SafePath AI
AI Powered Crime Prediction and Safe Route
Recommendation System
Project Overview
SafePath AI is an AI/ML based smart city solution that predicts crime risk in urban areas and
recommends the safest travel route instead of only the shortest route. The system combines
historical crime data, location, time and machine learning models to estimate safety levels.
Problem Statement
Existing navigation systems mainly optimize distance and travel time. They do not consider crime
patterns or public safety. SafePath AI aims to fill this gap.
Objectives
• Predict crime probability for a location.
• Generate crime heatmaps.
• Recommend safer routes.
• Provide area safety scores.
• Display nearby emergency services.
Core Modules
1. Crime Data Collection
2. Crime Prediction Engine
3. Crime Heatmap Generation
4. Safe Route Recommendation
5. Area Safety Score
6. Emergency Service Locator
AI/ML Techniques
• Random Forest
• XGBoost
• K-Means Clustering
• Optional LSTM for future crime forecasting
Tech Stack
Frontend: React + Leaflet Maps
Backend: Python FastAPI
ML: Scikit-learn, XGBoost, Pandas
Database: PostgreSQL
Team Distribution
Member 1: Frontend & Maps
Member 2: Backend & APIs
Member 3: ML Models
Member 4: Testing & Documentation
Novel Features
• Women Safety Mode
• Night Travel Risk Analysis
• Community Safety Reports
• Dynamic Safety Index
• AI Chat Assistant
Expected Output
User enters source, destination and travel time. The AI predicts risk level, suggests the safest route,
shows safety score and nearby emergency services.
Future Scope
• Real-time crime feeds
• Voice assistant
• Smart city integration
• Mobile application
• Emergency SOS
Why This Project?
Strong AI/ML content, practical social impact, good for final year major project, publication potential
and strong placement value.