IoT alarm system: detecting & preventing microsleep events

Microsleep events while driving pose a significant risk for road accidents. To address this issue, this project proposed an Internet of Things (IoT) enabled anti-microsleep alarm system for drivers. The system features a Node-Red, Mongo DB and Slack API that is equipped with sensors, a gauge dashboard to monitor driver physiology in real-time, including EAR and Lips Distance, notifications for the driver, voice out alarm with eSpeak, MongoDB for showing timestamp of the driver for yawn and drowsiness. The sensor data is transmitted to an onboard alarm unit that employs machine learning algorithms to analyze the metrics and detect early signs of microsleep episodes. Once the algorithms calculate a drowsiness score exceeding a defined threshold, multi-modal alarms encompassing auditory, visual, and tactile feedback are activated to alert the driver and prompt preventive actions before microsleep commences. The system is designed to be adaptive and customizable based on driver preferences and changing road conditions, allowing for improved detection accuracy and minimal distraction. This project aims to leverage recent advances in IoT and sensor technologies to introduce an intelligent microsleep alarm unit that can significantly enhance road safety by addressing a major yet often overlooked factor in driver fatigue. One of the most important aspects of the project is the integration with the Message Queuing Telemetry Transport (MQTT) protocol, enabling seamless communication and data exchange between the Python-based microsleep alarm system and the Node-RED platform. Node-RED, a powerful visual programming tool used for developing IoT applications, facilitated the creation of an intuitive and informative dashboard.