AIM-Robotics

Unitree G1 robot development examples and utilities

---

Project Structure

AIM-Robotics/
├── SLAM/               # LiDAR SLAM system (Livox Mid-360 + KISS-ICP)
├── LiDAR/              # LiDAR streaming and visualization
├── RealSense/          # Intel RealSense camera utilities
├── YOLOv8n/            # YOLOv8 real-time object detection + streaming
├── audio/              # Audio recording and processing
├── Light/              # LED control utilities
├── debug_g1_loco.py    # Locomotion debugging utility
└── debug_g1_arm.py     # Arm control debugging utility

---

Component Documentation

• SLAM - Complete LiDAR SLAM system with real-time mapping
• LiDAR - LiDAR streaming and 3D visualization
• RealSense - RealSense camera integration
• YOLOv8n - YOLOv8 object detection with RealSense + UDP streaming

---

Prerequisites

Unitree SDK2 Python:

git clone https://github.com/unitreerobotics/unitree_sdk2_python.git
cd unitree_sdk2_python
pip3 install -e .

---

Development Environment

• Platform: Unitree G1 (Jetson Orin NX)
• OS: Ubuntu 20.04 (JetPack 5.1.1 / L4T R35.3.1)
• Python: 3.8
• CUDA: 11.4

---

GPU Setup for Deep Learning

Important: Standard pip install torch installs CPU-only PyTorch. For GPU acceleration on Jetson, use NVIDIA's official wheels.

Install PyTorch with CUDA Support (JetPack 5.1.x)

# 1. Remove CPU-only PyTorch if installed
python3 -m pip uninstall -y torch torchvision torchaudio
python3 -m pip cache purge

# 2. Install NVIDIA's PyTorch 2.1.0 for Jetson (CUDA-enabled)
export TORCH_WHL='https://developer.download.nvidia.cn/compute/redist/jp/v512/pytorch/torch-2.1.0a0%2B41361538.nv23.06-cp38-cp38-linux_aarch64.whl'
python3 -m pip install --no-cache-dir "$TORCH_WHL"

# 3. Install torchvision (may show version warning - safe to ignore)
python3 -m pip install --no-cache-dir torchvision

# 4. Reinstall PyTorch to ensure CUDA version is active
python3 -m pip uninstall -y torch
python3 -m pip install --no-cache-dir "$TORCH_WHL"

# 5. Verify GPU is recognized
python3 -c "import torch; print(f'CUDA available: {torch.cuda.is_available()}'); print(f'Device: {torch.cuda.get_device_name(0) if torch.cuda.is_available() else \"N/A\"}')"

Expected output:

CUDA available: True
Device: Orin

Troubleshooting

Issue: pip install ultralytics reinstalls CPU PyTorch

• Solution: Install PyTorch first, then install ultralytics with --no-deps flag

Issue: torchvision version mismatch warning

• Solution: Safe to ignore for YOLO - works with torch core dependencies only
• Optional: Build torchvision 0.16.0 from source for perfect compatibility (30-60 min build time)

  # Only if needed for torchvision.models or advanced features
  git clone https://github.com/pytorch/vision.git && cd vision && git checkout v0.16.0
  export TORCH_CUDA_ARCH_LIST="8.7" FORCE_CUDA=1 MAX_JOBS=4
  python3 -m pip install --no-build-isolation --no-deps -v .

References:

• NVIDIA PyTorch for Jetson
• PyTorch for Jetson Forum

---

Made with 💡 by AIM Robotics