This is an old revision of the document!
Control System Talking to ROS
In the EEROS library you will find a directory with examples. For this example see RosTest1.cpp.
Open a shell in the build directory of your EEROS library. This examples with ROS will only be present in your build directory if ROS was installed before building EEROS with -DUSE_ROS=TRUE
set, see Preparations and Building. Open a shell and run
$ roscore
Open another shell in the build directory of your EEROS library and run
$ sudo -E ./examples/ros/rosTest1
You have to run a program using real time threads with root privileges. Also make sure to have run the
setup.bash
script of ROS in your shell as described in Preparations and Building.
Publishing
This example runs your EEROS application as a ROS node under the name /eerosNode. It publishes three topics
- /test/vector (of type Matrix<7,1,double>)
- /test/val (of type double)
- /test/safetyLevel (of type unit32_t)
With
$ rosnode list
you will see the new node.
$ rostopic list
will list the two topics which are published be our node, /test/val, /test/vector, and /test/safetyLevel. Open another shell to show the messages sent by EEROS, e.g.
$ rostopic echo /test/vector
With
$ rostopic hz /test/val
you can easily check how often data is sent from out ROS node. Try to plot the data with
$ rqt_plot /test/val /test1/vector/data[0] /test/vector/data[1] /test/safetyLevel
Subscribing
Open another shell in the build directory of your EEROS library and run rosNodeTalker
. This is not a EEROS program but simply starts a ROS node which publishes a couple of test topics. As soon as rosNodeTalker
runs, your example application will receive the two topics
- /rosNodeTalker/vector (of type Matrix<2,1,double>)
- /rosNodeTalker/val (of type double)
and logs its transported message values into the console. Alternatively you could publish a single message from the command line with
$ rostopic pub /rosNodeTalker/val std_msgs/Float64 -- -2.3 $ rostopic pub /rosNodeTalker/vector std_msgs/Float64MultiArray "{layout: {dim: [], data_offset: 0}, data: [2.5, 1.76]}"
Use
$ rqt_graph
to show a graph with all involved nodes together with their topics.