How watonomous/wato_asd_training works
Project Type and Purpose
• This is a ROS-based robotics monorepo designed for developing, simulating, and visualizing autonomous robot behaviors.
• It uses Docker to create reproducible environments for:
• Gazebo simulation (src/gazebo)
• ROS2 robot nodes (src/robot: bringup, control, costmap, map_memory, odometry_spoof, planner)
• Sample data pipelines (src/samples in C++ & Python: producers, transformers, aggregators)
• Visualization tools (Foxglove)
File Structure Analysis
• Top-level:
• config: JSON/YAML parameters for services & nodes
• docker/: Dockerfiles for each service category
• modules/: docker-compose YAML grouping containers by function (gazebo, robot, samples, vis_tools)
• src/: ROS2 packages & samples
• watod & watod_scripts/: CLI helpers for env setup
• src sub-directories follow ROS2 package conventions (package.xml, CMakeLists.txt, launch files), indicating a distributed pub/sub architecture.
README Insights
• Emphasizes Docker usage for reproducibility.
Elements
• Clear labels on each container and ROS2 package.
Detailed Capture of Architectural Elements
• Launch files under each container to indicate startup sequence.
• Message definitions (Filtered.msg, Unfiltered.msg, etc.) to show topic schemas.
• Module boundaries: e.g., “modules/docker-compose.robot.yaml” controlling which containers run together.
• Watod CLI as a gateway to environment setup, not part of the runtime graph.