The aim is to automate the process of manual health inspection and report generation of roads in Karachi. A device attached to a moving vehicle that captures geo tagged images of the road and sends them to the cloud. The data is passed through a Convolutional Neural Network based detector that detects and classifies the road damage into five categories (i.e. line cracks, potholes, alligator cracks, raveling and no damage). After the damages are detected, they are displayed on a map. A GUI is created that allows report generation of a selected road. It also gives the road priority for repair based on the area selected.