Stress Detector System Using IoT and Artificial Intelligence

This work presents the design and implementation of an IoT stress detection and classification system. Three sensors, a skin conductance sensor, an ECG sensor, and a simple skin temperature sensor are integrated into a wearable device for measurement of physiological features. The measurements are communicated to a cloud server through the user's mobile phone. On the cloud, Artificial Intelligence algorithms analyze the sensor data and determine the user's current state of stress. The predicted state is fed back to the user's mobile for display and suggestion of stress-relieving activities. In case of emergency stress levels, a message is forwarded to the physician who can access the data through the cloud server's interface. The system is able to achieve 97.6% accuracy binary classification based on real-time sensor data.