Tiny-ML and IoT Based Early Covid19 Detection Wearable System

The COVID-19 pandemic has unleashed significant global challenges, straining healthcare systems and impacting millions. Traditional diagnostic methods, such as PCR tests and chest X-ray scans, while effective, face accessibility and scalability issues, particularly in developing regions. Furthermore, these methods do not offer continuous monitoring, critical for early detection and intervention to prevent severe disease progression. Addressing these gaps, our research introduces an innovative AI and IoT-based wearable health monitoring system. Designed for affordability and ease of use, this wearable device integrates vital signs sensors—heart rate, blood oxygen saturation, and body temperature—alongside a small touch screen for symptom reporting. A distinctive aspect of our device is its embedded TinyML model, utilizing an XGBoost algorithm, that facilitates immediate on-device COVID-19 infection prediction. This feature significantly enhances the device’s utility by enabling real-time health status updates and early warning signals for potential COVID-19 infection, which is particularly vital in settings lacking robust healthcare infrastructure. Moreover, by eliminating the need for external data processing, our approach not only ensures data privacy and security but also addresses the critical delay in diagnosis, offering a scalable solution adaptable across various healthcare scenarios. Through this system, we aim to democratize access to COVID-19 screening and monitoring, presenting a significant advancement in wearable healthcare technologies and their application in pandemic response and beyond.