Embedded Multi-Sensor Smartwatch for Computationally Intensive Biosignal Processing
METADATA ONLY
Loading...
Author / Producer
Date
2024
Publication Type
Conference Paper
ETH Bibliography
yes
Citations
Altmetric
METADATA ONLY
Data
Rights / License
Abstract
Over the past decades, the rapid expansion of the wearable market has led to significant advances in integrated devices and technologies, particularly in smartwatches. These devices can unobtrusively acquire various physiological signals such as ECG, EDA, and PPG, resulting ideal for real-time monitoring of vital parameters in medical diagnostics, consumer health, and sports performance analysis. However, existing commercial solutions often do not focus on health-specific applications and face challenges in sensor integration, hardware/software optimization, and onboard signal processing. This paper introduces GAPWatch, a smartwatch-shaped platform with a PCB size of 49×47×1.2 mm based on GAP9 SoC for efficient onboard signal processing and designed to acquire ECG, PPG, inertial data, and 16-channels sEMG. The platform also features an ESP32-based radio module for connectivity and an STM32U5 MCU acting as a gateway. The system is validated in three scenarios: high-rate sEMG acquisition and streaming, low-rate ECG, PPG, and IMU data streaming, and onboard heart rate (HR) estimation using a TCN. In the HR estimation scenario, GAPWatch performs an inference every 2 seconds, transmitting results via BLE while consuming only 10.02 mW, yielding a battery life of over 92 hours.
Permanent link
Publication status
published
Editor
Book title
2024 IEEE Biomedical Circuits and Systems Conference (BioCAS)
Journal / series
Volume
Pages / Article No.
10798302
Publisher
IEEE
Event
20th IEEE Biomedical Circuits and Systems Conference (BioCAS 2024)
Edition / version
Methods
Software
Geographic location
Date collected
Date created
Subject
Embedded systems; Low-power; TinyML; Wearable; Smartwatch
Organisational unit
03996 - Benini, Luca / Benini, Luca