1. System Architecture

Core computing (IK, Vision, VLA Policy inference) runs either on your Host Laptop or the onboard NVIDIA Jetson Orin Nano.

┌──────────────┐                  ┌────────────────────────┐
│  Bus 1 & 2   │      USB-C       │    Compute Node        │
│ (Follower    │◄────────────────►│                        │
│    Arms)     │                  │  - Jetson Orin Nano    │
└──────────────┘                  │         OR             │
                                  │  - Laptop / PC         │
┌──────────────┐      USB-C       │                        │
│    Bus 3     │◄────────────────►│   Running:             │
│ (Neck/Wheels)│                  │   - LeRobot Stack      │
└──────────────┘                  │   - Analytical IK      │
                                  │   - YOLO / VLA         │
┌──────────────┐      USB         └───────────┬────────────┘
│ Cameras (3x) │◄─────────────────────────────┘
└──────────────┘

2. Framework Installation

We recommend Ubuntu 22.04 and Python 3.10.

Clone & Install

Terminal
# Clone the unified pro repository
$ git clone https://github.com/Minko82/xlerobot-pro
$ cd xlerobot-pro

# Install dependencies in editable mode
$ pip install -e .
JETSON USERS: Ensure JetPack 5.1.x or 6.0 is installed. Enable "Max Power Mode" via sudo nvpmodel -m 0 for policy inference.

3. Environment & Port Configuration

A. USB Serial Permissions

Grant Permissions
$ sudo chmod 666 /dev/ttyACM*

# Verify your connected ports:
$ python find_port.py

B. Camera Verification

V4L2 Check
$ v4l2-ctl --list-devices

4. Control & Teleoperation

Test hardware integration using scripts in the examples/ directory.

ModeScript NameNotes
Joint Test0_joint_control.pyBasic movement validation
Cartesian IK1_ee_control.pyEnd-effector movement via IK
Dual Teleop2_dual_arm_teleop.pyKeyboard control for both arms
Full Platform3_full_xlerobot_control.pyArms + Neck + Base control

Troubleshooting Quick Reference

Full Diagnostic Guide ➔
Permission Denied

Run sudo chmod 666 /dev/ttyACM* to unlock USB ports.

Import Error

Ensure you ran pip install -e . inside the xlerobot-pro folder.

Ready to train your robot?

Start the First Reach Tutorial ↗