[ros-users] Advanced ROS Network Introspection (ARNI)

Andreas Bihlmaier andreas.bihlmaier at gmx.de
Tue Sep 30 11:57:07 UTC 2014

Dear ROS community,

I'm pleased to announce http://wiki.ros.org/arni - a collection of tools for Advanced ROS Network Introspection.
>From the wiki page:
"Advanced ROS Network Introspection (ARNI) extends the /statistics
features introduced with Indigo and completes the collected data with
measurements about the hosts and nodes participating in the network.
These are gathered from an extra node that has to run on each host
machine. All statistics or metadata can be compared against a set of
reference values using the monitoring_node. The rated statistics allow
to run optional countermeasures when a deviation from the reference is
detected, in order to remedy the fault or at least bring the system in a
safe state."

No modification of existing nodes is required in order to use the
monitoring features. Therefore, the barrier of entry is very low:
- See the arni tutorial
- git clone https://github.com/ROS-PSE/arni into your catkin_ws
- roslaunch arni_core init_params.launch
- start all your other nodes
- rosrun rqt_gui rqt_gui
- Plugins -> Introspection -> Arni-Detail
  (Click on an item (host, node, topic or connection) in the tree view
  to get more details and graphs in the other widget)
- Enjoy out of the box distributed metadata-based monitoring

If you want to use the more advanced features in your own ROS network,
see the documentation on how to write "specifications" and "constraints".

The documentation can be found in the wiki including the tutorials (http://wiki.ros.org/arni/Tutorials).

Please give feedback and report any bugs found.

Many thanks to my students that worked hard on this:
Matthias Hadlich, Matthias Klatte, Sebastian Kneipp, Alex Weber, Micha Wetzel

Kind regards,
Andreas Bihlmaier

Dipl.-Inform. Andreas Bihlmaier
Group Leader
Cognitive Medical Technologies Group

Karlsruhe Institute of Technology (KIT)
Institute for Anthropomatics and Robotics (IAR)
Intelligent Process Control and Robotics Lab (IPR)

Room: 002.2
Engler-Bunte-Ring 8, Building 40.28
76131 Karlsruhe, Germany

eMail: andreas.bihlmaier at kit.edu
Tel: +49 721 608 44263
Fax: +49 721 608 47141


More information about the ros-users mailing list