Wearable Adhesive Biosensor Utilizing Electrospun PVDF Nanofibers Using ESP32-S3

Commonly used traditional wearable sensors like optical PPG (Photoplethysmogram) sensors located in smartwatches do not always work well due to issues related to "motion artifacts" — interference generated by the user moving — making it difficult to accurately capture data from gentle mechanical vibrations created by the beating heart (seismocardiography) and tiny amounts of muscle activity generated by slight movements.

In order to accurately measure and capture these two examples mentioned above, an exceptionally sensitive level of detection of mechanical vibrations is required that will so rarely deliver using normal silicon-based sensors without significant bulk.

Project Goals:

• Develop a flexible, skin-conformable sensing interface.
• Achieve microvolt-level signal resolution.
• Design a modular architecture to reduce "e-waste" and per-use costs.
• Ensure wireless data integrity for real-time clinical monitoring.

The core innovation of this project lies in its multi-layered sensing membrane. To achieve the required sensitivity, we moved beyond off-the-shelf components to custom-fabricated nanomaterials.

A. Piezoelectric Nanofiber Membrane

We utilized Electrospun Polyvinylidene Fluoride (PVDF). PVDF polymer is electrospun into highly fine (diameter of several nanometers) fibers.

This allows for a high surface area to volume ratio, meaning that even small changes in mechanical deformation are extremely detectable with this structure.

As a result of its highly flexible and breathable characteristics, the fiber mesh of this device can be used as a true alternative to conventional rigid piezo-ceramic materials.

B. Hybrid Electrode Integration

To extract signals from the PVDF layer, we printed Silver Nanowire (AgNW) electrodes in a serpentine pattern. This specific geometry allows the electrodes to stretch and flex without losing conductivity.

Additionally, a carbon-ink Force-Sensitive Resistor (FSR) grid was layered beneath the piezo-membrane. This allowed the system to measure the "contact pressure" of the device against the skin, providing a baseline to filter out false signals caused by the device shifting.

C. Substrate and Acoustic Coupling

The sensor was mounted on a hybrid PET/TPU substrate. To ensure the mechanical vibrations from the heart and lungs reached the sensor without loss of energy, we utilized PDMS (Polydimethylsiloxane) as an acoustic coupling agent. This material's mechanical impedance is tuned to match human skin, acting as a "bridge" for physiological signals.

Capturing a microvolt signal is a wasted effort if the electronics introduce "hiss" or quantization errors. The electronics pod was designed as a reusable "brain" that snaps onto the disposable sensing strip.

The Analog Front-End (AFE)

To achieve this goal we utilized a highly sophisticated signal chain that utilized an INA128 Instrumentation Amplifier as its primary component. The INA128 Instrumentation Amplifier is superior due to its exceptionally high Common-Mode Rejection Ratio (CMRR) - CMRR is extremely important when trying to eliminate the 50/60Hz electrical noise present in every modern building.

The resulting amplified analog signal was inputted to the ADS1256, a 24-bit Delta-Sigma A/D converter. The decision to use 24-bit resolution instead of the more conventional 8- or 10-bit A/D conversion in most microcontrollers allowed us to preserve the very fine details associated with the mechanical opening and closing of the heart's mechanical valves.

Computation and Connectivity

ESP32-S3 is the heart of the processing system, the two cores give us one core for the fastest possible data acquisition via SPI and another core for doing:

• DSP - real time implementation of Notch and Bandpass filters
• BLE 5.0 streaming - packaging data for wireless transmission to the mobile app or gateway
• Power Management - managing the LiPo battery and the integrated NFC antenna to retrieve data passively.

Raw data from a piezoelectric sensor is "dirty." To make it "investor-ready," we developed a proprietary DSP pipeline:

• Baseline Wander Removal: Using a high-pass filter to remove the slow "drifting" of the signal caused by the user's breathing.
• Wavelet Denoising: This advanced mathematical technique separates the "sharp" peaks of a heartbeat from the "random" noise of muscle movement (EMG).
• Peak Detection Algorithms: On-device processing calculated Heart Rate Variability (HRV) and Respiratory Rate in real-time, reducing the bandwidth needed for transmission.

The project has used a rigorous V-Model development process to achieve the medical-grade reliability of the system.

Phase I: Material Validation

Test PVDF nanofibers at controlled mechanical loads on a shaker table. This will allow for characterization of the voltage-to-pressure ratio.

Phase II: Signal Chain Optimization

Iterate PCB layout to minimize trace length for EMI and the possibility of them acting as antennas for the signals.

Phase III: Form Factor Integration

Transferring the electronics to a compact biocompatible "pod." The pod uses gold-plated snap connectors to maintain an electrical connection to the disposable strip reliably.

Phase IV: Human Subject Trials

Testing the prototype with human subjects in 3 different conditions: resting, walking and talking. This will validate the performance of the PDMS coupling and the DSP noise rejection.

The final deliverable from the project is a functioning, high fidelity prototype of our product. The SCG and BCG (e.g., captured by our device) are now present in sufficient clarity to approach that of a benchtop type laboratory apparatus.

Key Deliverables Achieved:

• Functioning hardware (reusable electronics housing and disposable adhesive sensors)
• Technical manuals (detailed design specifications including schematics, gerber files, and bills of materials for scalability)
• Firmware stack (documented C++ code for an ESP32-S3 with modular drivers for ADS1256 and Bluetooth Low Energy services)
• Market viability (affordable architecture where the reusable expensive electronic components are utilized, whilst only the inexpensive bio-interface is discarded, thus achieving the ideal balance from a healthcare economic perspective).

The application of advanced nano-free technology combined with high-precision electronics confirmed that wearable sensors can be comfortable and clinically accurate.

The utilization of nano PVDF fibres enables the development of new types of "smart adhesives" that will allow for the measurement of many different variables including athletic performance and recovery after surgery without the use of bulky straps or conductive gels.

The prototype demonstrates the effectiveness of collaborative engineering, applying elements from various disciplines including material science, analog electronics, and computer software, towards finding solutions to medically related problems in everyday life.