Open yani-rl-ai opened 3 years ago
Hi @ROSNovice, sorry for the late reply. This would depend on how you have structured your code for the HSR and whether your robot components are compatible with Python 3. If they are, it's rather easy, as you could just instantiate a model directly in your code and then pass the required data to obtain execution parameters.
In my case, I used mas_domestic_robotics
for the experiments, where the components of interest only work with Python 2 at the moment. I thus used ROS for communication to obtain execution parameters. For instance, in the case of the handle grasping action:
Object
message)Note that this workflow was just so that I can do the experiments, but, in the long run (once we migrate our domestic robotics repository to a newer ROS version and Python 3), I will simply instantiate an execution model directly in the implementation of the action.
In any case, I will push the scripts that were used to obtain execution parameters ~later this week~* so that you can get an idea about what was happening there. You can then hopefully adapt them to your needs.
* I will only have time to do this after the ongoing ICRA.
Hi @alex-mitrevski thank you for sharing your work. I interested with this development. i also the HSR user. Can you share the detail how to deploy your code in Toyota HSR? Thanks