLAM - mrcd_orb3slam_container.tar RTAB-MAP - mrcd_rtabmap_container.tar NAV2 SLAM Toolbox - mrcd_slamtoolbox_container.tar TUHH Pointcloud TUHH_campus_subsampled-1cm.e57 (5cm resolution: https://tore.tuhh.de/entities/product/55dc651d-63ab-4d38-8ba8-5375ffe005b9) Description: {SEQUENCENAME}_compressed.zip - bag file includes all sensor data (for VI- and LI-SLAM) {SEQUENCENAME}_nocamera_compressed.zip - bag files have no data from cameras (for LI-SLAM) {SEQUENCENAME}_gt_sampled_10Hz.csv - continuous ground truth sampled at 10Hz {SEQUENCENAME}_gt_spline_log.csv - continous ground truth in spline form TUHH_campus_subsampled_{RESOLUTION}.e57 - Point cloud with intensity and color information ## METHODOLOGICAL INFORMATION Devices used for data acquisition: Transport Robot Laura equipped with Velodyne VLP-16, Stereolabs ZED2, Intel Realsense D435 and EMLID Reach, FARO Focus S 70, FARO Focus Premium 150 Devices used for post processing: Workstation with Intel Core i0 13900K processor, 128 GB RAM and two NVIDIA RTX 4090 GPUs, Ubuntu 22 and ROS 2 Humble Point cloud registration: The registration was performed using the manufacturer's software FARO Scene. Initially, three large clusters were individually registered using artificial sphere targets. The combination of these three clusters to form the final point cloud was done in FARO Scene using cloud-to-cloud registration. Downsampling: Downsampling of the registered point cloud to 5 cm resolution using CloudCompare open source software.