ros2 autonomous robots in production
As developers and engineers we learn the most by reading high quality source code. In the last three years I have been immersed in self-learning of autonomous ground vehicles, and can confidently say that I am familiar with the ROS2 ecosystem.
That includes TF2 and its benefits, the localization package and its use with the Nav2 package, ros2 controls packages, gazebo, and of course DDS and the node lifecycle. I have also studied SLAM (both traditional SLAM and deep learning augmented), as well as deep learning for object detection, semantic segmentation, lane following, and freespace (drive-able) detection.
Is anyone aware of a good open source project that is based on ros2 and is deployed in a production environment? The target system would be for a unmanned ground vehicles, however what is more important is the code quality and system design.
One example is the robot bring-up. I am looking for high quality code that demonstrates system checks upon startup, sensor quality checks and redundancies, etc.
Over the years I have implemented combinations of ros2 packages and deployed them on Gazebo and some basic robot kits from Amazon, with RaberryPi's and Nvidia Jetsons.
I have also learned from many good educational and research platforms, however I am looking for an open source ros2 project, that is actually deployed in production.
Btw: I also spent time on the AutowareAuto project, however there are many inconsistencies in that project (which is a topic for another post)
Edit:
I think this clearpath husky code maybe close to what I was looking for; albeit not ros2 based
Hello i saw that your want some project which uses ros2 in production level. For this you can use baxtor ros packages. One of the repo is here: https://github.com/CentraleNantesRobo...