Smart Cradle Monitoring System with Infant Cry Detection Using AI

The issue of infant safety in sleep is a burning topic, as the traditional baby monitors imply only simple audio or video surveillance with no evaluation of physiological wellbeing or the analysis of behavioral indicators. In the given paper, the Smart Cradle Monitoring System with AI-based Infant Cry Detection is proposed, which combines the real-time physiological monitoring, environmental monitoring with intelligent infant crying analysis and automated cradle control in one Internet of Things (IoT) platform. The system measures vital parameters such as the heart rate, blood oxygen saturation (SpO 2), and body temperature with the help of biomedical sensors, whereas the environmental factors of temperature and humidity and the intensity of light are measured with help of special sensors. A cry classification model, which is an artificial intelligence-based model created on the basis of a Convolutional Neural Network (CNN), is an analysis of infant cry patterns that are subsequently divided into a category of hunger, pain, discomfort, or sleep-related cry. The cradle is activated to perform automated soothing movements based on the information of the fused sensor data and AI output, and realtime alerts and notifications are sent to the caregivers via a cloud-enabled dashboard. The experiments show that there is a high classification accuracy and high sensing performance in real-time conditions. The suggested system is a low-cost, scalable, non-invasive, and intelligent system that contributes to infant safety significantly as well as proactive caregiving and overcomes the drawbacks of the classical monitoring systems.