One of the major problems in developing countries is monitoring the road and traffic condition. Potholes on the road can affect the large proportion of society either directly or indirectly. Getting the pothole fixed, informing the authorities and asking them to take action is very time consuming. Implementing a system that involves citizens in the process of detecting pothole is what is being envisioned. many image processing techniques were proposed to inspect pavement defect for improving accuracy and efficiency for the onsite inspection by the humans but it gives you unacceptable stability. Therefore, an application of convolutional neural network is used for pothole detection using digital images in this study. This paper concludes earlier pothole detection methods that have been developed and suggest a cost-effective solution to detect the potholes on the roads and provides timely alerts to drivers to avoid critical problem that causes numerous car accidents annually is driver fatigue. Due to incapacity of driver to halt or minimize the impact accidents caused by driver’s sleepiness are much more inclined to result in fatalities or severe accidents. this paper focuses on accident detection and avoidance system which detect accidents and alerts the driver by buzzer. It calculates the eye open ratio by using eye blinking algorithm.