Wearable IoT Electronic Nose for Urinary Incontinence Detection

This paper presents the development of a low cost and low power wearable Electronic Nose with IoT connectivity for detecting synthetic urine. This system was designed with the ultimate aim to relieve the burdens associated with urinary incontinence for both sufferers and healthcare workers in hospitals and care homes. Commercially proven metal oxide (MOX) gas sensors for Volatile Organic Compound (VOC) detection such as the SGP30 and MiCS-5524 series of gas sensors were used, as well as the custom-made WO
₃
, SnO
₂
and CuO MOX sensors developed at the Microsensors and Bioelectronics laboratory at the University of Warwick. A web-application was created to facilitate real-time monitoring over the MQTT protocol and indicate when urine has been detected. The capabilities of the IoT for this purpose allow for enhanced patient mobility and simultaneous patient monitoring. An artificial neural network was used to classify synthetic urine from other odours. The model was able to identify synthetic urine with 100% accuracy based on prior training and testing datasets.