<aside> ⏰

시간 : 매주 수 오후 10시

</aside>

<aside> 📢

세미나 진행 시 잊지 말고 사진 촬영부탁드리고, 발표자분들께서는 자료 업로드해주세요! 또한 아래 사진란은 만지지 말아주세요!

</aside>

https://github.com/kaist-include/include_hexapod

팀장 임시현

Isaac Lab Tutorials - Basics

I Tried To Build a Robot Like Boston Dynamics With Isaac Sim

include_hexapod

Welcome! This is the main repository for Include Hexapod RL project.

We focus on practicing deep reinforcement learning to control a physical hexapod robot (Hiwonder SpiderPi).

Our short-term goal is to train the controller to perform various tasks:

• Basic real-time locomotion with stable base frame with different choice of active legs
• Vision-aided localization and locomotion

Specifications

• Platform: Hiwonder SpiderPi; 18 DOF hexapod robot with IMU, ultrasonic sensor, and a simple camera.
• GPU/Traning machine: Remote server (insert spec here)
• Environment and RL platform: NVIDIA Isaac Gym / Stable-Baselines3

Useful Links

List of documents that we reference through our journey:

Physical Platform Introduction and Setup