Pypot - V-REP sneak peek

Switch between real and simulated poppy in a single line of code…


Thanks Pierre it seems that V-rep works on most of platform!

Well the licensing is not fully free, but that seems to be clearly advance as possible as they can in an educative way. (which is great!)

Hello to Ice Cube a great philosopher of the XXI century.

Great @Pierre, I’m really impressed you managed to do it in just a couple of day.

Can you explain a bit how you plug it with pypot ? Juste another io ?

@Matthieu, to sum up, yes it’s another IO with the associated controller. The rest is rather straightforward.

I haven’t pushed anything yet as I’m taking this opportunity to refactor some code and makes it more obvious how to add other IOs (for instance to add support for other types of motor ;)). It should be ready really soon with some architecture explanations in a dedicated section of the doc.


Really great @Pierre !

So if i have well understand, it will be possible to simulate evaluation of poppy without having one, like `Yarp’ for Icub?

Does it will include the cobot part?

Yes that’s the broad idea. But I still don’t know how far we will be able to simulate poppy into V-REP: accuracy of the physical dynamics? sensors? motors behaving as the dynamical ones?

I guess you mean Human-Robot Interaction? That’s another topic that need investigation.

Yarp do it approximately correctly on all of that. Yes some incoherence exist, but it is really impressive. The missing side is more what V-REP seems to offer the way to modify “directly” part of the bot and software of the bot. (which is a bit more “tricky” with Icub)

Not only, but interaction with all the things around it. As Pierre-Yves Oudeyer qualify the “robot curiosity” (that is part of cobot (fr: cobotique).

Hi all,

@aurelien, I do not understand what you mean by saying that “robot curiosity is part of cobot”. Could you precise your point please?

Hi @Clement i speaking for part of

more information at
and wikipedia … and certainly INRIA.

If that type of software are not part of Poppy, it should be great to have them and share :slight_smile:
Then if V-REP offer an approximately true “simulation” it could be used without building the bot but improving that part. (which interest me (a lot) Thanks.

The paper you mention doesn’t seem to be related to cobot (more precisely, searching ‘cobot’ in the pdf doesn’t return any result). Could you please tell me what part of the paper you’re talking about?

If by “that type of software” you mean “software implementing artificial curiosity”, you could have a look at the Explauto library we are currently working on with @Pierre, available here:
But please note that it is just a beta (if not alpha) version for the time being.

Thanks for the link (i have already got this one) is it the full software on the subject use by poppy?

Maybe the term “Cobot” is not the right one
and it is more

The paper do not mention that word, but you will find at lot of works that are not qualify to be cobotique and cognitique but that is what it seems they are to be.


@Clement but you make me see i have miss the important

This topic was automatically closed after 24 hours. New replies are no longer allowed.