ros2_controllers
ros2_controllers copied to clipboard
Generic robotic controllers to accompany ros2_control
Analysis: ======= Working on ros2_controller_demo I happened to have systematically the following exception: `[ros2_control_node-1] ros2_control_node: /opt/ros/foxy/include/rcppmath/clamp.hpp:41: constexpr const T& rcppmath::clamp(const T&, const T&, const T&) [with T = double]: Assertion...
`experimental/optional` is not well supported on OSX and thus the complete package is disabled on OSX. Ticket for tracking a workaround solution to enable the compilation on OSX. relates to...
There is a quite a bunch of same code and logic needed in those controllers. We could add an abstraction layer to it. Here is a short proposal on this:...
As @jordan-palacios wrote: current implementation of JTC is not real-time safe. The memory operations, i.e. vector manipulations and resizing has to be reconsidered. See below his comment: I know this...
#140 implements a generic JointTrajectoryController supporting joint, velocity, and acceleration interfaces and their combinations. For the brevity of PR, pure `velocity` or `effort` controllers are not implemented. Their commands are...
Check the following comment by @matthew-reynolds: I think this needs further work, probably in a follow-up PR. In both the old and new code, I worry that we might end...
A long-standing shortcoming in the realtime_tools::RealtimeBuffer is an inability to provide lock-free writes from the RT context. (See https://github.com/ros-controls/realtime_tools/issues/14). The `joint_trajectory_controller` in particular is currently using `writeFromNonRT()` within the RT...
I just got reported that the `joint_state_controller` has `resorted` joint_names compared to the `ros2_control`-tag defined in the URDF file. They could be confusing and we should "debug" this if possible....
Hello everyone, congratulations on the amazing work with ROS1 and ROS2 control. We are migrating our robot to ROS2 . For a first test, I am trying to use diff_drive_controller...