Join the Discord Server
An experimental open-source application that integrates the OpenAi gpt-3.5-turbo-0125 or gpt-4-turbo models via API, NLP, TTS, STT, and an OBD-II ELM327 device to create a voice-activated, hands-free, vehicle diagnostic assistant.
03/17/2024: Updated to use OpenAI v1
11/14/2023: Added "development" branch with the option to use text input in the terminal instead of voice commands.
git clone https://github.com/<your_username>/Automotive-AI.git./install.shSet your API keys and variables in .env.template and save it as .env
cp .env.template .envWithout a vehicle communication interface:
python -m appWith an ELM327 device connected:
python -m app --device elm327.install.batSet your API keys and variables in .env.template and save it as .env
copy .env.template .envWithout a vehicle communication interface:
python -m appWith an ELM327 device connected:
python -m app --device elm327To start a conversation that uses JSON for conversation history, use the "start a conversation" command.
After a conversation has been started you can use the following voice commands to manage the conversation history:
After installing com0com, run:
elm -p COM6 -a 500000Set the COM port in the .env file to COM7.
python air_fuel_datastream.pyStreams data from the OBD-II ELM327 device to the console, but there's currently no way to stop the stream other than closing the application.