The implementation of a smart safety protection system using DL and IoT technologies

Safety has always been a concern. Thus, employing and utilizing technology is vital to help society address this concern. This paper looks at the design and implementation of a Smart Safety Protection System (SSPS), which employs the Internet of Things (IoT) to solve the issue of real-time monitoring, geolocation tracking, and emergency alert features. The proposed system is a wearable device that uses sensors and GPS to send continuous data to both an application on smartphones and a database on the cloud. The system's features include detecting unauthorized removal of the device as well as falls and abnormal temperature or heart rate. If anything happens as mentioned, the system will send alerts to the linked mobile application. A Deep Learning (DL) model is integrated into the SSPS system using multimodal WESAD sensor data. A trained Multi-Layer Perceptron (MLP) neural network achieved 93% accuracy when optimized with Adam and Softmax, enabling reliable identification of Monitored Individuals' emotional states. Steps of functional and integration testing of the prototype proved the reliability of sensor accuracy, data transmission speed, and real-time emergency response. Ultimately, this work emphasizes the potential of using DL and the IoT to improve Monitored Individuals' safety.