Open-TI: Open Traffic Intelligence with Augmented Language Model

Introduction

Open-TI is a revolutionary traffic intelligence model that bridges the industry-academic gap in intelligent transportation. It leverages large language models to execute complex traffic analysis tasks, making it the first to seamlessly integrate external packages based on conversations. Beyond analysis, Open-TI can train traffic signal control policies, explore demand optimizations, and communicate with control agents like ChatZero for efficient task execution. With a formal structure and open-ended design, Open-TI invites community-driven enhancements, emphasizing its pivotal role in advancing intelligent transportation systems.

The overview of the Open-TI functionalities

The design framework of Open-TI

The Open-TI conversation interface

Installation

Source

Our code is based on Python version 3.9 and Pytorch version 1.11.0. Open-TI provides installation from the source code. Please execute the following command to install and configure our environment.

clone https://github.com/DaRL-LibSignal/OpenTI.git
cd OpenTI
pip install -r requirements.txt
sudo apt-get install python3-tk

Configuration

LLM Configuration

First, you need to configure OpenAI-Key. Please create a ./config.yaml file in the root directory and add the following content to the file (please modify the content to your own settings):

OPENAI_API_TYPE: 'openai' #'azure'  OR 'openai'
# for 'openai'
OPENAI_KEY: '' # your openai key
Google_Map: '' # your googleAPI key

Here we recommend using at least ChatGPT-3.5-turbo to run as LLM. If you want to use your own LLM, please refer to LangChain-Large Language Models to define Your own LLM. In this case, please modify the following sections in ./DataProcessBot.py and ./SimulationProcessBot.py to configure your own LLM.

Simulator environment configuration

Though CityFlow, SUMO, and LibSignal are stable under Windows and Linux systems, we still recommend users work under the Linux system.

CityFlow Environment

To install CityFlow simulator, please follow the instructions on CityFlow Doc

sudo apt update && sudo apt install -y build-essential cmake

git clone https://github.com/cityflow-project/CityFlow.git
cd CityFlow
pip install .

To test configuration:

python3
import cityflow
env = cityflow.Engine

SUMO Environment

To install SUMO environment, please follow the instructions on SUMO Doc

sudo apt-get install cmake python3 g++ libxerces-c-dev libfox-1.6-dev libgdal-dev libproj-dev libgl2ps-dev swig

git clone --recursive https://github.com/eclipse/sumo

export SUMO_HOME="$PWD/sumo"
mkdir sumo/build/cmake-build && cd sumo/build/cmake-build
cmake ../..
make -j$(nproc)

If you cannot make the directory sumo/build/cmake-build, you could try the command below

cd sumo
mkdir build
cd build
mkdir cmake-build
cd cmake-build

To test installation:

cd ~/sumo/bin
./sumo

To add SUMO and traci model into the system PATH, execute the code below:

export SUMO_HOME=~/sumo
export PYTHONPATH="$SUMO_HOME/tools:$PYTHONPATH"

To test configuration:

python3
import libsumo
import traci

LibSignal Environment

To install LibSignal environment, please follow the instructions

cd LibSignal
pip install -r requirements.txt
pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu113
pip install cmake
pip install lmdb

Start Running

If you install all the requirments below, you can run Open-TI now.

cd OpenTI
cd pivotalAgent
python3 executor.py

If you find this project helpful, please use below Citation

@article{da2024open,
	title = {Open-TI: Open Traffic Intelligence with Augmented Language Model},
	author = {Da, Longchao and Liou, Kuanru and Chen, Tiejin and Zhou, Xuesong and Luo, Xiangyong and Yang, Yezhou and Wei, Hua},
	date = {2024/10/01},
	doi = {10.1007/s13042-024-02190-8},
	id = {Da2024},
	isbn = {1868-808X},
	journal = {International Journal of Machine Learning and Cybernetics},
	number = {10},
	pages = {4761--4786},
	url = {https://doi.org/10.1007/s13042-024-02190-8},
	volume = {15},
	year = {2024}
}