Skip to content

ArynShah/aurai

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

23 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

aurai

AURAI – AI-Driven Transit Optimization Simulator

AURAI is an interactive transit simulation that models a large urban subway system and uses reinforcement learning to optimize train frequencies based on passenger demand, congestion, and operating cost. The project combines graph theory, simulation modeling, and real-time visualization.

FEATURES

Large-scale subway network modeled as a graph of stations and connections

Real-time passenger flow simulation with time-of-day demand patterns

Live performance metrics including congestion, cost, and customer satisfaction

Reinforcement learning agent that dynamically adjusts train frequencies

Optional AI-generated daily performance summaries

Interactive graphical interface for visualization and control

TECH STACK

Language

Python 3

Libraries

pygame – graphics, UI, and real-time interaction

networkx – graph modeling and routing

numpy – numerical operations and calculations

Core Concepts

Graph theory

Reinforcement learning (Q-learning)

Simulation modeling

AI-assisted decision making

INSTALLATION

Requirements

Python 3.9 or newer

pip

Install dependencies pip install pygame networkx numpy

HOW TO RUN

Run the main application file python app.py

The application window will automatically scale to your screen resolution.

HOW THE SIMULATION WORKS

The subway system is represented as a graph where nodes are stations and edges are track segments. Passengers are generated dynamically based on the time of day, with higher demand during morning and evening rush hours.

Passengers travel through the network using shortest-path routing. Each connection tracks passenger load, train count, capacity, and congestion levels. These values directly influence system performance metrics.

AI AGENT

The AI agent uses reinforcement learning to decide whether to increase, decrease, or maintain train frequency on different subway lines. Decisions are based on current congestion levels and time-of-day conditions.

The reward function balances customer satisfaction against operating cost, allowing the agent to improve its strategy gradually over time.

NOTES

This project is a simulation intended for experimentation and visualization. Passenger behavior and system dynamics are simplified to improve clarity and performance.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 100.0%