During the recent times, increase of fatalities due to vehicle accidents has extended remarkably [1]. These accidents were due to the result of drowsy drivers behind the wheels. Drivers who have sleep insufficiency or who may fall asleep while driving put both themselves and other passengers in danger [2]. For that reason, the examination of driver conduct on the roads evolved into one of the trending research subjects. The other techniques involved in this study require the driver to put on health sensors. But this technique is falling short of success as the sensors should be in specific positions for well founded results. The main theme of this project is to use the Machine Learning and IOT to design and implement intuitive alert machine that is suitable to identify drowsiness among drivers and send alerts for them in order to mitigate the risks. Hence this project suggests to design a driver drowsiness detection system using a behavioural technique [3] to alert driver. This System provides facilities like Real-Time Detection, Non-Invasive Approach, High Accuracy, Low Cost and Scalability, Reduced False Positives and reduces unnecessary alarms in order to ensure timely responses when drowsiness is genuinely detected [4]. This system can be used in Fleet Management, Automotive Industry, Medical field, Mining and Construction, Transportation and Logistics. In this system Convolutional Neural Network model is used to evaluate in case if the driver is feeling drowsy or whether he is awake. This model achieves an accuracy of 95%, which facilitates to design an instantaneous driver monitoring system to avoid road accidents.