Urban Policy Simulator ( AI-Driven City Planning )

Simulate city traffic, emissions, and citizen satisfaction using AI to test policies before implementation, enabling data-driven urban planning and sustainable mobility.

Comment

About the Project

Inspiration

Urban planning is complex and often relies on intuition or historical data. Mistakes can be costly for traffic, emissions, and public welfare. We wanted to create a safe, interactive, AI-driven simulator that allows city planners and policymakers to test policies before implementing them in the real world.

What it Does

Urban Policy Simulator lets users:

  • Select pre-set scenarios like “EV-first City” or “Green Mobility Plan.”
  • Adjust parameters such as bus frequency, bike lanes, EV adoption, or congestion tolls.
  • Predict the impact of these policies on traffic flow, emissions, and citizen satisfaction using a TensorFlow.js neural network.
  • Visualize results on a 5×5 city grid, with real-time charts for comparison.

How We Built It

  • Frontend: HTML, CSS, JavaScript for interactive city grid and UI.
  • AI Model: TensorFlow.js neural network trained to predict traffic and emissions based on policy inputs.
  • Scenarios: Pre-set templates for quick experimentation.
  • Data Integration: Option to import open datasets from real cities for realistic simulations.

Challenges We Ran Into

  • Balancing simulation speed vs. accuracy of predictions.
  • Designing a simple UI that communicates complex data clearly.
  • Integrating AI predictions with dynamic visualizations in real time.

Accomplishments We're Proud Of

  • Functional AI-driven simulator that updates instantly.
  • Pre-set policy templates demonstrating diverse city strategies.
  • Clear visualization of policy impacts, helping users make informed decisions quickly.

What We Learned

  • Integrating predictive modeling with user interaction is challenging but highly rewarding.
  • Even a small neural network can provide actionable insights when designed carefully.
  • Clear data visualization is critical for non-technical users to trust AI predictions.

What's Next

  • Integrate real-world city datasets for more realistic simulations.
  • Expand the neural network to predict long-term impacts of policies.
  • Enable collaborative planning for teams of city planners and stakeholders.
  • Add more scenarios and policy options for broader experimentation.

Built With

Share this project:

Updates