ForesightBot – AI Public Space Risk Predictor

• 

  ForesightBot is an AI-powered system that predicts and prioritizes problems in public and campus spaces, helping identify risks early and en

Inspiration

What it does

How we built it

Inspiration

Public and campus spaces can develop problems such as crowding, waste, blocked paths, accessibility issues, and safety risks. We wanted to create a simple AI-based system that can identify and prioritize these problems before they become bigger issues.

What it does

FORESIGHTBOT is an AI-based prediction and prioritization system for public and campus spaces. It analyzes location and problem-related factors to predict a Risk Score and Priority Score. It then classifies problems as LOW, MEDIUM, or HIGH priority so that areas needing attention can be identified quickly.

How we built it

We built FORESIGHTBOT using Python in Google Colab. Pandas and NumPy were used for data processing, Matplotlib for visualization, and Scikit-learn's Random Forest Regression for risk prediction.

The dataset includes locations such as Main Gate, Library, Canteen, Block A, Block B, Parking Area, Bus Stop, and Lab Area. The system considers problems such as Crowding, Waste, Blocked Path, Accessibility, and Safety.

Results

The model achieved an MAE of 4.77 and an R² score of 0.85 on the test data. The system can convert predicted risk into practical priority levels for faster inspection and response.

What we learned

We learned how to prepare data, train and evaluate a machine-learning model, interpret predictions, and connect AI results with a practical real-world problem.

What's next

In the future, FORESIGHTBOT could be connected with cameras and sensors to collect real-time information and automatically monitor public and campus spaces.

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for ForesightBot – AI Public Space Risk Predictor

Built With

• artificial
• google-colab
• machine-learning
• matplotlib
• numpy
• pandas
• python
• random-forest-regression
• scikit-learn