emerald01-tech/Nigeria-RoadForecast-AI
Domain:
mobilityRecord type:
model
Creator:
eme
Host:
AI-powered traffic crash prediction system using Nigerian FRSC data (2020-2024)
# Nigeria-RoadForecast-AI
---
An AI-powered solution to reduce road fatalities using FRSC crash data (2020-2024)
---
🚗 Predicts accident risks by road type, weather, and location
📊 Analyzes 43,262+ historical crashes with 84% accuracy
🛠 Built with: Python, Random Forest AI, and Tkinter GUI
# Nigeria-RoadForecast-AI Dashboard
---
## See Live Dashboard
Live xlx Dashboard
## Project Overview
An intelligent system that predicts high-risk road conditions using 5 years of FRSC crash data (2020-2024). Designed for drivers, policymakers, and transport agencies to proactively reduce accidents.
## 📌 Key Features & Visuals
| Feature | Specification |
|-----------------------|------------------------|
| Algorithm | Random Forest |
| Accuracy | 84% (F1-score) |
| Input Features | RoadType, Weather, LightCondition |
| Prediction Output | Fatality Risk (0-100%) |
## 📌Feature Impact & Dashboard KPIs Visual Reference (Mockup)
| Feature | Specification |
|------------------------------|------------------------------------------|
| 📊 Crash Hotspot Mapping | Flags Top 3 states: |
| | - Lagos (11,264 crashes) |
| | - Oyo (7,401) |
| | - FCT (8,805) |
| 🌧️ Weather Risk Alerts | Predicts 38% of crashes occur in rain |
| | 24% in fog |
| 🛣️ Road-Type Analysis | Highways = 52% fatalities |
| | Urban roads = 30% |
| | Rural roads = 18% |
| 🛣️ Road-Type Analysis | Processes 5,184 Q1-2024 crashes in <1sec|
| ⏱️ Real-Time GUI | 84% prediction accuracy |
## Data Cleaning & Trans …