Senior design, 2022 · Wearable sensing
A noninvasive, battery-powered sensor pod that sits in a fabric wrap and continuously streams an infant’s heart rate, blood oxygen, temperature, breathing and sleep position to a caregiver’s computer, with alarms when a reading crosses a threshold. Aimed at SIDS prevention.
My role
PCB design, Bluetooth data encoding, MATLAB GUI
Team
“5 Guys”, five-person senior design team
Course
BME 452 Senior Design, NC State / UNC Joint BME
Output
Working prototype + provisional patent draft
The assembled double-sided board: ATmega328P, sensor front-ends, LiPo charging over USB-C.
Heart rate & SpO₂
Pulse-oximeter module read through an acrylic window in the enclosure
Body temperature
Non-contact IR temperature sensor behind the same window
Breathing & lung sounds
Electret microphone coupled to a spring-loaded stethoscope head
Sleep position
On-board accelerometer tracks body orientation
Wireless
Packaged sensor data streamed over Bluetooth to a monitoring computer
Power
LiPo battery with USB-C charging, regulator and status LEDs
The electronics sit in a sealed plastic pod: a rubber gasket and stethoscope cap keep moisture out, an acrylic window gives the optical and IR sensors a clear view, and a sprung platform presses the stethoscope head against the body. The pod slides into a fabric wrap whose snap buttons line the sensors up with the right spot on the infant.
Exploded view of the sensor enclosure.
The pod embedded in the swaddle.
An ATmega328P ties together six modules: accelerometer, microphone front-end, pulse oximeter, IR temperature sensor and the power stage that charges the LiPo battery and regulates the supply. The board is double-sided, with the optical sensors and microphone port on one face and the controller, accelerometer, USB-C and data ports on the other.
Layout, controller side.
Layout, sensor side, with the stethoscope coupling.
Module-level schematic around the microcontroller.
On the board, the firmware manages charging states and status LEDs, reads every sensor, packages the samples with a compact encoding scheme and sends them over Bluetooth. On the computer, a MATLAB program unpacks each stream, uses open-source algorithms for breathing-cycle detection and pulse oximetry, logs lung sounds, and raises audio and visual alarms with a timestamp whenever position, breathing, temperature, heart rate or SpO₂ leaves its safe range.
Microcontroller program.
MATLAB monitoring program with alarm thresholds.
As part of the course, the team drafted a provisional patent for the system. Its claims cover:
© 2026 Mohammad Riahi
Raleigh, NC