Reinforcement Learning with rhino (not grasshopper)

Hi all. As part of my research, i am looking into setting up a RL model that would (ideally) send modelling commands to rhino and receive the geometry rhino outputs (for further processing).

note: For this research, i am omitting grasshopper altogether.

If anyone has experience with that, or has any clue on how to go about this, please provide some hints on how to set up a bidirectional communication channel between Rhino and an RL model. Cheers.

Hi,

Do you mean something like this Agents doing 2d drawing in Rhinoceros (adding dimensions and markings) ?
If you need custom tutoring to reach that result, or if you wish to have a software developed that integrates ML models / RL agents you can reach out to me at farouk.serragedine@gmail.com

Farouk

1 Like

Hi @farouk.serragedine. Your software looks nice, good work, and it falls under the same umbrella as what i’m after. I’m not looking to replicate a specific functionality at the moment, and this is for academic research, hence there is no funding to develop the software externally, nor will it ever be used commercially (this also comes from the fact that in academic research we might be interested in testing hypotheses that might never see practical implementation)

If you would be kind enough to provide an overview, in a few words, on how you went on about setting up the RL agent to communicate with rhino, that would be a lot of help. Cheers!

If anyone else with enough technical knowhow has a clue, any contribution is appreciated

a final attempt at getting some help on this :sweat_smile:
anyone?

Hi @adel.albloushi ,

I made a book about it, thought you might be interested.
It contains actually working sourcecodes to classify objects inside Rhinoceros with great explanations. 48 pages of actual content.

Farouk