This article focuses on the research and development of a model for a drowsiness detection and monitoring system while driving based on computer vision. The product can be applied to current vehicles (such as cars or vehicles with similar functions). The study utilizes several state-of-the-art sleep detection models to collect and process data and train the model to provide quantitative results. The model is evaluated through this process, and improvements are proposed to make it suitable for vehicles in Vietnam. Experimental results show that the system can capture images via a camera, detect human faces, predict sleep states, and send alerts to a server, displaying warning data on a website interface. Essential evaluations indicate that the system operates stably and relatively accurately according to the pre-established design and programming requirements. The results of this research can be applied in practical scenarios or serve as a practical model for students in artificial intelligence, robotics, and automation control fields to enhance their ability to apply theoretical knowledge to real-world problem-solving.