Changelog

5.0.0: Refactor MoCap pipeline and sensors (Dec 2025) - Team/Tech Lead: Hamid Ebadi

  • Working demo video scenario: Hamid Ebadi, Pär Aronsson
  • Refactor and improve prediction pipeline: Pär Aronsson
  • Review and update the mocap architecture: Pär Aronsson, Hamid Ebadi
  • Pipeline for Prediction on real data, images, and videos: Pär Aronsson
  • Improve motion player: Pär Aronsson
  • New demo scenario: Pär Aronsson
  • Unify "single" and "multi" pipeline: Pär Aronsson, Hamid Ebadi
  • Documentation and presentation (mkdocs, pandas): Hamid Ebadi

4.0.0: Improvements (Nov 2025) - Team/Tech Lead: Hamid Ebadi

  • System integration tests: Pär Aronsson
  • Overall system diagrams (launch files and bringups): Anton Stigemyr Hill
  • PyTorch model for humanoid MoCap: Pär Aronsson
  • Improved documentation and AutoGluon configuration: Marwa Naili
  • Demo scenario: Anton Stigemyr Hill
  • Semantic segmentation, depth, color camera sensors: Anton Stigemyr Hill, Pär Aronsson
  • Humanoid navigation: Anton Stigemyr Hill
  • Rework GPSS/ArUco navigation: Anton Stigemyr Hill

3.0.0: Jazzy, humanoid (Oct 2025) - Team/Tech Lead: Hamid Ebadi

  • Collision sensor: Anton Stigemyr Hill
  • Automatically generated parameter-files for robot_agents and world fidelity: Anton Stigemyr Hill
  • Shared map and obstacle generation for robot_agents: Anton Stigemyr Hill
  • Geofencing and safety stop: David Espedalen
  • Humanoid motion capture: Casparsson and Siyu Yi, Hamid Ebadi
  • Humanoid integration (pymoveit2 to moveit_py): Siyu Yi, Pär Aronsson
  • Mocap (humanoid) with multi-camera training pipeline: Siyu Yi, Pär Aronsson
  • Panda arm integration: Pär Aronsson
  • Dyno-robotics trajectory replay integration, scenario: Sebastian Olsson

2.1.0: Namespace cleanup and multi-agent navigation (Aug 2025) - Team/Tech Lead: Hamid Ebadi

  • Implemented generic robot-agent control supporting multiple meshes (e.g., pallet truck, forklift) and distinct ArUco IDs: Pär Aronsson
  • Robot-agent and navigation2 (nav2) namespace: Pär Aronsson
  • Multi-agent navigation capabilities using GPSS: Pär Aronsson

2.0.0: GPSS (May 2025) - Team/Tech Lead: Hamid Ebadi

  • Upgrade Gazebo from classic to ignition, adapting new sensors: Nazeeh Alhosary
  • ArUco_localization, added localization and nav2 to new robot: Pär Aronsson
  • Navigation: Pär Aronsson
  • Bird's Eye View ros package: Converting projection.ipynb to camera_bird_eye_view: Hamid Ebadi, Pär Aronsson
  • Deprecated: InfoBot, replaced with Pallet_truck (based on Jackal by Clearpath Robotics)

1.0.6: Collision (Feb 2025)

  • Add scenario execution library for ros2
  • Add action servers for set_speed, teleport_robot and collision
  • Add Node for TTC calculation
  • Add package for custom messages

1.0.5: Delivery 4 (Dec 2024)

  • D3.5 Disentanglement: One-parameter Object Movements.
  • Trajectory visualization: Hamid Ebadi
  • feat: added DynoWaitFor to make sure Jackal is launched last (synchronisation issue)
  • simlan_bringup pkg created: Christoffer Johannesson

1.0.4: Delivery 3 (Oct 2024)

  • SMILE-IV and ArtWork projects contributions
  • Updated camera extrinsic, fixing OpenCV camera calibration to Gazebo simulator conversion: Erik Brorsson, Volvo
  • D3.4 Out distribution data collection: Hamid Ebadi
  • D3.3 In distribution data collection: Hamid Ebadi
  • D3.2 Images for stitching: Hamid Ebadi
  • D3.1 Dataset draft for HH: Hamid Ebadi
  • CI/CD and Kubernetes integration by Filip Melberg, Vasiliki Kostara
  • Jackal integration: Christoffer Johannesson, Hjalmar Ruscck
  • Docker/vscode
  • Disable GPU support by default
  • POC for camera projection to pixel coordinates (Jupyter notebook): Hamid Ebadi
  • Camera image dump and image stitching: Hamid Ebadi

1.0.3: Jackal Robot (Mar 2024) - Christoffer Johannesson

  • Added dyno fork of jackal repo from Clearpath Robotics.
  • Updated to Humble, added bringup and support for namespacing. Jackal can be spawned in Gazebo and controlled through the keyboard.
  • Added .devcontainer folder with Dockerfile and devcontainer.json to set up project container in VS Code.
  • Added docker-compose to link all needed files and set environment variables.
  • Added .vscode folder with settings and tasks for easy building of the project.
  • Updated README with info on how to use Docker setup in VS Code, and some features to make it easy to share the same setup with others.
  • Features include: python3 dependency install with pip, cloning of other git repositories and how to make changes to those repositories.

1.0.2: Delivery 2 (Feb 2024)

  • Volvo warehouse 0.0.1: Hamid Ebadi
  • Volvo camera calibration in Gazebo 0.0.1: Hamid Ebadi
  • Integrate Infobot_agent 0.0.2: InfoBot differential-drive AMR (Autonomous Mobile Robot) URDF and ROS launcher (GOPAL and forklift): Hamid Ebadi
  • Integrate Infobot_cartographer 2.1.5: cartographer for creating PGM maps
  • Integrate nav2_commander 0.0.2: ROS package to command Infobot where the destination is: Hamid Ebadi
  • Integrate Infobot_navigation2 2.1.5: Standard Nav2 stack launcher: Hamid Ebadi
  • Integrate Infobot_teleop 0.0.2: Teleoperation for InfoBot

1.0.1: Delivery 1 (Dec 2023) - Team/Tech Lead: Hamid Ebadi

  • Basic warehouse model 1.0.0: Anders Bäckelie
  • CAD modelling (eur-pallet, boxes, shelf, support_pole, traffic-cone, steel_drum) 1.0.0: Jacob Rohdin
  • Physics (collision, inertia), visuals and Gazebo compatible mesh creation 1.0.0: Anders Bäckelie
  • Walking actor using scripted trajectories 1.0.0: Anders Bäckelie
  • Infobot_Gazebo_environment 1.0.0: ROS2 launcher to start Gazebo world: Hamid Ebadi
  • static_agent_launcher 1.0.0: Camera and ArUco tags: Hamid Ebadi
  • camera-viewer 1.0.0: Python code to get Gazebo camera feed: Hamid Ebadi