← Home

Uses

Uses

Updated 2025-05-15

Workstation

  • OS: Ubuntu 22.04 LTS — everything in robotics runs on Linux
  • Machine: [TODO: add hardware spec] — running ROS2, Docker, and CUDA workloads
  • Editor: VS Code with C/C++, Python, and ROS extensions

Development

  • Languages: C++17 for inference and robotics; Python for prototyping and data pipelines; TypeScript for tooling
  • Build: CMake + colcon (ROS2 workspaces)
  • Containerisation: Docker — all deployment targets use containers; dev environments use Dev Containers
  • Version control: Git, GitHub

Robotics stack

  • Middleware: ROS2 (Humble / Iron)
  • Inference runtime: ONNX Runtime — runs the same model on x86 dev and ARM robot hardware
  • Computer vision: OpenCV 4.x
  • Point cloud: PCL (Point Cloud Library)
  • Simulation: Gazebo / Ignition

ML / AI

  • Training: PyTorch — export to ONNX for deployment
  • Quantisation: ONNX Runtime quantisation tools (INT8, FP16)
  • Experiment tracking: [TODO: MLflow / W&B — confirm]

Hardware I work with

  • AGILOX AMR hardware (proprietary)
  • LiDAR sensors: [TODO: specific models]
  • RGB-D cameras: [TODO: specific models]
  • Edge compute boards: [TODO: Jetson / other]

Terminal / Shell

  • Shell: zsh + Oh My Zsh
  • Terminal: [TODO: add terminal emulator]
  • Multiplexer: tmux — essential for multi-pane ROS2 development

Updated May 2025.