Embedded based Smart Accident Pre-Alert and Prevention System with Machine Learning

Accidents in the world increases every day in different ways. In this scenario, 50 percent of accidents were caused by the driver's inattention. Driver play as a heart of the vehicle; therefore, the driver needs to be in steady and active state during the travel. At times, the drivers used to work for more than the usual timing, which leads to drowsiness causing the base for accident. In case of goods transportation for longer distance via road ways, drivers used to consume alcohol to overcome tiredness leads to accident. This paper explains about the pre-alert system and prevention of accidents by drowsiness and alcohol consumption. The hardware model incorporates Arduino Uno board. The pre alert system uses webcam to detect drowsiness by monitoring the face of the driver using object detection algorithm and MQ3 sensor placed in front of driver senses the alcohol consumption. If the threshold exceeds and alcohol and drowsiness is detected, immediately the proposed system uses to alert the driver until driver comes to active state. At that situation the model identifies the current location and intimates to the owner/blood hood through short message service. If the vehicle undergoes an accident, the vibration sensor detects it and sends a message to the emergency contact.