Using Internet of Things and Artificial Intelligence in the Design of an Advanced Smart Shower Space

This paper presents the architecture of an Internet of Things (IoT) and Artificial Intelligence (AI)-enabled smart shower space, designed to assist individuals who have fallen by sending automated SOS calls. The architecture addresses three cases: Case I involves the use of a CZN-15E microphone to detect calls for help, with a Convolutional Neural Network (CNN) employed to differentiate a ‘cry for help’ from pre-processed audio signals. Case II explores the potential of the LM567 IC to detect thuds in the shower. Case III proposes using CNN to classify images of the patron as either ‘Fall’ or ‘Standing’ based on binary cross-entropy. The initial model achieved a validation accuracy of 80 %, but overfitting was observed. Hyperparameters were subsequently adjusted, including changes to dropout rate and the number of epochs, to reduce overfitting. The proposed architecture provides a potential solution for enhancing safety in shower spaces