Awards:
Best Poster Award in ECE at UC San Diego’s Research Expo 2025
Poster:
Contributions:
Overview of Fabrication Challenges I Solved
| Problem | Solution |
|---|---|
| Sensor-to-sensor operating frequency would vary dramatically | Decreased cavity size, allowing more of the sensor to be surrounded by the polymer to maintain a more uniform shape |
| Sensor was not sensitive enough at baby-level forces (0–~2 N) | Changed material type to Ecoflex-OO10 (this change is possible due to smaller cavity size = more material, which allowed for a more flexible polymer (10A to OO10)) |
| Air bubbles forming on sensor mold | Designed and 3D printed pacifier holder that was fitted with an airtight seal for degassing chamber |
| Air bubbles between the mold and pacifier (bad adhesion) | Use isopropyl alcohol to wipe off residual “man” spray from mold fab step |
| No current fabrication that supports vacuum pressure sensing for our sensors | Interfaced feeding tube with current force pacifier sensor to create a dual pressure and vacuum sensor |
Test setup
With this setup I was able to collect and analyze force vs. phase data through a host PC using Python scripting. With this data, I was able to improve sensor quality.
Improved Fabrication of Sensor
- Enhanced performance and manufacturability:
- Increased sensor batch yield by 80% by improving polymer adhesion techniques and eliminating air bubbles
- Improved sensor consistency by reducing operating frequency standard deviation by 80% through optimized polymer-based geometry
- Reduced fabrication time by 25% by streamlining processes and eliminating redundant steps
- Increased sensitivity by 167% by selecting and optimizing polymer material composition
- Testing and validation:
- Conducted vacuum pressure performance testing to benchmark the sensor against industry-standard pacifier sensors
→ Validated dual functionality for both force and vacuum pressure sensing - Built and tested multiple prototypes using PCBs and 3D-printed fixtures; developed Python scripts to control VNA/Arduino/force actuator, automate S11 data collection while step forces applied, and generate calibration curves for sensitivity/linearity analysis.
- Conducted vacuum pressure performance testing to benchmark the sensor against industry-standard pacifier sensors
- Automation and data processing tools:
- Developed a Python script to:
- Communicate with a VNA, Arduino, and linear actuator
- Apply step forces and collect S11 parameter data
- Generate phase vs. force calibration curves for performance analysis
- Compare sensor performance across individual sensors, within the same batch or across multiple batches
- Created an algorithm to:
- Automatically identify the optimal operating frequency (maximizing phase change)
- Ensure linearity and control for differential magnitude > −5 dB
- Developed a Python script to:
On the left is an old fabrication process sensor with no response between 0–~2 N, which is the active force range of a newborn. On the right is the improved fabrication process sensor with a linear force vs. phase response from 0–~4 N, with about 10 degrees phase change per 1 N of force.
PCB Design
Tested transmission line (TL) flex PCB in Ansys HFSS and designed in Altium. TL was designed for 900 MHz.