Skip to content

Question About Real-Robot Experiments and Code Availability #2

Description

@imchy0212-ui

Dear authors,

Thank you for sharing your work and the implementation of “Waypoint-Based Reinforcement Learning for Robot Manipulation Tasks.”

I am a student at South China Normal University, and my research also focuses on applying reinforcement learning to robotic manipulation. I am currently trying to reproduce your work and adapt the waypoint-based approach to my own robotic arm.

So far, I have successfully reproduced the simulation experiments. However, while preparing to transfer the method to a real robot, I encountered several questions, particularly regarding:

  • the design and computation of the reward function in the real-robot setting;
  • the safety constraints or protective mechanisms used during robot-environment interaction and online training;
  • any additional practical considerations when transferring the policy from simulation to a physical robot.

Would it be possible to share the code used for the real-robot experiments, or any relevant implementation details, for academic reference? If the real-robot code cannot be made public, I completely understand. In that case, may I ask you a few specific questions about the real-robot setup and training procedure?

Any guidance would be greatly appreciated. Thank you very much for your time and for making this project publicly available.

Best regards,
A student from South China Normal University

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions