Pothole Patrol.
Turning a road photograph into a clearer picture of the repair work ahead.
01 / The experience
A closer look.
Actual screens.
From the idea to the interface.
Finding the damage
Detected defects are highlighted in the uploaded road photograph.
02 / Behind the build
The thinking
underneath.
What it does
An AI-assisted road inspection tool that identifies potholes in uploaded images and estimates the material needed for repairs. It connects visual detection to an engineering question: what would it take to fill the damage?
How it comes together
A Mask R-CNN computer vision pipeline detects road defects. User-provided road width and expected depth turn image measurements into estimated volume and asphalt requirements. Those outputs depend on the input assumptions; they are estimates rather than field measurements.
The cloud-hosted demo may need a moment to wake up.