🚀 Teaching the SO-100 to Lift: Human Teleoperation in Isaac Lab

Isaac Lab allows developers to simulate and train robotic behaviors efficiently. In this post, we’ll show how to use keyboard teleoperation to control the SO-100 robotic arm for a cube-lifting task. These expert demos can be used to train imitation learning policies.

👉 Follow this blog for full environment setup: Training SO-100 Robot for Cube Lifting in Isaac Lab

---

🧠 Why Teleoperate?

• Collect training data for imitation learning
• Prototype and debug new tasks
• Safe and repeatable in simulation

---

🛠️ Run the Teleoperation Script

python environments/teleoperation/teleop_se3_agent.py \
  --task Template-So-100-CubeLift-v0 \
  --num_envs 10 \
  --teleop_device keyboard

---

🎮 Keyboard Controls

Action	Keys
Move XYZ	W / A / S / D / Q / E
Rotate Axes	Z / X / T / G / C / V
Toggle Gripper	K
Reset	R

---

💾 What Gets Recorded?

Saved to datasets/:

• RGB-D camera data
• Joint & gripper states
• End-effector 6-DoF actions

Ready for use in:

• Behavioral Cloning
• DAgger
• Offline RL

---

✅ Summary

Use teleoperation in Isaac Lab to collect expert demonstrations and kickstart policy training — all within a scalable simulation framework.

---