Improving Health Monitoring Based on Smartwatches with Advanced Sensors and TinyML
Year: 2025
Type: article
Abstract: HWatch, a revolutionary smartwatch, addresses sensor constraints, cloud dependence,and high-power consumption in wearable health systems. It has three sensors: theMAX30100 for heart rate and SpO2, the MAX30205 for body temperature, and an IMU formotion tracking. The RP2040 microcontroller efficiently collects and processes data fromthese sensors. HWatch uses TinyML for real-time health analysis without cloudinfrastructure, improving privacy and reaction time. The device has 92% cardiacabnormality detection accuracy and 90% sleep monitoring accuracy with 50 ms delay and200 mW power consumption. HWatch monitors hypoxia, arrhythmias, fever, and sleepquality 24/7, making it a trustworthy tool for preventive health management. For real-timehealth insights without regular recharging, the system's edge-based processing and lowpower utilization make it efficient and practical. HWatch costs and prioritizes privacy overMedAi, which requires 11 sensors and cloud computing. Its small size and simplisticdesign make it scalable and accurate for disease detection and sleep analysis. In conclusion,HWatch is a breakthrough in wearable health technology, providing a simple andresource-efficient continuous health monitoring solution. Hardware validation, diseasedetection, and functionality improvements are planned. The proposed system determineshealth anomalies based on biomarkers (SPO2, Body Temperature, Body Motion) and sleeptracking accuracy of 75%, early disease diagnosis 99.99%, and user lifestyle using TINYMachine Learning Technology.
Authors: Chandini Mutta, Sk.Shameer, S. Meghana, N.S.R. Santosh, B.B.N.S.V. Anil, P. Munna, B. Anushka
Language: English
Subfield: Biomedical Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6930867753
FWCI:
Cites:
Cited by:
Related to: