Software Setup

Configure Raspberry Pi

1. Open a terminal and run:

   sudo raspi-config
2. Update the configuration tool through Update.

3. Under Interface Options, enable VNC, SPI, and I2C.
4. Set audio settings under Advanced Options → Audio Config → Pipewire.
5. Reboot the system:

   sudo reboot

Install and Configure LCD-show

1. Clone and install the LCD-show driver:

   git clone https://github.com/goodtft/LCD-show.git
   cd LCD-show
   chmod +x LCD35-show
   sudo ./LCD35-show

2. Modify boot configuration for HDMI resolution with VNC:

   sudo nano /boot/config.txt

   Add or update:

   hdmi_force_hotplug=1
   hdmi_group=2
   hdmi_mode=82  # 1920x1080 resolution
   framebuffer_width=1920
   framebuffer_height=1080

3. Modify boot configuration for screen rotation:

   sudo nano /boot/config.txt

   Add:

   dtoverlay=tft35a:rotate=270

4. Restart the system:

   sudo reboot

Touchscreen Calibration

1. Install the Calibration Tool

sudo apt-get install xinput-calibrator

2. Run the Calibration Tool

xinput_calibrator 

Touch the targets as instructed. The tool will generate calibration offsets.

3. Save the Calibration Data Copy and Paste the calibration data and save the file. Save the output data to a configuration file:

sudo nano /usr/share/X11/xorg.conf.d/99-calibration.conf 

4. Reboot

sudo reboot

5. Verify Calibration Interact with the desktop to ensure accurate touch. If needed, rerun the calibration tool or manually adjust the values in 99-calibration.conf.

Clone the TARS-AI Repository

1. Open a terminal on your Raspberry Pi.
2. Clone the TARS-AI repository:

   sudo apt install python3-venv -y
   git clone https://github.com/TARS-AI-Community/TARS-AI
3. Navigate to the cloned directory:

   cd TARS-AI

Install System Dependencies

To ensure smooth operation, use the included install script.

1. Update your system:

   sudo apt update
   sudo apt upgrade -y
2. Update script permissions:

   chmod 777 Install.sh
3. Allow the script to be executable:

   chmod a+x Install.sh
4. Run the provided script:

   sudo ./Install.sh
5. Fix permissions:

   sudo chown $(id -u):$(id -g) .env
   sudo chown $(id -u):$(id -g) src/config.ini

Connect Hardware

1. Microphone: Plug a USB microphone into your Raspberry Pi.
2. Speaker: Connect an external speaker via audio jack or Bluetooth.

Install Audio AMP

1. Edit the boot configuration file:

   nano /boot/config.txt

2. Add the following lines at the end of the file:

   dtparam=audio=on
   dtoverlay=hifiberry-dac
   dtoverlay=i2s-mmap

3. Install required dependencies and setup I2S amplifier from the TARS-AI Folder:

   source src/.venv/bin/activate
   sudo apt install -y wget
   pip3 install adafruit-python-shell
   wget https://github.com/adafruit/Raspberry-Pi-Installer-Scripts/raw/main/i2samp.py
   sudo -E env PATH=$PATH python3 i2samp.py

   ⚠️ DO NOT run audio in the background, as it will break VAD (Voice Activity Detection) and RMS (Root Mean Square) processing.

4. Reboot the system:

   sudo reboot

5. After rebooting, activate the environment and re-run the setup script from the TARS-AI folder:

   source src/.venv/bin/activate (if not already activate)
   sudo -E env PATH=$PATH python3 i2samp.py

6. Test the speaker output:

   • When prompted, say yes to the speaker test and listen for sound.
   • If the audio sounds distorted, reduce the volume using alsamixer.

7. Enable volume control in Raspbian Desktop or RetroPie:

   • Reboot the system again:

     sudo reboot
   • After the second reboot, open alsamixer:

     alsamixer
   • Use the arrow keys to lower the volume (50% is a good starting point).

8. If audio issues persist:

   • Re-run the i2samp.py script.
   • When prompted, choose N (disable) for the /dev/zero playback service.

Configure API Keys

To enable LLM and TTS functionality, configure your API keys.

1. Edit the .env file and replace "your-actual-api-key" with your valid keys:

   # LLM API Keys
   OPENAI_API_KEY="your-actual-openai-api-key"
   OOBA_API_KEY="your-actual-ooba-api-key"
   TABBY_API_KEY="your-actual-tabby-api-key"
   
   # TTS API Keys
   AZURE_API_KEY="your-actual-azure-api-key"

Get Your API Keys:

• OpenAI API Key
• Azure Speech API Key (FREE)
  • Make sure to create a Free Azure account Free Azure Signup
  • Follow all the steps in the video up to Install Azure speech Python package.

Set Up Configuration Parameters

1. Open src/config.ini and update the LLM settings:

   [LLM]
   llm_backend = openai
   base_url = https://api.openai.com
   openai_model = gpt-4o-mini
2. Update the TTS (Text-to-Speech) settings:

   [TTS]
   ttsoption = azure
   azure_region = eastus
   tts_voice = en-US-Steffan:DragonHDLatestNeural

Run the Program

1. Navigate to the source directory:

   cd src/
2. Start the TARS-AI assistant:

   python app.py
3. TARS-AI is now running! Speak into the microphone and listen to responses from the speaker.

TARS-AI can integrate with Home Assistant to enable smart home automation via voice commands.

The Home Assistant integration allows TARS to:

• Process voice commands and trigger smart home actions.
• Interact with connected devices (lights, thermostats, security systems, etc.).
• Leverage Home Assistant’s Assistant API to send prompts and execute actions.

To enable this integration, TARS-AI must be configured to communicate with your Home Assistant server.

Prerequisites

Before setting up integration, ensure:

1. Home Assistant is installed and running on your local network.
2. TARS-AI is fully set up on your Raspberry Pi.
3. A Home Assistant long-lived access token is generated.

Configure Home Assistant in config.ini

1. Open src/config.ini and locate the [HOME_ASSISTANT] section.
2. Enable the Home Assistant module and configure the connection URL:

   [HOME_ASSISTANT] # HA Module
   enabled = True
   # If set to False, the Home Assistant module will be disabled.
   url = http://homeassistant.local:8123
   # Example: http://192.168.2.5:8123
   # URL to access Home Assistant (set Token in .env file!).
3. Adjust the URL if your Home Assistant is running on a different address.

Generate and Add the Access Token

TARS-AI needs a long-lived access token to authenticate with Home Assistant.

1. Log into Home Assistant.
2. Click on your User Profile (bottom left corner).
3. Scroll down to the Security section.
4. Locate “Long-lived access tokens” and click “Create Token”.
5. Enter a name for your token and click OK.
6. Copy the token immediately, as it won’t be visible again.

Store the Access Token in .env

1. Open the .env file in your TARS-AI directory.
2. Add your Home Assistant Token:

   HA_TOKEN="INSERT-YOUR-TOKEN-HERE"
3. Save the file.

Restart TARS-AI to Apply Changes

1. Navigate to the src/ directory:

   cd src/
2. Restart the application:

   python app.py