Pothole Patrol — AI-powered road damage detection and mapping system landing page mockup

Live on MUJ → Jaipur Railway Station route

Know what's on
the road before
you hit it.

Google Maps tells you about traffic. We tell you about the road itself. AI-powered pothole detection mapped in real-time — so you slow down before the damage does it for you.

Route · MUJ Campus → Jaipur Railway Station
📍 MUJ Campus 🚉 Jaipur Railway Station
Clear road
Moderate damage — slow down
Heavy damage — be careful
5
damage types detected
45%
confidence threshold
10GB
training data merged
Why this exists

You shouldn't have to discover a pothole by hitting one.

Driving an unfamiliar road in India means playing a guessing game. The data to map road damage exists — it's just trapped in PDFs and dashcam footage nobody processes. Pothole Patrol takes that footage, runs it through a YOLOv8 model trained on 10,000+ annotated road images, and turns raw detections into a colour-coded map anyone can read in seconds. The same intuition as Google Maps traffic layers — applied to the road surface itself.

YOLOv8m detection
Optimised for 30fps dashcam inference. Detects potholes, longitudinal cracks, alligator cracks and more.
GPS-mapped output
Every detection plotted on an interactive Folium map with severity colour-coding — built for real decisions.
Unified dataset
Custom pipeline resolved label conflicts across RDD2022 and Pothole.v1 — 10GB+ of merged, clean training data.