Quick Start
Get from unboxing to your first recording or stream in five steps. This page covers the end-to-end workflow; each step links to a detailed guide.
Setup Device → Install SDK → Authenticate → Record → Process with MPS
↘ Stream (real-time)
Prerequisites
- Aria Gen 2 glasses (powered on and charged)
- Aria Companion App installed and paired with your device (Companion App guide)
- USB cable connecting the device to your PC
- Python 3.10 - 3.12 installed (python.org)
Step 1: Install the Client SDK
Create a virtual environment and install the SDK:
python3 -m venv $HOME/projectaria_gen2_python_env
source $HOME/projectaria_gen2_python_env/bin/activate
pip install projectaria-client-sdk
See Installation for platform-specific details.
Step 2: Configure Your System
Run aria_doctor to open the required ports and configure USB networking:
aria_doctor
Step 3: Authenticate Your Device
Connect the device via USB, then pair it with your PC:
# Verify the device is detected
aria_gen2 device list
# Authenticate (approve the request in the Companion App)
aria_gen2 auth pair
# Confirm authentication
aria_gen2 auth check
You only need to authenticate once per device-PC combination. See Device Authentication for troubleshooting.
Step 4a: Record Data
Start a recording, collect data, then download it:
# Start recording
aria_gen2 recording start --profile profile8 --recording-name my_first_recording
# ... perform your activity ...
# Stop recording
aria_gen2 recording stop
# List recordings and note the UUID
aria_gen2 recording list
# Download the recording
aria_gen2 recording download -u UUID -o ~/Downloads/
See Recording Control for profiles and advanced options.
Step 4b: Stream Data
Stream sensor data in real time and visualize it:
# Start streaming (USB)
aria_gen2 streaming start
# In a second terminal, open the viewer
aria_streaming_viewer --real-time --interpolate --rerun-memory-limit 4GB
# When done, stop streaming
aria_gen2 streaming stop
See Streaming Control for wireless streaming and thermal management.
Step 5: Process with MPS (Optional)
Submit recordings to Machine Perception Services for cloud-based 3D trajectory and hand tracking:
See Machine Perception Services to get started.
What's Next?
| Goal | Page |
|---|---|
| Control the device with Python scripts | Program with SDK |
| Use wireless streaming | Streaming Control |
| Stream to a remote server | Remote Streaming |
| Validate recording quality | Data Validation Tools |
| Troubleshoot issues | Support & Troubleshooting |