Hello there,
I am creating the digital twin of the SES-V2 arm. I implemented a simulator environment and controlled a real robot through IsaacLab .
It is quite handy because I did not need to use MoveIt of ROS. The IsaacLab has a built-in Inverse Kinematics function.
Please check the source code from the GitHub link below.
Below is an experiment testing the conveyor belt function. It is very convenient because the conveyor can be controlled entirely with Python code.
I need some feedback, so it would be great if you could leave it in the comments.
Thank you.
@dialfonzo Thank you for the response!
However, it is still good for simple actions like picking up and placing down a cube.
It was a really good experience to learn about a manipulator robot.
kimbring2:
I need some feedback
Very cool! Would be great to see the setup with some additional views. This is largely the same type of setup humanoid manufacturers are using to train their robots on tasks.
One trend I’ve been seeing is to use external cameras and an external system to the robot:
https://arxiv.org/pdf/2604.10096
What kind of feedback are you looking for?
Hello there,
Thank you for sharing an interesting project about the OpenClaw. I tried to use it last week to create the game.
The OpenClaw worked really well. I was pretty impressed by the possibility of it.
Besides, I gave up using it because I can not use the OpenClaw through a monthly subscription of the Claude, Gemini, and ChatGPT.
Instead, I should use it through the API, and that was expensive. I spent $10 for a simple few commands.
That method of the paper you shared might be a bit difficult use in my work to for that budget reason.
P.S. Not tried, but I heard that running OpenClaw locally doesn’t have good performance.
Thank you.
