Skip to content

Automata Deployment Tools for ROS2. Designed to aid in the development and deployment of neural networks and data collection system utilizing edge devices.

License

Notifications You must be signed in to change notification settings

Rnfudge02/ADTR2

Repository files navigation

Automata Deployment Tools ROS2

Automata Deployment Tools ROS2 is a ros2 package used to develop and deploy edge robotics in a minimal, autonomous framework.

Installation

  1. Install the AARDK prerequisities.
  2. Clone the AARDK.
  3. Clone the project under ./Projects/Deployment.
  4. Run ./CC.sh -b auv-deployment.
  5. Run ./CC.sh -s auv-deployment.
  6. Compile the project with the following command.
/home/auv-deployment/ros_ws/DC.sh

Usage

If you want to visualize data collection and control the device via GUI, install Foxglove Studio using the following command.

sudo snap install foxglove-studio #x86_64 only, arm64 requires install with dpkg

AUVMonitor

Run the following command to start the AUV Monitor, the main entrypoint for the program. Running the DC.sh script mentioned above will launch the monitor

source /opt/ros/${ROS_DISTRO}/setup.bash && source install/setup.bash && \
    ros2 run auv_deployment_tools_ros2 auv-monitor

The AUVMonitor node can be interacted with by sending requests of type std_srvs::Trigger to the following topics.

  1. toggle_daq
  2. toggle_system

After launching, the AUVMonitor will report the status of all modules. The following topics are available at a configurable refresh rate (default = 4Hz).

  1. /auv_monitor/status/daq
  2. /auv_monitor/status/system

Documentation

Documentation for code can be generated using the included Doxyfile with Doxygen. The resultant webpage should contain documentation for all component .hpp and .py files.

To generate documentation for configuration options, the yamldoc pip package should be used

Contributing

This is an open source project maintained by Robert Fudge, 2025 -

To ensure Doxygen output is complete and precise, please follow Doxygen comment styling to ensure generated output contains all documentation

For specific styling preferences, refer to AUVMonitor.hpp for an example

This is currently a closed-source project maintained by Robert Fudge, 2025 -

Frequently Asked Questions

If pytorch encounters any CUDA errors, it could be the container is having trouble accessing the hardware due to some unexpected condition (sleep, etc.), to fix, restart the device.

License

Apache-2.0

About

Automata Deployment Tools for ROS2. Designed to aid in the development and deployment of neural networks and data collection system utilizing edge devices.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published