Abstract- Pothole detection isone of the most important tasks for road maintenance. Computer visionapproaches are generally based on either 2D road image analysis or 3D roadsurface modelling. However, these two categories are always used independently.Furthermore, the pothole detection accuracy is still far from satisfactory .Theproposed pothole detection and avoidance system focuses not only on thedetection of potholes but also assist the driver to avoid potholes; by givingdirections towards left or right to avoid the pothole .Moreover, it will alsohave an automatic control over the speed of vehicle which will reduce the speedof the vehicle as the car approaches towards the pothole. All the data will becollected in the local database/cloud. The data will be further used forvarious types of analysis such as the road that has a greater number ofpotholes, tracking the lifeline of the road and prioritize the roadreconstruction.

Keywords:-IR, CNN, MachineLearning