[ros-users] [Discourse.ros.org] [General] ROSCon 2018 Program Published

v4hn ros.discourse at gmail.com
Wed Aug 15 17:37:22 UTC 2018

Hello everyone,

It was quite hard for me to write a follow-up to @tkruse's last reply, so I exchanged some private messages with members of the program committee instead. Still, I fail to understand why @awesomebytes, @andreucm and @Martin_Guenther feel relieved by @gerkey's explanation and thus want to give my personal response here in public. I expect it represents the opinion of a significant fraction of the contributing ROS community.

I agree with a lot of what @tkruse means to convey, but the way he puts it is neither productive nor easy to agree with. The discussion about whether or not OSRF should be funded by military organizations in the future (too) is a valid one, but it is per se unrelated to the program of ROSCon and thus does not belong here. I am sure we will see a lot of discussion on that at the conference and I believe constructive statements in a different thread are welcome.

It seems there were a number of submissions for this ROSCon on military-related projects and the program committee had to decide whether or not they want to have these talks at the conference.
Looking through the program, the outcome was to accept two talks with obvious military funding or background, the one by Alejandro R. Mosteo (Centro Universitario de la Defensa, Zaragoza) on Ada support for ROS2 and the one by Jerry Towler, apparently summarizing military projects for the US military.

I object to the scheduled talk by Jerry Towler.

The abstract does not describe any project with civil use cases (or any concrete project at all for that matter).
After reading it many times, I would expect him
- to introduce ''ROS-M'', maybe comparable to ``ROS-I'' but with more centralized funding,
- summarize some non-classified portions of US-military robotics projects in the past,
- show how they can benefit from using ROS,
- explain how they have interesting research problems also in uncontroversial scenarios,
- state that they will open-source everything
- and thus not touch weapon controls.

Don't get me wrong, I believe this can be an interesting overview talk.
``ROS-M'' might even be an interesting movement, given they open-source all their code.
I would like to see the presentation as a 20-minutes video on vimeo.
But I do not want it to be part of the ROSCon conference.

ROSCon is an event to build and support lobbies and communities around ROS and software projects.
The committee chooses which lobbies are supported by deciding on talks, within the quality and number of submissions they receive.
But some communities and lobbies do not mix well.
Ryan Gariepy said at the end of last year's ROSCon something like
"We would never have imagined to have a conference where we talk about robot coloring books and fully autonomous S-classes at the same event".
Imagining this year's statement additionally contains "and autonomous convoy vehicles for border patrol", or anything comparable, makes me cringe.
In my opinion, military applications of ROS, though they obviously exist, have no place in the worldwide happy-go-lucky, lobby-supporting community that OSRF always advertises at their events.

Discussing military applications in the field of autonomous robotics breaks a taboo for many academics - especially in places where funding does not in large part come from the domain. At least in Germany, there is a long-standing debate about the so-called ["Zivilklausel"](https://de.wikipedia.org/wiki/Zivilklausel), which requires universities that accept it to reject all research and collaboration in projects with military objectives, sometimes even potential military applications.

All in all, I do not oppose talks by military-related researchers or projects in general.
They might present novel concepts or software modules that are useful for many civil applications as well, and if they do not only focus on how military robotic systems can benefit from them, they are clearly interesting for every researcher.
But I believe the ROS community - to a significant proportion represented at (and by) ROSCon - should not feature military applications and a talk titled ''ROS-Military: Progress and Promise'' crosses that line.
This has nothing to do with ''hiding objectionable projects'', it is a matter of public support.

[Visit Topic](https://discourse.ros.org/t/roscon-2018-program-published/5537/9) or reply to this email to respond.

More information about the ros-users mailing list