Driver drowsiness is the most critical cause of road accidents so detection of drowsiness play a vital role in preventing road accidents. We are developing an android app that will alert drivers before an accident occurs. This will reduce the number of road accidents on a road. Drowsiness is a natural phenomenon that happens in human body due to different factors. Machine learning was applied to predict drowsiness and improve drowsiness prediction using facial recognition technology and eye-blink recognition technology. In this app, front camera will take a picture of drowsy driver then this picture will be taken as input. In processing the detected image, we are using OpenCV Library. OpenCV Library uses Haar Cascade Classifier for detection images such as eyes and face. Eyes and face will be the target in this system. This application will be implemented on Android Operating System. Drowsiness detection system will send alert to the driver when the driver feels asleep while driving a car, this can avoid accidents. Driver which is the user in this application, if they close their eyes within one second, the sensor which is the front camera in the smartphone will catch and process this event and then trigger the system to give voice alert to the user. Moreover, if the driver is willing to turn on back camera then it will detect the lane detection violation and will calculate the distance from the vehicle ahead of it. If the distance is too close, then it will generate an alarm. It will also generate an alarm if there is a violation of the lane on the road.