GMapping and Cartographer with YOLOv8: ROS 2 Simulation Code and SLAM Evaluation Artifacts

Abstract

This research software archive provides ROS 2 Jazzy simulation code, configuration files, occupancy-grid maps, TUM trajectory data, and EVO evaluation artifacts for four experimental configurations: GMapping GMapping with YOLOv8n Cartographer Cartographer with YOLOv8n The project uses TurtleBot3, Gazebo Sim, and ROS 2 to investigate SLAM mapping workflows with optional deep-learning-based object detection. Local runtime tests demonstrated successful occupancy-grid map generation for both baseline SLAM algorithms and concurrent operation with YOLOv8 image processing. These tests do not establish that object detections directly modify or improve the generated maps. The archive includes documentation of the software environment, ROS 2 launch files, experimental outputs, evaluation provenance, and known reproduction limitations. Two EVO evaluation archives for GMapping with YOLOv8 were recomputed from retained TUM trajectories and are explicitly identified. Pretrained YOLO weights, raw ROS bag recordings, certain external simulation assets, and complete upstream software repositories are not included. This archive is a research software release candidate. Independent clean-room reproduction and some third-party licensing checks remain outstanding. Users should consult the included README and release documentation before using or redistributing its contents. Licensing: Original software code authored by Eslam Ibrahem Hamed is licensed under the MIT License. Original experimental occupancy-grid maps, TUM trajectories, and EVO evaluation artifacts identified in the archive's data-license inventory are licensed under Creative Commons Attribution 4.0 International (CC BY 4.0). Third-party components retain their respective licenses and copyright notices.