Involvement of 3rd parties in students robots

1

Hello everyone,

in the last years we noticed an increasing amount of schools in Germany that develop their own internal robotics hardware/software platform for the use of their students. I think this is generally a great thing. These platforms do no more than i.e. the LEGO or Fischertechnik systems (make the hardware more accessible to the students). So basically the hardware drivers and parts of the hardware are made by the mentors (or possibly even other students in the school) and the core RoboCup software is made by students. This should be in compliance to line rule 2.2.1/maze rule 2.2.3:

Any robot kit or building blocks, either available on the market or built from raw hardware, may be used as long as the design and construction of the robot are primarily and substantially the original work of the students

I am not sure about Soccer and OnStage, but I think this is an issue that might be relevant to all leagues, why I am publishing this under General. There is a rule quiet similar in soccer (refer 8.2.2).

In our opinion this is an unfair advantage over schools that don’t have partnerships with companies or teachers with the expertise to develop such a kit or software.
In my opinion, it should be mandatory for the schools to publish their development to make it accessible to other schools. Otherwise at some point it will be very hard for new schools to participate in the competition.
Alternatively the rule I quoted could be enforced by completely prohibiting 3rd parties involvement. But I don’t think that is feasible, because the border between self made and made by 3rd party is not clear at all (technically all students would have to built their robot from scratch and kits would have to be completely forbidden).

What is all TCs opinion on this? I am looking forward to your responses!

Best,
Jan

2

Hi @jblumenkamp,

The biggest problem with this rule that I have encountered are usually connected to the word substantially. The decision then boils down to the subjective view on what is and is not substantial original work, which does not necessarily need to be the same even among the members of the TC.

Making it mandatory is to some extend problematic, because as I see it we could run into issues with what does it actually mean to publish their development? Are blueprints enough? Would technical documentation accompanied with photos be sufficient? While we are at it, what actually is technical documentation?

I believe we are looking at a pretty big problem here and we are not alone. The scientific world has problems coping with at well. Despite various advantages of modern science communications, it is still quite problematic (and sometimes outright impossible) to reproduce published and well received scientific papers.

My personal opinion is that the least expensive way out may be via much bigger push for openness. I do not think that can be done on the rules side of things, but it can probably be incentivized by having specific rules for awards at the international competition, similar to what the scientific community does. A “No best paper award if we cannot reproduce your work” can be translated in RCJ terms as “No trophy if you do not publish sufficient information for building your robots”.

As we can see at this very forum, the interest for technical documentation does exist.

I believe this is sadly already happening to some extend, as the international competition seems to host more or less the same teams over the past few years. That is not necessarily a bad thing, but we may also view it as somewhat alarming, as it probably is increasingly difficult for new schools to participate as you note.

Thank you for staring this thread – I hope we can get something tangible out of it.

– Marek

3

Making RCJ a completely open source event would go a long way to solving this. The TC could hold all materials in escrow until the international events are complete and then release them on somewhere like git.

I have been advocating for this for 4 years now. It is not that hard to do as we collect everything for the technical evaluations.

4

I do agree as well, as I am not sure how could one disagree with something like this :slight_smile: Did you meet any substantial opposition? I would be interested in hearing such arguments.

My problem with making it mandatory as @jblumenkamp suggested above is that just releasing everything we receive for the technical evaluations may not actually be that helpful to other teams, but would certainly be a start. The Soccer OC for instance usually does not get technical portfolios from all teams, so releasing just the materials of those that provide something may fell a bit uneven.

Long story short, I am all for making RCJ a completely open source event, but I would also like to make sure that the source that gets opened is actually useful.

– Marek

5

Great that there is so much support regarding this!

I agree it is hard to define. The published information should be sufficient to at least understand what the team was doing and why, so that other teams can understand. In case hardware and software is developed by 3rd parties, everything should be published in a way that it is reproducible (i.e. a detailed Bill of Material for a PCB, manufacturing data for a PCB, the Software and at least some basic instructions).[quote=“mareksuppa, post:2, topic:570”]
My personal opinion is that the least expensive way out may be via much bigger push for openness. I do not think that can be done on the rules side of things, but it can probably be incentivized by having specific rules for awards at the international competition, similar to what the scientific community does.
[/quote]

I just checked, in Rescue we actually have this rule:

4.3.2 Teams awarded with certificates are required to post their documents and presentation online when asked by the OC/TC.

It was probably just not enforced. This rule mentions documents, I think software and hardware documentation are documents as well, but it is not clear.

This forum would be a great place to publish this, but a new section for sharing would be necessary. Teams that want to share their robots could do it there voluntarily or after being asked.

The question is if it is good if teams are forced to publish everything. It would be better to encourage them, but it should be the teams decision.
Actually in the Majors SPL league teams are only required to release their code without code that defines the behavior of their robot and perhaps configuration files. Otherwise all teams would use the last years winners code, since the hardware platform is the same. I see a similar problem in Junior - if a team develops a great robot based on LEGO and needs to publish detailed building instructions and all the core software, there would be many lookalike robots with identical software in the next year. I understand that this is generally desirable in a research competition, but in Junior possibly many teams would not learn as much as if they have built it on their own.

Also my original question referred to the work not done by the students. It’s a difference, since what the mentors etc… provide is on the same level like LEGO, but usually does not solve core RoboCup tasks.

To summarize, I am a supporter of open source, but there are some understandable reasons why it might not be good to make the whole RoboCup Junior competition open source. Teams should have a platform to share and be encouraged to use it (like the forum). But it should be mandatory to publish work not done by the students and not already accessible for everyone else (if, for example a mentor develops a robotics platform and makes it available commercially world wide that is fine, but if the mentor makes it only for the own school, it needs to be published in a way that every other team and competition organizers can reproduce and understand the work).

Best,
Jan

6

I would love to hear more opinions and thoughts on this :slight_smile: