I built the BioWatch during my second year of my master's degree at the De Vinci Innovation Center. My objective was to create a watch from scratch in the continuity of my work on wearable technologies. So the BioWatch is the wearable that allowed me to implement the biosensors I developed during my master's degree.

Wearable biosensors are becoming increasingly advanced yearly, while smartwatches are increasingly functional. The BioWatch is an open-source prototype smartwatch for the implementation of wearable biosensors. This project is part of the trend of private companies to develop molecular monitoring as a new functionality of smartwatches.

BioWatch

DeVinci-Innovation-Center ⋅ 6 months ago

Introduction

Wearable technology is undergoing constant development, entrenching our daily lives to capture rich contextual information produced for a personalized experience [1]. Smartwatches were the first socially accepted wearable devices and are still the most popular. In 2016, smartwatch sales in the wearables market were the second product over intelligent devices, with 50 million units sold. They cover a broad application scope, including connectivity, sports, and health [2]. Human activity detection for well-being applications has become a central stake in the smartwatch and fitness tracker market. Heart rate, temperature, and blood pressure are examples of already monitored physiological data by commercial wearable devices [3]. Wearable biosensor technology has advanced significantly [4], but significant obstacles remain to overcome their full integration into our smartwatches regarding new materials, power sources, and communications systems [5].

Related Works

Few works on watch-integrated biosensors and multianalyte measurement devices have been recently published. Kim et al. developed in 2018 a wearable and flexible patch that monitors two analytes in two different biofluids: alcohol in sweat and glucose in interstitial fluid (ISF) [6]. In 2022, Wang et al. presented a sweat cortisol sensor with an LCD screen that prints in real-time the measured concentrations [7]. Finally, the French company PKVitality (Paris, France) developed the K'Watch Glucose® and the K'Watch Athlete®, which can respectively monitor glucose and lactate levels thanks to minimally invasive biosensors [8]. Thus microneedles-based biosensors access the interstitial fluid without any sensation of pain or discomfort. The smartwatch tracks the wearer's glucose or lactate level in real-time for seven days before replacing the microneedles module with another one. The wearable seems promising in modularity and compatibility with all minimally-invasive electrochemical biosensors and multi-analyte measurements.

Contribution

The BioWatch is a smartwatch prototype for developing wearable biosensors and real-time data visualization. This electronic project aims to easily demonstrate proof of concept of smartwatch-integrated or wirelessly connected wearable biosensors. It allows rapid implementation and on-body demonstration of non-invasive (sweat) or minimally invasive (ISF) electrochemical biosensors. Beyond electrochemical sensors, BioWatch's contribution extends to the development and proof of concept of any wearable sensor connected to or implemented in a smartwatch.

The BioWatch is also a learning platform for IoT technology for students at the De Vinci Innovation Center. Master students are working on this project as part of their training to learn by developing new functionalities for the BioWatch.

Functionalities

The system has five main functionalities: over-the-air programmability, Wi-Fi server-client connection connectivity, sensor data visualization, sensor modularity, and rechargeability.

• With over-the-air programmability, users can remotely update and modify the software on the device, which allows for easy integration of new features and bug fixes.
• The Wi-Fi server-client connection allows easy access to data from the device, enabling users to monitor sensor readings in real time.
• Sensor modularity is an essential feature of the BioWatch, as it allows for easy replacement and upgrading of sensors.
• The device includes a rechargeable battery to ensure 3 hours of autonomy.

Hardware

The display of the bioWatch is a 1.28" round RGB LCD module from WaveShare with an integrated GC9A01 driver. It communicates with Serial Peripheral Interface (SPI) bus protocol.

• In the first version of the BioWatch, the display is controlled by an Arduino WeMos D1 mini Lite. The WeMos D1 Mini is a microcontroller board based on the ESP8266 Wi-Fi module. The board is particularly small (34 x 25 mm), and suitable for wearable applications.
• The battery shield, soldered on the Wemos, allows to power the board with a 3.7V LiPo battery thanks to a mini JST connector. The battery can be charged via the micro-USB port on the shield. Two red and green LEDs indicate when the battery is charging and when the charge is complete.
• The watch case is modeled in CAD software and 3D printed with PLA. Magnets of 2 millimeters in diameter are integrated into the case. Thus, a sensor can easily be positioned, such as a micro-needle module.

PCB Design

The second version of the BioWatch has a PCB explicitly designed for the BioWatch. It uses an ESP32 Wroom, a compact and low-power chip. The ESP32 Wroom module includes a dual-core Tensilica LX6 microcontroller with up to 240 MHz clock speeds and various peripheral interfaces such as SPI, I2C, UART, and ADC. It also includes Wi-Fi and Bluetooth 4.2 connectivity, making it an ideal choice for wearables applications.

UART pinout allows programming the ESP32. Seven digital pins are provided for programming the GC9A01 driver in SPI. A LiPo battery can power the chip via the pin reserved. A switch button allows to turn on and off the smartwatch. A charging circuitry similar to the Wemos battery shield enables charging the battery through the micro-USB connector.

User Interface Design

The display has been designed based on the preferences of diabetic patients. The BioWatch interface should ensure a great user experience for diabetes to monitor their glucose levels and for other patients to track health data. Three patients with type 1 diabetes, including two adults and one child, were consulted for their preferences. They reported wanting three pieces of information on the BioWatch:

• The last glucose concentration value in mg/dL,
• The evolution of their glucose level over the last hour,
• A visual indicator for identifying hypoglycemic (<70mg/dL) and hyperglycemic (>140mg/dL) situations.

Based on this consultation, four designs were proposed to 30 healthy subjects who evaluated the designs based on aesthetics and intuitiveness. Design number 2 received the highest score.

Software

The collection of user requirements was the basis for developing the watch software. The Arduino GFX library is used to control the screen graphics. The measured data are converted into coordinates of a pixel on the screen (240 x 240 pixels), allowing to draw a history curve. An algorithm displays a prediction of the next value.

The clock is synchronized with the network thanks to an NTP connection, allowing to display the time and date. The update frequency can be easily updated, between one second for pulse sensor applications and one minute for biosensors. The conversion of the measured data into coordinates on the screen is also easily adaptable. The BioWatch is thus compatible with all types of sensors implemented in I2C or connected wirelessly.