[ros-users] [Discourse.ros.org] [ROS-Industrial] Announcement: ros_control RobotStatus publisher

[gavanderhoorn via Discourse.ros.org]

(as this is rather ROS-I specific I opted to post this here instead of over in the #ros-projects category)

Various robot drivers in ROS-Industrial periodically publish [RobotStatus](http://docs.ros.org/latest/api/industrial_msgs/html/msg/RobotStatus.html) messages that encode controller/robot state (ie: e-stop active, is the robot in motion, etc). For drivers that build on-top of `ros_control`, publishing these messages is typically done by the `hardware_interface` part of the stack, but that always seemed like a work-around to me.

As I needed it for a `ros_control`-based driver recently, I decided to write a `ros_control` robot status controller that can be configured similar to the `joint_state_controller` and takes on the responsibility of publishing `RobotStatus` messages, so `hardware_interface` implementations don't need to do that any more. This separates responsibilities and reduces complexity of `hardware_interface`s.

The controller may be found here: [gavanderhoorn/industrial_robot_status_controller](https://github.com/gavanderhoorn/industrial_robot_status_controller) and the readme explains how to integrate it into a `hardware_interface` for a specific robot and how to add it to a configuration. I'm hosting it on my personal account for now, but that may change in the future.

Note: in case you're wondering: the `industrial_robot_status_controller` does not actually *control* anything. The name follows the (unfortunate) convention in `ros_control` to call everything that derives from `controller_interface::Controller` a *(XYZ)Controller* (such as the [joint_state_controller](http://wiki.ros.org/joint_state_controller) which publishes `JointState` messages).





---
[Visit Topic](https://discourse.ros.org/t/announcement-ros-control-robotstatus-publisher/8170/1) or reply to this email to respond.